Insights in the effects of an admissions schedule on the wards

Size: px
Start display at page:

Download "Insights in the effects of an admissions schedule on the wards"

Transcription

1 Insighs in he effecs of an admissions schedule on he ards Auhor: Supervisors: Annemaaike Hooijsma B. van den Akker (Isala Klinieken) Dr. ir. I.M.H. Vliegen (UT) Dr. ir. E.W. Hans (UT) Dae: 10 Sepember 2012

2 1

3 Managemen summary Background In Isala Klinieken he processes a he ards are negaively affeced by oher processes in he hospial. The admissions schedule is supposed o be he main reason for his. The deparmen Paien Logisics ihin Isala Klinieken already performed research abou his on a acical level. The curren research akes he nex sep by broadening he scope o he operaional offline level. The objecive of his research is: Generaing insighs in he effecs of operaional offline scheduling decisions made for hospial admissions on he capaciy planning processes a he ards. Approach To generae hese insighs a ool is (parly) developed o sho he admissions planners he qualiy of heir schedule. This research consiss of muliple seps. Firs, e consruced a lis of performance measures hich are imporan for he ards and should herefore be shon o he admissions planners. We based his lis on inervies and a survey on all hierarchical levels ihin Isala Klinieken. This resuled in a long lis of performance measures. Therefore, e reduced he lis by using evaluaion crieria and by looking a he imporance based on he inervies and he survey. Secondly, e inervieed muliple planners and he deparmen Paien Logisics o deermine ho hese performance measures should be displayed for he admissions planners. The hird par of his research focused on ho o forecas he informaion needed o calculae he performance measures. We parly based his on models from lieraure. Finally, e developed par of he ool according o he sofare developmen mehod Exreme Programming. Conclusions The performance measures ha should be shon o he admissions planners are: Average uilizaion per ard Variabiliy in he number of used beds over he eek Saff produciviy Variabiliy in amoun of ork during he day Consan paern in amoun of ork over he eek Variabiliy in admissions during he day Variabiliy in admissions over he eek Saff availabiliy Overuilizaion per ard Overuilizaion on cerain momens of he day A furher requiremen of he ool is ha i should summarize hese measures ino a raing for he oal qualiy of he admissions schedule. Nex o his he ool should sho he rends of he oal qualiy and of he performance measures. Anoher imporan requiremen for he ool is ha i should be flexible. This is ranslaed ino a possibiliy for admissions planners o change muliple seings in he ool. During his research e developed par of he ool, namely he resuls for he admissions over he eek and for he overuilizaion. For he calculaions of he overuilizaion e build upon he acical model of Smeenk (2011). 2

4 Recommendaions We recommend Isala Klinieken o coninue developing he proposed ool and o implemen i. Nex o his, e recommend changing he focus in he organizaion from he operaing rooms more o he ards. Our final recommendaion is o le he admissions for one ard be scheduled by one admissions planner or scheduling eam. Currenly he admissions are scheduled per specialism. Especially for he ne hospial ih larger ards his ould mean ha muliple planners schedule for he same ard. Topics for fuure research are: ho o change he focus ihin Isala Klinieken from he OR deparmen o he ards generaing insighs in oher resources han ards, like for example ORs and IC examining if here is correlaion beeen he performance measures improving he inpu for he ool improving he mehod for inegraing he performance measures ino a raing for he oal qualiy of he schedule 3

5 Managemen samenvaing (Duch) Achergrond Binnen de Isala Klinieken orden de processen op de verpleegafdelingen negaief beïnvloed door andere processen binnen he ziekenhuis. De opnameplanning ord gezien als de voornaamse reden hiervoor. De afdeling Paiënen logisiek van de Isala Klinieken heef hier al onderzoek naar gedaan op een acisch niveau. Di onderzoek gaa hiermee verder op he operaioneel offline niveau. He doel van di onderzoek is: He genereren van inzichen in de effecen van operaioneel offline beslissingen me berekking o de opnameplanning op de planningsprocessen op de verpleegafdelingen. Aanpak Voor he genereren van deze inzichen is een ool gecreëerd die de planners de kaliei van hun planning laa zien. Di onderzoek besaa ui een aanal sappen. Ten eerse hebben e een lijs opgeseld me belangrijke indicaoren voor de verpleegafdelingen die geoond moeen orden aan de opnameplanners. Deze lijs hebben e gebaseerd op inervies en een enquêe op alle hiërarchische niveaus binnen de Isala Klinieken. Di resuleerde in een lange lijs me presaie indicaoren. Daarom hebben e deze lijs gereduceerd door gebruik e maken van evaluaie crieria en door e kijken naar he belang van de indicaoren blijkend ui de inervies en de enquêe. Ten eede hebben e meerdere planners en de afdeling Paiënen logisiek geïnervied om e bepalen hoe deze presaie indicaoren eergegeven moeen orden aan de opnameplanners. He derde deel van he onderzoek rich zich op hoe de informaie die nodig is voor he berekenen van de indicaoren kan orden voorspeld. Di baseren e gedeelelijk op modellen ui de lierauur. Uieindelijk hebben e een deel van de ool onikkeld door gebruik e maken van de sofare onikkelingsmehode Exreme Programming. Conclusies De presaie indicaoren die aan de opname planners moeen orden geoond, zijn: Gemiddelde bezeing per verpleegafdeling Variabiliei in he aanal gebruike bedden over de eek Produciviei personeel Variabiliei in voorhanden erk gedurende de dag Consan paroon in voorhanden erk over de eek Variabiliei in opnames gedurende de dag Variabiliei in opnames over de eek Beschikbaarheid personeel Overbezeing per verpleegafdeling Overbezeing op bepaalde momenen van de dag Een exra eis aan de ool is da he deze indicaoren moe samenvaen in een cijfer voor de oale kaliei van de opnameplanning. Hiernaas moe de ool een rend voor de oale kaliei en voor de presaie indicaoren eergeven. Een andere belangrijke eis is da de ool flexibel is. Di is veraald in een mogelijkheid voor de opnameplanners om meerdere insellingen in de ool aan e passen. Tijdens di onderzoek hebben e een deel van de ool onikkeld, namelijk de resulaen 4

6 voor opnames over de eek en voor overbezeing. Voor de berekeningen van overbezeing hebben e he acische model van Smeenk (2011) verder onikkeld. Aanbevelingen We bevelen Isala Klinieken aan om door e gaan me he onikkelen van de ool en deze e implemeneren. Daarnaas bevelen e aan om de focus binnen de organisaie e verplaasen van de operaiekamers meer riching de verpleegafdelingen. Onze laase aanbeveling is om de opnames voor één afdeling e laen plannen door één opnameplanner of eam. Op di momen orden de opnames gepland per specialisme. Vooral voor he nieue ziekenhuis me groere verpleegafdelingen zal di beekenen da meerdere planners voor dezelfde afdeling plannen. Ondererpen voor verder onderzoek zijn: hoe de focus binnen Isala Klinieken moe orden verplaas van de OK afdeling naar de verpleegafdelingen he genereren van inzichen in andere middelen dan de verpleegafdeling, zoals bijvoorbeeld de OKs en de IC onderzoeken of er correlaie is ussen de indicaoren he verbeeren van de invoer van de ool he verbeeren van de mehode voor he samenvoegen van de presaie indicaoren in een cijfer voor de oale kaliei van de planning 5

7 Preface When I sared my sudy Indusrial Engineering and Managemen six years ago I never expeced o end up in healh care. Afer geing acquained ih his fascinaing discipline by chance during my Bachelor hesis, I couldn le go of i. When searching for a Maser hesis, i as clear ha i should be in a hospial. The enhusiasm of Erin and Ingrid abou Isala Klinieken and he ideal locaion of Zolle made he choice for Isala Klinieken easy. A fe monhs laer I moved o Zolle and I did no regre i for even a minue. I an o use his opporuniy o hank some people for heir help and suppor during hese las fe monhs. Firs of all I an o hank Bernd. Your innovaive ideas and our ineresing conversaions brough his research o a higher level. Every problem could be solved ih one of our discussions. Wha made my ime ihin Isala exra special ere he opporuniies you gave me o experience healh care logisics in pracice. Thank you for ha. I also an o hank all employees of Isala Klinieken ha ere involved during my research. Everyone as alays prepared o help me and o anser my quesions, his made orking on my hesis even more fun. Of course I also an o hank my supervisors from he universiy, Ingrid and Erin. You gave me he freedom o make i my on research. Alhough e only had a fe appoinmens, our conversaions alays helped me o make a large sep forard a he righ momen. I do no only an o hank you for your help during my Maser hesis, bu also for our collaboraion before ha. I learned a lo from you during my courses and as a suden assisan. Thank you boh for hose opporuniies. Las bu no leas, I an o hank my family and my friends for he suppor hey gave me during my sudy. I as a rollercoaser of evens and emoions and you ere alays here o fall back on. I hope you ill enjoy reading my repor jus as much as I did riing i. Annemaaike Hooijsma Zolle, Sepember

8 7

9 Conens Chaper 1 - Inroducion Isala Klinieken Zolle Problem Research objecive and research quesions Chaper 2 - Conex analysis Process Planning and conrol Chaper 3 - Performance measures Measures from lieraure Measures in pracice Overvie performance measures Evaluaing performance measures Measuring approach Curren performance Chaper 4 Tool expecaions Deparmen Paien Logisics Admissions planners Concepual model Chaper 5 Exising models from lieraure Surveys Aricles Chaper 6 - Tool for insighs abou he ards Forecass Inegraing performance measures Tool for insighs abou he ards Chaper 7 - Implemenaion Implemenaion seps Sakeholder analysis Chaper 8 - Conclusion and recommendaions Conclusions Discussion Recommendaions Bibliography Appendices Appendix A OR block schedule Appendix B Inervie design acical level Appendix C Inervie design operaional level Appendix D Survey quesions Appendix E Inervie design admissions planners Appendix F Concepual model Appendix G Overvie variables forecass Appendix I Explanaion disribuion of he number of used beds per hour Appendix J Explanaion expeced number of discharges

10 9

11 Chaper 1 - Inroducion Usually a paien has o go hrough several seps for his/her reamen in a hospial. This process is called he clinical pahay. For some paiens saying a he ard is par of his pahay. Wards inerac a lo ih oher deparmens, for example he operaing room (OR) deparmen. These oher deparmens have a large influence on he ards. Despie of his clear ineracion beeen deparmens, mos research in healhcare logisics focuses on a single deparmen (Vanberkel, Boucherie, Hans, Hurink, & Livak, 2009) and ofen his single deparmen is no he ard (Cardoen, Demeulemeeser, & Beliën, 2010). This means ha improvemens are made a oher deparmens and ha ards jus have o cope ih hese changes. This conradics ih he fac ha saffing a he ards is an expensive resource for a hospial (Hurs, 2010) and sufficien saffing is imporan because i has a large influence on he moraliy rae (Aiken, Clarke, Sloane, Sochalski, & Silber, 2002) and on paien and employee saisfacion. For hese reasons Isala Klinieken in Zolle iniiaed research on he ineracion beeen ards and, among ohers, he OR deparmen. Wih his maser hesis e hope o conribue o his research line. Secion 1.1 describes he conex of he research, Isala Klinieken in Zolle. Secion 1.2, gives a descripion of he problem invesigaed in his research. From his descripion e deermine he objecive of he research in Secion 1.3. This chaper ends ih he research quesions ha ill be ansered during he res of he repor. 1.1 Isala Klinieken Zolle The conex in hich his research akes place is Isala Klinieken in Zolle. This is he larges op clinical hospial in he Neherlands. I has approximaely employees o suppor 260 medical specialiss. In he Figure Logo Isala Klinieken hospial here are approximaely 1000 beds available. Wih hese resources Isala Klinieken helps more han oupaiens and inpaiens per year. The hospial has yearly revenues of 442 million euro. Nex o reaing paiens Isala Klinieken also focuses on innovaion, raining and research. Isala Klinieken has o locaions; hese are Sophia and Weezenlanden. Nex o ha here are o oupaien clinics in Heerde and Kampen, o laboraories and a Diaconessenhuis Meppel (.isala.nl/overisala, 2012). A he momen Isala Klinieken is building a ne hospial in Zolle a he locaion Sophia. In Augus 2013 he ne hospial ill be ready for use. Figure 1.2 shos an aris impression of he ne building (.isalabou.nl, 2012). This research projec is commissioned by he deparmen Paien Logisics of Isala Klinieken. This deparmen is concerned ih he paien flo and capaciy managemen in he hospial. Figure Aris impression ne hospial (source: hp://.isalabou.nl/isalabou/srucuur/pages/defaul.aspx#x1) 10

12 1.2 Problem This secion describes he problem on hich his research focuses. Subsecion gives he moivaion for he research and Secion describes he problem in more deail Moivaion In Isala Klinieken he processes a he ards are negaively affeced by he admissions schedule. One of he problems experienced is a high variabiliy in he orkload for nurses a he ards. According o Livak e al. (2005) reducing unnecessary variabiliy a ards ill reduce he sress for paiens and nurses and i ill improve he safey and qualiy of care for paiens. Large variabiliy in he amoun of ork makes ha he number of nurses ha is acually needed, differs from he number ha is scheduled. During some shifs here is no enough saff and during ohers here is oo many. This ill no only resul in loer safey and qualiy a he ards bu i also leads o frusraions, hich in urn has is effecs on he processes a he ards. The amoun of ork a he ard originaes from, among oher hings, he number of paiens a he ards, he inensiy of care of hese paiens and he number of admissions and discharges in a cerain period of ime. The number of paiens, admissions and discharges are also used as measures in earlier research projecs abou his opic (Vanberkel, Boucherie, Hans, Hurink, Len, & Haren, 2011a). The schedules for he ORs and for non-surgical reamens have a large influence on hese variables and herefore also on he amoun of ork a he ards. A beer ineracion beeen he deparmens could reduce he variabiliy in his. A Isala Klinieken ools already exis for forecasing he amoun of ork based on an exising schedule. These forecass are no used o improve he schedule. Therefore, i does no reduce variabiliy and he ards sill have o adap o he schedules for he ORs and for non-surgical reamens. For he ards i is no alays possible o adap because of capaciy (nurses and beds) and ime resricions. I ould be beer o also look a reducing he variabiliy of he amoun of ork. Previous research a Isala Klinieken (Vlijm, 2011) focused on reducing he variabiliy by making adjusmens o he OR schedule on a acical level. This research akes he nex sep by broadening he scope o oher hierarchical levels and o oher aspecs of he processes a he ards Problem descripion To give a more deailed descripion of he problem e use he frameork for hospial planning and conrol from Hans e al. (2011). This frameork is displayed in Table 1.1. The frameork consiss of some managerial areas and hierarchical levels. In he able an example is given for each combinaion. The area and level ha our research focuses on are marked green in he able. Laer in his secion e explain hese choices. Hans e al. (2011) give four managerial areas, hese are: Medical planning This area is abou medical decisions made in a hospial. I especially involves he decisions made by specialiss and nurses. Resource capaciy planning Resource capaciy planning involves he planning of non-reneable resources, like saff and rooms. I also includes scheduling paiens. 11

13 Maerials planning Nex o non-reneable resources hospials also use reneable resources, like blood, medicines and equipmen ha are only used once. Planning of hese resources belongs o he maerials planning area. Financial planning The final managerial area is financial planning. This includes decisions abou coss and revenues of he hospial. From he moivaion described above e can conclude ha his research focuses on resource capaciy planning. This area includes scheduling for he ORs and non-surgical reamens and saffing a he ards. Hans e al. (2011) also give four hierarchical levels. The levels are: Sraegic level The sraegic level includes decisions made for he long erm. Tacical level Medium erm decisions are made a he acical level. These decisions do no ye concern acual paiens. Operaional offline level A he operaional offline level decisions are made for he shor erm, bu he decisions are sill made in advance. In conras o he acical level, decisions a his level do include acual paiens. Operaional online level The decisions made a he operaional online level are also for he shor erm. Hoever, hese are no made in advance, bu during he process. Table Frameork for hospial planning and conrol Medical planning Resource capaciy planning Maerials planning Financial planning Sraegic Research Case mix planning Warehouse design Invesmen plans Tacical Treamen selecion Block planning, saffing Supplier selecion Budge allocaion Operaional offline Diagnosis Appoinmen scheduling Maerials purchasing Cash flo analysis Operaional online Triage Emergency coordinaion Rush ordering Billing complicaions To decide hich hierarchical level o focus on, e compare he curren siuaion of he resource capaciy planning area ih he desired siuaion. Figure 1.3 shos he curren siuaion. We divide he resource capaciy planning in o pars, he ORs and non-surgical reamens and he ards. For he ORs and non-surgical reamens he focus of he research is on scheduling paiens and no on oher resources needed for he reamens. The capaciy planning processes a he ards can be divided in saffing and bed planning. 12

14 Figure Curren siuaion A a sraegic level producion agreemens are made for he reamens. A he ards he bed capaciy is deermined and sraegic decisions for he orkforce are made, like for example he number of nurses o hire. In he curren siuaion no many changes are made on a acical level. Block schedules and a basic orkforce schedule are used, bu hese are ofen adoped from he previous period. Also he allocaion of beds o he ards and specialisms is almos never changed. Isala Klinieken has a acical planning ool (TPT), hich as consruced by Ronald Vlijm (2011). This ool is based on he model of Vanberkel e al. (2011a). The ool can predic he amoun of ork a he ards based on he OR block schedule and i can make improvemens o he block schedule based on his. This ool is no ye used in he organizaion. 13

15 A he operaional offline level he planners for he differen specialies make he admissions schedules, hich means filling he block schedules ih paiens and reservaions for emergencies. Based on his schedule he planner assigns he paiens o he ards. A his level he basic orkforce schedule of he ards is adaped. Isala Klinieken has buil an inensiy of care ool ha can predic he amoun of ork a he ards based on an exising admissions schedule and on he curren ard occupancy. In his i akes ino accoun he expeced inensiy of care of he paiens. A some ards his ool is used o deermine he orkforce schedule on he offline and online level. A he operaional online level changes are made in he schedule for he reamens and based on hese changes he orkforce schedule and he bed planning of he ards is adaped. During execuion of he schedules daa is sored in he daa arehouse. This daa is ofen only used for decisions a a sraegic and acical level. A lo of improvemens are possible for hese processes. The desired siuaion according o Isala Klinieken is displayed in Figure 1.4. In his figure improvemens are marked green. We describe he improvemens per hierarchical level: Sraegic level An improvemen a he sraegic level is ha he producion agreemens are considered more hen performing he sraegic orkforce planning and hen deermining he bed capaciy. Bu also he oher ay around is imporan, hen producion agreemens are made consrains concerning he orkforce and bed capaciy should be considered. Tacical level Previous projecs of sudens a Isala Klinieken ofen focused on his level. Therefore here are already some ools available for improvemens here. The main aspec is he acical planning ool ha can predic he amoun of ork a he ards based on he OR block schedule and can improve he block schedule based on his. A large improvemen a he acical level is using his ool every fe monhs o evaluae he block schedule, he basic orkforce schedule and he bed allocaion. Furher improvemens a his level ould be improving his acical planning ool. I is, for example no clear heher he objecive ha his ool uses, is he righ one. Operaional offline level A his level here already exiss an experimenal inensiy of care ool. Hoever, his ool is currenly only used by a fe ards, in he desired siuaion he ool is used for deermining he orkforce schedules of all ards. The admissions schedules are no deermined based on he block schedule and he aiing liss. Effecs of he schedule for he ORs and nonsurgical reamens are parly clear for he planner, because hese are he main resources hey use. Hoever, he effecs for he ards are no considered a all hen making he admissions schedule. Therefore, an improvemen a he operaional offline level ould be o consruc a ool ha gives he planner insighs in he effecs of he consruced schedule. Alhough effecs for he ORs and non-surgical reamens are already more clear han hose for he ards, he ool could also give some more insighs in hese effecs. I could for example calculae a probabiliy of overime for a given schedule. 14

16 Figure Desired siuaion Operaional online level Daa from he daa arehouse is no mainly used for decisions a he sraegic and acical level. I ould be beer o learn from resuls of previous periods also on oher levels. Feedback should be given o he planners and o he people ha are responsible for he block schedules. Besides, he daa should be used for improving he differen ools. The figure for he desired siuaion does no sho improvemens in he processes a he operaional online level, i only shos improvemens for feedback on his level. A his level he ineracion beeen he processes could be invesigaed. Hoever, i is no ye clear heher improvemens are possible here, because you canno influence emergencies much. I is also expeced ha 15

17 improvemens a oher levels are more urgen. Therefore, in his research e do no furher look ino possible improvemens for he processes a he operaional online level. We ill focus on he operaional offline hierarchical level. Because of he ne building for he hospial, curren projecs especially focus on decisions on he sraegic and acical level. For example, he bed capaciy and he allocaion of beds o he ards are deermined no. These decisions ill no have he expeced opimal resuls hen he capaciy planning processes a he loer levels and in oher areas do no perform sufficienly. A he acical level Isala Klinieken already did a lo of research. Alhough some adjusmens migh be necessary, he exising acical planning ool can already lead o large improvemens a his level. Therefore, a focus on he operaional offline planning is more urgen no. I is expeced ha focusing on he ool on he operaional offline level leads o he bes resuls. The feedback from he daa arehouse is also imporan, bu his is only possible hen processes a higher levels are performing adequaely. The ool on he operaional offline level should give he planner insighs in he effecs of an admissions schedule on he ards, he ORs and non-surgical reamens. Hoever, research necessary for hese hree aspecs is very differen, because differen performance measures should be used and oher deparmens of he hospial are involved. To make sure he research does no become oo broad, e only focus on he ards. The planner already sees some effecs for he ORs and non-surgical reamens, because ha is he main resource hey schedule. This is no he case for ards. Therefore i is no more imporan o focus on hese insighs. To ge from a schedule o he insighs differen seps are necessary, hese seps are shon in Figure 1.5. The firs sep is o forecas ha ill happen a a ard hen applying a cerain schedule. The nex sep is calculaing performance measures from hese forecass and he final sep is displaying i o he admissions planners. Schedule 1. Forecas 2. Performance measures 3. Display Insighs abou he ards Figure Seps in he process of generaing insighs 1.3 Research objecive and research quesions Based on he problem descripion e formulae he folloing research objecive: Generaing insighs in he effecs of operaional offline scheduling decisions made for hospial admissions on he capaciy planning processes a he ards. In order o reach his objecive e anser he folloing research quesions hroughou his hesis: 1. Ho are he curren capaciy planning processes for he ORs, he non-surgical reamens and he ards organized and ho are hey planned and conrolled? The firs sep in he research is finding ou ho he curren processes are organized. A shor descripion is already given in Secion Chaper 2 expands his descripion. 16

18 2. Wha are performance measures for an admissions schedule for he hospial ha need o be considered hen looking a he capaciy planning processes a he ards? The research objecive is o generae insighs, bu i is no ye clear hich insighs are imporan. These insighs should be based on performance measures ha are imporan for he ards, on an organizaional level, and can be influenced by he admissions schedule on an operaional offline level. Finding ou ha hese performance measures are is needed for he second sep in Figure 1.5. Chaper 3 describes hese performance measures and also gives he curren performance of he ards based on hese measures. 3. Wha are he expecaions of he organizaion for he proposed ool? The research should resul in insighs for he planners. Ho should hese insighs be presened according o he deparmen Paien Logisics and he planners hemselves? Ansering his quesion is imporan for he hird sep in Figure 1.5. Chaper 4 gives a concepual model of he ool. 4. Which models have been proposed in lieraure for generaing insighs in he effecs of an admissions schedule? In Chaper 5 e look for exising lieraure abou models ha combine admissions schedules ih he processes a he ards. We could use hese models as a saring poin for he forecasing sep of our ool. 5. Which models are suiable o be used as he basis for he proposed ool and ho can hese models be adaped o be applicable in Isala Klinieken? Previous research quesions give models from lieraure and resricions from pracice for he ool. Chaper 6 combines his and gives a descripion of he proposed ool. Because of ime consrains e develop only par of he proposed ool during his research. 6. Wha are he issues hen implemening he proposed ool? The final sep in he process is implemening he proposed ool. We ill no be involved much in he implemenaion. Hoever, in Chaper 7 e do give a proposal for he implemenaion phase. 7. Wha conclusions and recommendaions follo from he research? Finally, Chaper 8 gives he conclusions and recommendaions ha e can address based on his research. 17

19 Chaper 2 - Conex analysis A problem canno be solved ihou knoing is conex. Tha is hy his chaper focuses on analyzing he conex of he problem ha is described in he previous chaper. In Subsecion 2.1 he process flo of he paien hrough he hospial is addressed. This process flo needs o be planned and conrolled. Subsecion 2.2 discusses hese aspecs. 2.1 Process This secion describes he process flo of a paien hrough he hospial. There are a lo of possibiliies for his, bu he focus in his research is on paiens ha ill go o he ard in some poin of he process. Firs e describe he differen seps of he process flo and hen e give some more informaion abou hese seps. A he momen Isala Klinieken is building a ne hospial, in his hospial some of he characerisics of he process ill change, e describe hese changes a he end of his secion. When a paien arrives a he hospial he/she firs goes o he oupaien deparmen (possibly afer a reference from he general praciioner) or he emergency deparmen. When he paien arrives a he oupaien deparmen, he paien has an appoinmen ih a specialis. The specialis deermines heher he paien needs reamen and heher his is urgen. If i is urgen, he paien goes o he ard direcly, oherise he paien is pu on he aiing lis and an admission is scheduled hen capaciy becomes available. A he day of he surgery or reamen, he paien firs goes o he ard. From he ard, he/she goes o he OR deparmen or o a non-surgical deparmen. A he OR deparmen he paien is firs kep in holding and hen he/she is prepared for surgery. Afer he reamen he/she has o go ino recovery. Finally he paien goes back o he ard, or firs o he inensive care and hen o he ard, unil he/she is ready o leave he hospial. This oal process flo is shon in Figure 2.1. Oupaien deparmen Urgen? no Waiing lis Admission scheduled Emergency deparmen yes Oher reamen Ward OR deparmen Holding Preparing for surgery Surgery Recovery IC needed? yes no Leaving he hospial IC Figure Process flo paien The hospial has a lo of differen oupaien deparmens. Every oupaien deparmen belongs o a specialism and each specialism has cerain ards ha hey preferably use. The reamens ha are performed by a specialism can be surgical or non-surgical. Table 2.1 gives an overvie of he specialisms of Isala Klinieken. I also gives heir main locaion, hich ards hey preferably use, 18

20 hich percenage of all admissions hey cause, hich percenage of heir admied paiens needs a surgical reamen and hich percenage is emergencies. Boh day care paiens and clinical paiens are included. Paiens ha only go o he emergency deparmen or acue admissions ard are no considered as admissions here. Table Overvie specialisms (source: izis, 2011, n = paiens) Specialism Locaion Preferred ards % of all admissions % of admissions ha needs OR Emergency % of admissions Aneshesiology Weezenlanden A2, A5, A6 1,5% 98% 3% Cardiology Weezenlanden B1, B1IC, 10,1% 12% 48% B3, B5, B6 Ear, nose and hroa surgery Sophia and SZ A1, 4,4% 97% 3% (ENT) Weezenlanden WL A2, WL A2K Ophhalmology Weezenlanden A2 5,1% 94% 4% Gasroenerology Sophia B1, M3 11,5% 2% 51% General surgery Sophia A1, A1P, 8,0% 77% 33% B3, B4, B5 Gynecology Sophia A1, A5, 9,2% 40% 44% A5C, A6, A6W, B5G Inernal medicine Sophia A1P, A3, 16,0% 4% 51% B2, B6, M5 Lung medicine Weezenlanden A2, B4, B6 5,1% 3% 48% Neurology Weezenlanden A2, A5, A6 5,0% 8% 49% Neurosurgery Sophia D3, IC 1,9% 91% 10% Ja surgery Weezenlanden A2, A2K 1,1% 98% 7% Orhopedics Sophia and Weezenlanden SZ B3, WL A2, WL A4 5,4% 90% 17% Pediarics Sophia A5W, A6W, 5,5% 5% 49% K1B, K2, K3, NEO Plasic surgery Sophia A1, B3, K3 3,7% 98% 12% Psychiary Sophia SZ H0 0,7% 14% 50% Radioherapy Sophia B1 0,02% 8% 77% Rheumaology Weezenlanden A2, A5 0,7% 11% 38% Special denal surgery Weezenlanden A2, A2K 0,4% 99% 0,3% Thoracic surgery Weezenlanden A7, IC 1,5% 95% 7% Urology Weezenlanden A2, A2K, A5 2,8% 65% 21% As already can be seen in Table 2.1, he hospial has a lo of differen ards. Tables 2.2 and 2.3 give an overvie of hese ards per locaion. The ards are differen because of he specialisms ha use hem, bu also he funcion of he ards differs. The ards can be divided ino day care ards and 19

21 clinical ards. Paiens ha only need o say in he hospial on he day of heir reamen ofen say a he day care ards. These ards are closed during he nigh and he eekend. Paiens ha need o say in he hospial longer go o he clinical ards. A Isala Klinieken his division is no ha sric, here are also day care paiens ha go o a clinical ard. Clinical ards can be divided ino inensive and specific ards, general nursing ards and psychiaric ards. Table Overvie ards locaion Sophia (source: izis, 2011, n = paiens) Ward Number of beds Day care / Clinical Specialisms A1 10 Day care Gynecology, Ear, nose and hroa surgery, Plasic surgery, General surgery A1P 15 Clinical (general) Inernal medicine, General surgery A3 40 Clinical (general) Inernal medicine A5 18 Clinical (inensive and specific) Gynecology A5C 9 Clinical (inensive and specific) Gynecology A5W 8 Clinical (inensive and specific) Pediarics A6 17 Clinical (inensive and specific) Gynecology A6W 17 Clinical (inensive and specific) Gynecology, Pediarics B1 41 Clinical (general) Radioherapy, Gasroenerology B2 41 Clinical (general) Inernal medicine B3 42 Clinical (general) General surgery, Orhopedics, Plasic surgery B4 42 Clinical (general) General surgery B5 11 Clinical (general) General surgery B5G 17 Clinical (general) Gynecology B6 23 Clinical (general) Inernal medicine D3 30 Clinical (general) Neurosurgery H0 27 Clinical (psychiaric) Psychiary IC 22 Clinical (inensive and specific) Neurosurgery K1B 16 Clinical (inensive and specific) Pediarics K2 8 Day care Pediarics K3 30 Clinical (inensive and specific) Pediarics, Plasic surgery M3 12 Day care Gasroenerology M5 12 Day care Inernal medicine NEO 14 Clinical (inensive and specific) Pediarics Isala Klinieken has an OR deparmen on boh locaions. A locaion Weezenlanden here are elve ORs of hich o are day care ORs. On hese day care ORs only shor surgeries are done on paiens ha can go home a he end of he day. Locaion Sophia has en ORs. From hese en, o are day care ORs and one OR is compleely reserved for emergencies. The capaciies for he non-surgical reamens are less cenralized and herefore oo complex o describe here. 20

22 Table Overvie ards locaion Weezenlanden (source: izis, 2011, n = paiens) Ward Number of beds Day care / Clinical Specialisms A2 41 Clinical (general) Aneshesiology, Special denal surgery, Ja surgery, Ear, nose and hroa surgery, Lung medicine, Neurology, Ophhalmology, Orhopedics, Rheumaology, Urology A2K 6 Clinical (inensive and specific) Special denal surgery, Ja surgery, Ear, nose and hroa surgery, Urology A4 41 Clinical (general) Orhopedics A5 44 Clinical (general) Aneshesiology, Neurology, Rheumaology, Urology A6 35 Clinical (general) Aneshesiology, Neurology A7 36 Clinical (general) Thoracic surgery B1 12 Clinical (inensive and specific) Cardiology B1IC 4 Clinical (inensive and specific) Cardiology B3 39 Clinical (general) Cardiology B4 42 Clinical (general) Lung medicine B5 39 Clinical (general) Cardiology B6 41 Clinical (general) Cardiology, Lung medicine EHH 14 Clinical (acue admissions) Cardiology, Neurology, Urology IC 19 Clinical (inensive and specific) Thoracic surgery Isala Klinieken is building a ne hospial a locaion Sophia. There ill be feer ORs in he ne building. No here are 22 ORs in oal, in he ne building here ill be 14 ORs, bu probably six ORs ill be added in a reamen cener. I is expeced ha beds ill become he main boleneck. The ne hospial ill have approximaely 780 beds insead of he 1000 beds ha are no available. Alhough hese 1000 beds are no all used no, some changes are needed o cope ih having a capaciy of only 780 beds. One of he changes is ha he ards become larger. In ha ay i is easier o capure he variabiliy in he number of beds needed (porfolio effec). Some of he curren ards are shared beeen specialisms, bu every specialism has is on beds and saff a he ards. In he ne hospial specialisms also share beds and saff a a ard, such ha a high number of beds needed for one specialism can be solved by a specialism ha has a lo requiremen of beds a ha momen. Anoher large change in he ne hospial is he acue admissions ard. A his ard emergency paiens ill be admied. These paiens say a his ard for a mos o days. During hese days a diagnosis and a reamen are performed. In his period he paien is eiher discharged or ransferred o he preferred ard. Because of his acue admission ard, here ill be no unplanned admissions a clinical ards. A hese ards here ill sill be some emergency paiens admied, bu hese admissions are knon up o o days in advance. This makes admissions a he ards more predicable. 21

23 2.2 Planning and conrol The process flo of he paien described in he previous secion needs planning and conrolling. Figure 1.3 in Chaper 1 already shos he general picure of he curren planning and conrolling of admissions, ard saffing and bed capaciy. This secion gives a more deailed descripion. This descripion is divided ino he admissions schedule (Secion 2.2.1) and scheduling of he ards (Secion 2.2.2) Admissions schedule The admissions schedule is he schedule for ORs and non-surgical reamens. A he sraegic level producion agreemens are made. These sraegic decisions fall ouside he scope of his research. The acical level conains consrucing block schedules. The block schedules are almos alays adoped from he previous period. Hoever, hen he hospial moves o he ne locaion nex year his block schedule needs o be changed. The curren block schedule for he ORs is displayed in appendix A. On he operaional offline level, he block schedules are filled ih paiens. This scheduling is decenralized, his means ha every specialism has is on planner and makes is on admissions schedule. Because of he block schedules every specialism has is on periods in hich i can schedule paiens. A planner deermines per block heher here are suiable paiens in he aiing lis. The schedules are based on, among oher hings, he aiing lis, he ishes of he paien, resricions regarding he specialis, OR ec. and someimes also on he number of admissions a he ards. For some specialisms he schedule is made a fe eeks in advance and he paiens are noified approximaely one eek in advance. For oher specialisms he admissions are scheduled immediaely afer he paien visied he specialis a he policlinic. On he operaional online level sill some changes can be made in he schedule of he admissions. This is necessary because emergency reamens need o be added o he schedule. This is done by he planner of a specialism or by he OR planner, depending on ho urgen he surgery is. In some cases surgeries need o be cancelled because of emergencies or surgeries ha ake longer han expeced Ward scheduling The ards have o main resources, beds and saff. This secion discusses he planning of boh resources. We sar ih he bed planning. A he sraegic level i is deermined ho many beds are needed in oal. This process falls ou of he scope of his research, so e ill no discuss i in more deail. A he nex level, he acical level, beds are allocaed beeen ards and specialisms. This allocaion almos never changes. Hoever, large changes ill soon be necessary for he ne hospial. The precise allocaion for he ne hospial sill has o be deermined. A he operaional offline level paiens are assigned o he ards. This is done by he planner of he specialism. The planner does his a he same ime as he/she makes he schedules for he reamens. These schedules are mainly based on he requiremens of he reamens. So he planner does schedule paiens a he ards, bu ofen he capaciy planning processes a he ards are no considered in his. A he operaional online level he schedule changes because of emergencies and variabiliy in lengh of say. Hoever, i also happens ha he proposed schedule appears no o be feasible a he ards. This also leads o las minue changes. 22

24 The oher par of planning a he ards is ard saffing. The sraegic level of ard saffing includes decisions like ho many nurses o hire. A he acical level a basic orkforce schedule is consruced. This schedule differs per ard and per specialism, because he nurses are generally no exchanged beeen specialisms. The basic orkforce schedules do almos never change. They are ofen adoped from he previous period. The basic schedule is adaped a fe days in advance; his is he operaional offline level. These changes mosly mean ha exra nurses are required. The hospial has nurses ha can be flexibly deployed. These nurses are used a he deparmens ha need exra saff. The operaional offline schedule is based on he bed schedule for he nex days. Nex o ha, a fe ards use he inensiy of care ool for deermining he schedule. The inensiy of care ool can predic he amoun of ork a he ards based on an exising admissions schedule and on he curren ard occupancy. A he operaional online level no many changes are possible. Hoever, if i is really necessary las minue requess are made for exra nurses. No e kno more abou he conex of he research e can sar ih he main research focus. The firs sep is o find ou ha defines a good admissions schedule. The nex chaper describes he performance measures for his. 23

25 Chaper 3 - Performance measures This research delivers a ool ha gives admissions planners insighs in he effecs of a schedule for he ards. The firs quesion ha arises is hich insighs should he ool give?. Since he admissions planner should be able o improve he processes a he ards, he insighs mus be based on performance measures for he ards. Secion 3.1 describes such measures from lieraure. Secion 3.2 uses hese measures from lieraure as a saring poin o analyze hich performance measures are imporan according o he hospial. These firs o secions resul in a long lis of performance measures. Secion 3.3 gives an overvie of hese measures. Presening all hese measures o he admissions planners ill sill no give much useful insighs. Secion 3.4 herefore reduces his lis by evaluaing he measures using evaluaion crieria and he imporance from he inervies and he survey. Before he performance measures can be used, you need o kno ho hey should be measured. Secion 3.5 describes his. Wihou knoing he performance of he curren siuaion (before he ool is implemened) i is no possible o kno heher he performance is improved by implemening he ool. Therefore, Secion 3.6 describes he curren performance. 3.1 Measures from lieraure Lieraure abou performance measures can be divided ino o ypes: lieraure ha gives relevan measures for his research and lieraure ha discusses evaluaion crieria for performance measures in general. This secion is also divided according o his. Secion gives performance measures for ards hile Secion discusses crieria on hich hese crieria can be evaluaed Performance measures Mos overvies of performance crieria for ards come from lieraure revies abou healh care logisics. All hese lieraure revies have a broader scope han ards alone. Hoever, e only discuss measures for he ards, and hen specifically he measures ha can be influenced by an admissions schedule. Li and Benon (1996) performed a lieraure revie abou healh care performance measures used in lieraure. They do no menion a lo of measures ha are ineresing for he ineracion beeen admissions schedules and ards. Hoever, hey do give a useful frameork for displaying performance measures. They disinguish beeen inernal and exernal measures, and cos/financial and qualiy performance. Inernal measures are imporan for people inside he hospial and exernal measures are concerned ih exernal sakeholders, for example an insurance company and paiens. We evaluae performance measures from oher lieraure revies based on his frameork. Cardoen e al. (2010) analyze aricles abou operaing room scheduling based on differen aspecs, one of hese aspecs is he performance crierion used. In he survey of Guerriero and Guido (2011) aricles on differen organizaional levels are discussed. On he acical level hey discuss objecives used in aricles and on he operaional level hey menion crieria used. Table 3.1 shos he crieria from hese surveys in he frameork of Li and Benon (1996). Table 3.1 shos ha lieraure does no menion any useful measures for exernal cos/financial performance. This is performance according o for example healh insurance companies. This is logical because performance of he ards is less imporan for hose organizaions. 24

26 Table Performance measures from lieraure Cos/financial performance Qualiy performance Inernal measures Underuilizaion 1,2 Number of nurses 2 Occupancy raio 1,3 Number of beds used 2 Leveling ards/bed occupancy 1,2 Exernal measures Paien deferrals/ refusals 1 1 Cardoen e al. (2010) 2 Guerriero and Guido (2011) 3 Li and Benon (1996) Overuilizaion 1, Crieria for measures Lieraure also gives some guidelines for evaluaing performance measures. One heory ha is ofen used is he SMART-principle (Doran, 1981). Originally his heory as mean o evaluae objecives. Hoever, in lieraure i is also used o evaluae performance indicaors (Shabin & Mahbod, 2007). SMART sands for specific, measurable, aainable, realisic and ime-bounded. Aainable means ha he goal should be reachable, bu ha i is no oo easy. This is no applicable hen choosing performance measures, because here is no a goal o reach. The oher crieria can be used for performance measures. Van Hoorn and Wend (van Hoorn & Wend, 2008) give anoher lis of condiions ha performance indicaors should mee; indicaors should be relevan, ransparen, comparable, measurable, normaive and possible o be influenced. There is some overlap ih he SMART-principle. When combining hese o principles, e ge he folloing lis of crieria hich e ill use o evaluae he performance measures: Specific/Transparen: he performance measure should be clear and easy o undersand. Relevan: shoing he performance measure should resul in enough improvemen for he ards compared o he effor needed o deermine he measure. Measurable: i should be possible o monior he performance measure. This means ha needed daa is available and ha he effor needed o measure he indicaors does no increase oo much. Comparable: he measuremen should be reliable and unambiguous such ha a repeiion of he measuremen gives he same resuls. Realisic/Can be influenced: he admissions planners should be able o influence he performance measure. Time bounded: i should be possible o monior he performance measure in a reasonable period of ime. Normaive: here should be he abiliy o compare he performance measure o a norm. Every performance measure should comply ih all hese crieria. In Secion 3.4 e evaluae he final lis of performance measures based on hese aspecs. 3.2 Measures in pracice The previous secion gives he measures menioned in lieraure, bu he measures used in Isala Klinieken migh even be more imporan in his research. The opinion of he saff a he ards is of 25

27 course very imporan in his, bu also people in he oher hierarchical levels of he organizaion should be involved. Secion describes ha he sraegic level uses as performance measures for he ards. Secion does his for he acical level and finally Secion looks a i from he operaional poin of vie Measures on a sraegic level The sraegic level of Isala Klinieken consiss of o members of he board of direcors and hree direcors of operaions. To find ou hich performance measures are imporan from a sraegic poin of vie, e inervieed one of he direcors of operaions. From his inervie i appears ha he sraegic level of Isala Klinieken is no much involved in processes and decisions a he ards. Their main responsibiliy for he specialisms is conrolling he producion according o he producion agreemens made. This could have consequences for he ards, bu ha is no heir main focus (Hings, 2012). For he sraegic level of he hospial uilizaion a he ards should be as high as possible. The admissions planner should have insighs in his uilizaion of he ards. According o Hings (2012) i is imporan ha he planner alays schedules a consan mixure of paiens on a ard. I should be a consan mixure based on inensiy of care and on lengh of say, because i is expeced ha i is difficul o realize a high uilizaion if hese mixures are no consan. For he ne hospial his consan mixure should also be aken ino accoun hen deermining hich specialisms ill use a ard ogeher (Hings, 2012) Measures on a acical level The sraegic level of he hospial consiss of only a fe people, bu he acical level is already much larger. From his level e inervieed muliple people ha are involved in measuring performance of he ards. These are he Capaciy manager, he manager of he deparmen Qualiy and Safey and some managers of specialisms. We chose hree specialisms for he inervies ha are very differen based on size and use of he OR. A ranslaion of he inervie design is shon in Appendix B. We coded he aspecs menioned during he inervie ourselves. Because e only inervieed a fe people from he acical level, he performance measures menioned are probably no complee. To kno hich par of he measures e found, e used a Poisson model ha as also used by Nielsen and Landauer (1993). They used i for usabiliy problems, bu i is also applicable for he number of ansers found in inervies. This heory indicaes ha e found approximaely 86% of he performance measures on he acical level. We assume he measures no ye menioned are he leas imporan ones and herefore hese five inervies are sufficien (Hannink, 2012; Klappe, 2012; Broers, 2012; Veraar, 2012; Elbrech & Seenbergen, 2012). Uilizaion Uilizaion can be measured on muliple levels. The highes level is he oal uilizaion for he hole hospial. In one of he inervies i as menioned ha he admissions planner should have more insighs in he uilizaion on his higher level. An overvie of he hole hospial migh be oo complex, bu hey should a leas have insighs in he ards ha are preferred firs and second by heir specialism. Nex o he uilizaion in oal, also he uilizaion per ard is imporan. This uilizaion indicaes heher a schedule is efficien. I ould be useful if he ool gives he possibiliy o disinguish beeen day care and clinical paiens for his measure. This could be necessary if a 26

28 ard has separae beds per ype of paiens. A final level menioned for uilizaion is he uilizaion for each specific bed. On all levels he uilizaion should be as large as possible. Bed leveling Nex o increasing he average uilizaion also leveling his uilizaion over he days is menioned muliple imes during he inervies. The number of uilized beds could also be leveled during he day. This indicaor as menioned once during an inervie. Overuilizaion / Bed availabiliy Alhough he uilizaion is preferably as high as possible, here should no be oo many paiens in he hospial. The admissions planners should have insighs in his on a hospial level and on a ard level. No oo many paiens also means ha here is alays a bed available for an emergency paien. To reach his, beds should be reserved for hose paiens. Saff availabiliy Previous menioned performance measures are all abou beds, bu saff is also an imporan aspec. The planner should also kno ha enough saff of he righ specialism is available. Saff produciviy An indicaor moniored for mos ards is saff produciviy. This is he number of paiens on he ard per nurse. I is relaed o saff availabiliy. The difference is ha for enough saff o be available he number of paiens per nurse should no be oo high, hile for produciviy i should no be oo lo. An admissions planner can influence his by maching he number of paiens on a ard o he number of nurses scheduled. Paien saisfacion/access ime During he inervies i is menioned ha he paien saisfacion is also imporan. If a schedule is opimal for he saff and for he hospial, his does no alays mean ha i ill resul in a high paien saisfacion. Hoever, he ishes of paiens differ a lo and i is no alays possible o give he planner a clear overvie of he forecased paien saisfacion based on a schedule. Despie of his, performance for paiens should sill be an imporan indicaor for he admissions planner. One specific performance indicaor menioned for paiens is he access ime. The access ime is he ime beeen he momen he paien knos he has o undergo a reamen and he momen he is acually admied. The ards are no much involved in his performance measure. Hoever, for he admissions planners i is imporan o have insighs ino his, because hey can influence i ih he schedule. Amoun of ork A final aspec is he amoun of ork. I as menioned ha he amoun of ork a he ards does no only consis of he paiens presen. Also he inensiy of care of hese paiens and oher asks are imporan. This aspec as only menioned once during he inervies a he acical level. Hoever, he inervieee expecs ha his ill be oo difficul o measure, so i ould be bes o focus on he number of beds firs Measures on an operaional offline/online level The operaional level of he ards consiss of he saff of he ards, so he nurses and he managers of he ards. They are involved in boh he operaional offline and he operaional online level. We 27

29 explore heir opinion abou performance measures for he ards in o ays. We sen a survey o all managers of he ards. Nex o ha e inervieed a member of he VKI (Verpleegkundigen Kenniskring Isala) (van Apeldoorn, 2012); his is a orks council for nurses in Isala Klinieken. A ranslaion of he inervie design is displayed in Appendix C and a ranslaion of he quesions from he survey is given in Appendix D. VKI The VKI is concerned ih he ineres of paiens and nurses. They monior heher agreemens made in he hospial are me. The member of he VKI e inervieed is also manager of he Cardiology ard. According o her, muliple performance measures are imporan for he ards. These are parly based on he processes a Cardiology ards. The measures are he disribuion of admissions over he day, he disribuion of admissions over he eek and he uilizaion of he ards. The admissions planners of Cardiology already consider hese aspecs by keeping rack of he scheduled admissions and he resuling used beds hemselves (Huenesein & de Graaf, 2012). An example of improving he disribuion of admissions over he eek is also alloing admissions during he eekend. The Cardiology ards are divided ino muliple pars ih differen ypes of beds. These ypes are long say paiens, elecive shor say paiens and elecive day care paiens. The ards are judged based on uilizaion per ype of bed and uilizaion of all Cardiology ards in oal (van Apeldoorn, 2012). Managers of he ards The survey displayed in Appendix D as sen o all operaional managers of 35 ards of Isala Klinieken. We received a response from managers of 32 ards. In he beginning of he survey six performance measures are menioned. The respondens are asked o indicae ho imporan hese aspecs are for a good admissions schedule on a five poin scale from no a all imporan o very imporan. The main goal of hese quesions is o make clear ha he research is abou. The subjec of he research is relaively unknon for he saff a he ards. Therefore i is useful o give hem some examples firs. Figure 3.1 shos he resuls of hese quesions. I appears ha all he performance measures are indicaed by mos people as imporan or very imporan. Therefore, e include all hese aspecs in he lis of performance measures. In he survey e also asked he respondens o menion more performance measures in muliple ays. We coded hese performance measures ourselves, herefore i is subjecive. This resuled in he caegories given in Table 3.4. In he survey e paid more aenion o leveling he amoun of ork. This seemed an imporan aspec based on earlier conversaions ih saff a some ards. Also measuring his aspec migh be complex. A quesion asked abou his in he survey is heher he ork should be leveled, heher a consan paern in he ork is enough or ha boh opions are no necessary. This is asked for during he eek and for during he day. The resuls are given in Table 3.2 and 3.3. From his e can see ha a small majoriy prefers a consan paern in he amoun of ork, as ell for during he eek as for during he day. Hoever, also a large par of he respondens prefer he ork o be leveled during he eek and during he day. Therefore, e consider boh leveled ork and a consan paern in he amoun of ork as performance measures for he ard. 28

30 Sufficien beds for he scheduled paiens Sufficien beds reserved for emergencies Very imporan Imporan Neural No imporan No a all imporan 0 Very imporan Imporan Neural No imporan No a all imporan The paiens on righ he ard A consan amoun of ork during he day Very imporan Imporan Neural No imporan No a all imporan 0 Very imporan Imporan Neural No imporan No a all imporan The same amoun of ork on differen days Predicable amoun of ork Very imporan Imporan Neural No imporan No a all imporan 0 Very imporan Imporan Neural No imporan No a all imporan Figure 3.1 Resuls for imporance of aspecs Table 3.2 Resuls of survey quesion abou leveling he amoun of ork during he eek Anser I ould be ideal if he amoun of ork ould be equal for all (eek)days. 32% Amoun of ork does no have o be equal for all days, as long as here is consan paern in he amoun of ork during he eek (so on Monday alays he same ork as on oher Mondays ec.) Amoun of ork doesn have o be equal for all days or according o consan paern. 24% Table Resuls of survey quesion abou leveling he amoun of ork during he day Anser Amoun of ork should also be consan during he day. 34% Amoun of ork does no have o be consan during he day, as long as i follos a consan paern (for example alays a high amoun of ork beeen 10 and 11 am) I is no necessary ha he admissions planner ries o ge he amoun of ork consan or according o a consan paern during he day, his ill go auomaically. Percenage 44% Percenage 39% 27% 29

31 Table Performance measures from survey Performance measure Inernal financial performance Number of imes menioned Variabiliy in number of admissions during he day 14 Overuilizaion per ard 13 Variabiliy in number of admissions over he eek 9 Admiing paiens a he righ ime 8 Variabiliy in amoun of ork during he day hile considering oher asks a he 6 ards Overuilizaion on cerain momens of he day 6 Saff availabiliy 6 Paiens on he righ ard 4 Variabiliy in number of used beds over he eek 4 Uilizaion 4 Cancellaions 2 Availabiliy beds for emergency paiens 2 Variabiliy in amoun of ork over he eek 2 Admissions in he evening 1 Predicable amoun of ork 1 Variabiliy in amoun of ork during he year 1 Predicable number of admissions Overvie performance measures The previous o secions resul in a long lis of performance measures. To display his in a clear ay e use he frameork of Li and Benon (Li & Benon, 1996). For each quadran of he frameork e pu he indicaors in a KPI ree. In a KPI ree high level performance indicaors (like for example coss) are divided ino sub KPIs. In Isala Klinieken every deparmen deermined a number of key performance indicaors (KPIs) on hich hey monior he performance of heir deparmen. These KPIs are deermined using a KPI ree. The goal of displaying i in his ay is no o give an objecive overvie, jus o give a clear overvie of he indicaors. Figures 3.2, 3.3 and 3.4 sho hese KPI rees. There ere no exernal financial performance measures menioned, so for ha quadran here is no KPI ree. The ends of he rees are performance measures e ill furher consider. In he nex secion e ill reduce his long lis of measures hen evaluaing hem. Average uilizaion Variabiliy in number of used beds Saff produciviy Consan mixure of paiens For hole hospial Per ard Per ype of bed During he day Over he eek Based on inensiy of care Based on lengh of say Per bed Figure KPI ree inernal financial performance 30

32 Inernal qualiy performance Amoun of ork Admissions Variabiliy Consan paern Predicable Variabiliy In he evening During he day (hile considering oher asks a he ards) During he day Over he eek During he day Over he eek Over he eek Over he year Figure KPI ree inernal qualiy performance Exernal qualiy performance Availabiliy Access ime Admiing paiens a he righ ime Oher performance for paiens Saff availabiliy Overuilizaion Per ard Availabiliy beds for emergency paiens Cancellaions Paiens on he righ ard For hole hospial On cerain momens of he day Figure KPI ree exernal qualiy performance 3.4 Evaluaing performance measures In his secion e reduce he lis of performance measures. The firs sep in his is o evaluae he measures based on he crieria menioned in Secion If afer his sill oo many measures remain, e deermine hich measures are mos imporan according o he ranking from he inervies and he survey. So e firs look a feasibiliy and hen a imporance. The evaluaion of he performance measures is subjecive. To increase he objeciviy of he evaluaion e also le someone else evaluae he performance measures (van den Akker, 2012a). 31

33 Folloing, he level of agreemen beeen he o evaluaions can be calculaed per evaluaion crierion. The indicaor used for he level of agreemen is Cohen s kappa (Cohen, 1960). This indicaor excludes he level of agreemen ha is expeced by chance. Landis and Koch (1977) label he srengh of agreemen for differen ranges of kappa. We use hese labels o deermine ho srong he agreemens for he differen crieria are. If he level of agreemen is subsanial, almos perfec or perfec, e assume ha our evaluaion on ha crierion is correc. If he srengh of agreemen is smaller, he indicaors on hich e do no agree are discussed. If even hen he srengh of agreemen is no large enough, e ask a hird person o evaluae he indicaors on hich here is no agreemen. An indicaor should be included in he final lis if i complies ih all he crieria. Table 3.5 shos he values of Cohen s kappa before and afer he discussion beeen he o evaluaors. Before he discussion he values of Cohen s kappa are relaively small. The main reasons for his are he large number of performance measures and evaluaion crieria and he relaionships beeen he crieria, hich make i difficul o evaluae in a consan manner. Afer he discussion here appears o be sufficien agreemen for all indicaors. Table 3.6 shos he final evaluaion of he indicaors. The indicaors ha are seleced in he oal column comply ih all he crieria. The oher performance measures ill no be considered furher. Afer his evaluaion sill seveneen performance measures remain. Shoing all hese measures o an admissions planner ill be oo complex o use. The ool mus be simple o be effecive. To do his, he lis of performance measures needs o be furher reduced. We rank he indicaors per hierarchical level based on he ansers on he inervies and he survey from Secion 3.2. The sraegic and acical levels only consis of inervies. For hese levels e look a ho ofen he indicaors are menioned during hese inervies. A he operaional level e held one inervie and e sen a survey o he ards. In he survey specific quesions ere asked abou he amoun of ork. A majoriy preferred a consan paern in he ork during he eek. This is logical, because ih a consan paern you can base your basic saff schedule on his paern. Therefore e consider his indicaor as ranked high enough o be included. For he ork during he day also he majoriy prefers a consan paern, bu he difference ih leveled amoun of ork during he day is very small. Afer discussing his ih some employees of he ards e can conclude ha leveling he ork migh be more useful, because i is difficul o change your saff schedule every momen of he day. For ranking he oher aspecs on he operaional level e look a ho ofen he aspecs are menioned in he survey and during he inervie. All rankings are displayed in Table 3.7. To reduce he lis of indicaors e only use he indicaors ha are ranked highes per hierarchical level. For he sraegic level he ool ill only use he indicaor ranked firs, because only one aspec as menioned. For he acical level he ool ill sho he firs hree indicaors. The indicaors ranked fourh ere only menioned once. Finally for he operaional level e use he indicaors ranked firs o fifh. Figure 3.5 shos he final lis of performance measures in a KPI ree. 32

34 Specific/ Transparen Relevan Measurable Comparable Realisic/ Can be influenced Time-bounded Normaive Toal Table Cohen's kappas per crieria Crieria Cohen s kappa before discussion Cohen s kappa afer discussion Specific/Transparen 0,65 (subsanial) 1 (perfec) Relevan 0,10 (sligh) 0,91 (almos perfec) Measurable 0,37 (fair) 1 (perfec) Comparable 0,02 (sligh) 1 (perfec) Realisic/Can be influenced 0,14 (fair) 1 (perfec) Time bounded 0,28 (fair) 1 (perfec) Normaive -0,06 (poor) 0,78 (subsanial) Toal 0,11 (sligh) 1 (perfec) Table Evaluaion of indicaors Performance measure Uilizaion for hole hospial x x x x x x Uilizaion per ard x x x x x x x x Uilizaion per ype of bed (for example emergencies, elecive, x x x x x x x x day care ec.) Uilizaion per bed x x x Consan mixure of paiens based on inensiy of care x x x x x x Consan mixure of paiens based on lengh of say x x x x x x Variabiliy in number of used beds over he eek x x x x x x x x Variabiliy in number of used beds during he day x x x x x x x x Overuilizaion per ard x x x x x x x x Overuilizaion for hole hospial x x x x x x Overuilizaion on cerain momens of he day x x x x x x x x Availabiliy beds for emergency paiens x x x x x x Saff availabiliy x x x x x x x x Saff produciviy x x x x x x x x Paiens on he righ ard x x x x x x x x Cancellaions x x x x Access ime x x x x x x Performance for paiens Variabiliy in number of admissions over he eek x x x x x x x x Variabiliy in number of admissions during he day x x x x x x x x Admiing paiens a he righ ime x x x Admissions in he evening x x x x x x x x Variabiliy in amoun of ork over he eek x x x x x x x x Variabiliy in amoun of ork during he year x x x x x x x x Variabiliy in amoun of ork during he day hile considering x x x x x x x x oher asks a he ards A consan paern in he amoun of ork during he day x x x x x x x x A consan paern in he amoun of ork over he eek x x x x x x x x Predicable amoun of ork x 33 x

35 Table Ranking performance measures Performance measure Ranking sraegic level Ranking acical level Ranking operaional level Uilizaion per ard Uilizaion per ype of bed (for example emergencies, elecive, day care ec.) Variabiliy in number of used beds over he eek Variabiliy in number of used beds during he day Overuilizaion per ard Overuilizaion on cerain momens of he day Saff availabiliy Saff produciviy Paiens on he righ ard Variabiliy in number of admissions over he eek Variabiliy in number of admissions during he day Admissions in he evening Variabiliy in amoun of ork over he eek Variabiliy in amoun of ork during he year Variabiliy in amoun of ork during he day hile considering oher asks a he ards Performance of he ard Inernal financial performance Inernal qualiy performance Exernal qualiy performance Average uilizaion per ard Variabiliy in number of used beds over he eek Saff produciviy Variabiliy in amoun of ork during he day (hile considering oher asks a he ards) Consan paern in amoun of ork over he eek Variabiliy in admissions during he day Saff availabiliy Overuilizaion per ard Overuilizaion on cerain momens of he day Variabiliy in admission over he eek Figure KPI ree final lis of performance measures 34

36 3.5 Measuring approach Afer he evaluaion in he previous secion en performance measures remain. The ool ill sho hese performance measures. In he previous secion i is already said ha hese indicaors can be measured, bu an exac ay o calculae hem should sill be deermined. This is necessary o calculae hem in he ool, bu also o deermine he curren performance in Secion 3.6. Secion explains he ay of measuremen per indicaor. For hese calculaions a lo of informaion is needed. Secion gives a lis of his informaion Calculaion of performance measures Belo e describe per performance measure he ay in hich i ill be calculaed. We base his on mehods from lieraure and on he resuls of he inervies and he survey. Uilizaion per ard In Isala Klinieken muliple ays are used o measure he uilizaion (Hannink, 2012). The firs ay is he number of occupied beds a 10 am. 10 am is chosen because hen mos paiens are no discharged ye, bu here are already some ne paiens admied. A second ay of measuring uilizaion is he ime he beds are occupied compared o he ime he beds are available. Finally, uilizaion is also measured as he number of occupied beds a midnigh compared o he oal number of beds. This measure is especially used o say somehing abou he scheduling on compleely clinical ards. These ards only have paiens ha also say during he nigh. If a such a ard a bed is used during o subsequen nighs, i does no maer ha happens in beeen, because you canno schedule anoher clinical paien in i. Therefore, if a hese ards all beds are used a midnigh, he schedule as good regardless of ha happened he res of he day. For he ool i is imporan o sho he uilizaion in such a ay ha i gives a correc represenaion of he performance. Like jus explained, for purely clinical ards i is sufficien o sho he uilizaion a midnigh. For purely day care ards he uilizaion during opening hours is imporan. The easies ay o calculae his is o compare he ime he beds are occupied ih he ime hey are available. Isala Klinieken has o ards ha have boh clinical and day care paiens. For hese ards boh ypes of uilizaion are imporan. The exac formulas for hese ypes of uilizaion are: Uilizaion 0:00 = Uilizaion opening hours = The ool should sho hese numbers for all days in he scheduling horizon. The scheduling horizon is he period for hich he schedule is complee and for hich he admissions planner ans o kno he qualiy for he ards. For he curren performance in he nex secion e use he acual number of used beds and calculae he average uilizaion for he hole year. Overuilizaion per ard/ Overuilizaion on cerain momens of he day Overuilizaion indicaes ho much he number of expeced paiens exceeds he number of beds available. In he lis of performance measures o ypes of overuilizaion are menioned. If a ard is compleely occupied and on one day he same number of paiens is discharged as admied, his ould no resul in overuilizaion on he ard. Hoever, because some paiens migh be admied 35

37 before ohers are discharged, i could give problems on cerain momens of he day. Shoing boh measures in he ool migh be confusing. Therefore e only sho he overuilizaion on cerain momens of he day. The ool could sho he expeced number of paiens ha ill no fi. Hoever, here is much uncerainy abou his number. A beer ay o represen he overuilizaion is giving he chance ha here is overuilizaion. To calculae his, he available number of beds and he disribuion for he number of used beds are needed. The probabiliy of overuilizaion can hen be calculaed for every hour of he day like shon in he formula belo. The ool can sho his in a graph. Overuilizaion = For he curren performance and he realizaion e kno ih cerainy heher here as overuilizaion. The probabiliy of overuilizaion is eiher 0% or 100%. In Secion 3.6, e herefore calculae he percenage of hours during a year on hich here as overuilizaion compared o he opening hours (hours during hich he uilizaion is larger han 0%). Some ards have exra beds hich can be used in case of overuilizaion, for hese ards e can recognize overuilizaion from he realizaion daa. This is no he case for all ards, some ards ransfer paiens o oher ards in case of overuilizaion, his influences he daa abou he curren performance. Variabiliy in he number of used beds over he eek The variabiliy in he number of used beds over he eek should indicae ho much he number per day differs from he average. There are differen opions o calculae variabiliy. Table 3.8 liss he differen opions ih some characerisics. Table Opions o measure variabiliy Opion Formula Characerisics Mean absolue deviaion (MAD) Mean absolue percenage deviaion (MAPD) Easy o undersand (Gorard, 2005) Inuiive Efficien in he realisic siuaion (Gorard, 2005) Easy o compare See characerisics MAD Variance Quadraic uni No inuiive Sandard deviaion No inuiive Coefficien of variaion Easy o compare Tradiion (Gorard, 2005) Easy o manipulae algebraically (Gorard, 2005) Efficien under ideal circumsances (Gorard, 2005) See characerisics sandard deviaion 36

38 For he ool i is imporan ha he planner undersands ha he indicaors mean. From he opions above he mean absolue deviaion (MAD) is he mos easy o undersand. Hoever, i is also imporan ha he numbers are comparable. Weeks ih a high average number of paiens ill ofen also have a larger MAD han eeks ih feer paiens. To make sure i is comparable e use he mean absolue percenage deviaion (MAPD) insead. This is also useful hen you an o compare he resuls beeen deparmens. To calculae he variabiliy in he number of used beds e use he uilizaion like calculaed before, because hese numbers are already knon. Using hese numbers also makes i consisen ih he oher performance measures. The formula for his performance measure is shon belo. Variabiliy used beds = I is imporan o kno for hich days you deermine he variabiliy. If a ard is closed during he eekend i is clear ha you do no ake hose days ino accoun, bu for oher ards his is no clear. Wha is preferred differs beeen he hierarchical levels. Alhough his indicaor as menioned a fe imes by he operaional level, i is included in he final lis because of is imporance for he acical level. For he acical level i is imporan because ih a leveled number of used beds a higher uilizaion is possible. For his reason i is bes o level he used beds over he hole eek. Variabiliy in number of admissions over he eek This variabiliy can be calculaed in he same ay as he variabiliy in number of beds used, so displaying he MAPD. Only he expeced number of admissions during he eekdays should be used for his, because no admissions are scheduled during he eekend. Variabiliy number of admissions over he eek = Variabiliy in number of admissions during he day Because on many ards here are also hours ihou an admission, he variabiliy in he number of admissions per hour does no give he righ indicaion of he disribuion of admissions over he day. Therefore i is beer o calculae he variabiliy in he ime beeen admissions. This should hen also be calculaed as he MAPD. Variabiliy number of admissions during he day = A consan paern in he amoun of ork over he eek The firs problem for his performance measure is ho o deermine he amoun of ork on a ard. During he survey for he ards quesions ere asked abou his. According o 95% of he respondens he amoun of ork does no only depend on he number of beds occupied, bu also on he inensiy of care of he paiens in hose beds. Therefore e do ake his ino accoun. 37

39 In he inensiy of care ool of Isala Klinieken also inensiy of care of paiens is used. To make boh ools consisen and comparable, e use he same mehods as in he inensiy of care ool. In ha ool nurses of he ards should indicae per reason for admission and per day, he inensiy of care by giving hem 1 o 4 care poins. In he survey e checked heher i is necessary o differeniae he inensiy of care per day or even per hour. 55% of he respondens prefers he inensiy of care per day and 33% prefers i per hour. So i seems he righ choice o le he nurses indicae i per day. 93% of he respondens of he survey sae ha no only he reason for admission is imporan o deermine he inensiy of care, bu ha also oher paien characerisics should be considered. Hoever, i is no clear hich characerisics deermine he inensiy of care and ho hey should be combined o calculae i. Fasoli e al. (2011) give some indicaors for inensiy of care on a ard, like for example average lengh of say and he number of paiens ha are 70 years or older. Hoever, hey do no say ho o combine hese indicaors o deermine one number for he inensiy of care. Therefore i is no possible o include oher paien characerisics. For some paiens i migh in advance be clear for he planner ha he inensiy of care differs from he sandard, in ha case i should be possible o change i. In he survey also some oher aspecs are menioned ha cause ork. Some of hese aspecs canno be included because hey are difficul o predic. The oher aspecs consis of paien specific and ard specific asks. The paien specific asks are he admissions (95% of he respondens) and he discharges (85%). These asks should be prediced by he ool and care poins should be allocaed o hem. The inensiy of care ool also considers admissions and discharges. An example of a ard specific ask is he visis of he specialiss. This does ofen no differ per day. Therefore e do no consider his here. The amoun of ork can be calculaed using he folloing formula: Amoun of ork per day = In his formula he variables mean: A x A (h) D x D (h) I L,R x L,R care poins per admission number of admissions during hour h care poins per discharge number of discharges during hour h care poins per paien already L days in he hospial and ih admission reason R number of paiens a he ard ha are already L days in he hospial and have admissions reason R The inensiy of care for paiens is muliplied by 24 because hese care poins are based on he care needed per paien during one hour. For he number of paiens e again use he paiens a midnigh or he average beeen opening hours, depending on he ype of ard. From his ork per day he ool should deermine heher here is a consan paern over he eek. We use he seasonaliy index for his. The seasonaliy index is he ork on a cerain day divided by he average ork over he eek. The mean absolue deviaion from a sandard index per eekday indicaes heher i follos he consan paern. The formula for his is displayed belo. By comparing he seasonaliy index insead of he acual ork, his performance measure does no 38

40 depend on he oal ork during he eek. The sandard seasonaliy index per eekday should be deermined in advance. This could be based on hisorical daa or on preference of he ards. Deviaion from consan paern in amoun of ork over he eek = Variabiliy in he amoun of ork during he day hile considering oher asks a he ards For his performance measure also he amoun of ork is needed. For his he same approach is used as for he previous measure. The difference is ha no he ork should be deermined per hour. The ard specific asks should no be aking ino consideraion. We ill include his by changing he number of care poins ha one nurse can accomplish during an hour. The formula for he amoun of ork for his indicaor han becomes: Amoun of ork per hour(h) = The ork has o be consan over he day. For his e use he MAPD, jus like for he oher indicaors abou variabiliy. 86% of he respondens saes ha i is no necessary for he ork o be sable 24 hours per day, i is sufficien if i is consan during he day (opening hours). The admissions planner can only influence he variabiliy during he day, so e measure he variabiliy during hese hours. Variabiliy amoun of ork during opening hours = Saff availabiliy/produciviy The performance measures saff availabiliy and produciviy parly overlap. Saff availabiliy means ha on days ha feer personnel are scheduled also feer paiens should be scheduled. This acually means keeping he produciviy consan. Saff produciviy is he average number of paiens per nurse. For he saff availabiliy especially he number of nurses during he day is imporan, his is he number ha changes he mos beeen days. Therefore e calculae per day an average produciviy during he opening hours. Isala Klinieken does no ye have a saff scheduling sysem, so i is no possible o easily deermine he number of saff per shif. Hoever, ards do have a basic saff schedule. This schedule can be used o deermine he expeced produciviy per shif. To deermine heher he saff produciviy is consan, again e use he MAPD. Nex o ha he produciviy should be high enough such ha i mees a cerain norm. Therefore he average produciviy is also imporan. Saff produciviy = Saff availabiliy (variabiliy in produciviy) = 39

41 3.5.2 Lis of required informaion For he calculaions described above a lo of informaion is required. Some of his informaion should be deermined by he ards, he planners or he managemen of a specialism. Oher aspecs should be forecased based on he schedule. Table 3.9 gives a lis of seings needed and Table 3.10 describes he daa ha should be prediced by a model. Table Seings Seing Type of ard (purely day care, purely clinical or combinaion) Opening hours Number of beds available Lengh of say per reamen ype Sandard seasonaliy indices for amoun of ork per day Inensiy of care per reason of admission and per day Changes in inensiy of care for specific paiens Care poins per admission Care poins per discharge Number of care poins possible per nurse for each hour of he day Basic saff schedule per hour and per day of he eek Needed for Uilizaion Variabiliy number of used beds over he eek Uilizaion Variabiliy number of used beds over he eek Variabiliy amoun of ork over he day Uilizaion Overuilizaion Uilizaion Consan paern in amoun of ork over he eek Consan paern in amoun of ork over he eek Variabiliy amoun of ork over he day Consan paern in amoun of ork over he eek Variabiliy amoun of ork over he day Consan paern in amoun of ork over he eek Variabiliy amoun of ork over he day Consan paern in amoun of ork over he eek Variabiliy amoun of ork over he day Variabiliy amoun of ork over he day Saff availabiliy/produciviy Table Forecass Forecas Expeced number of used beds per hour Disribuion of he number of used beds per hour Expeced number of admissions per day Expeced imes beeen scheduled admissions Expeced number of discharges per day Expeced number of paiens per hour, for each reason of admission and for each number of days already in he hospial Expeced number of admissions per hour Expeced number of discharges per hour Needed for Uilizaion Variabiliy number of used beds over he eek Overuilizaion Variabiliy number of admissions over he eek Consan paern in amoun of ork over he eek Variabiliy number of admissions during he day Consan paern in amoun of ork over he eek Consan paern in amoun of ork over he eek Variabiliy amoun of ork over he day Variabiliy amoun of ork over he day Variabiliy amoun of ork over he day 40

42 Average uilizaion beeen 7 am and 6 pm Average uilizaion a midnigh Average uilizaion a midnigh 3.6 Curren performance Secion 3.4 gives a lis of performance measures hich he admissions planners should see abou he ards. We an o kno ha he curren performance is based on hese measures. The differen subsecions describe his curren performance per indicaor Uilizaion per ard Mos ards of Isala Klinieken are clinical ards. For hese ards he uilizaion a midnigh is imporan. This is displayed in Figure 3.6 and 3.7. The ards are divided ino general ards and inensive/specific ards. For he day care ards he uilizaion during he day is relevan. Figure 3.8 shos for hese ards he average uilizaion beeen 7 am and 6 pm. Wards A2 and A2K on locaion Weezenlanden are clinical ards, bu hey are also mean for day care paiens. Therefore for hese ards boh ypes of uilizaion are imporan. 100% 80% 60% 40% 20% 0% Figure Uilizaion a midnigh general ards (izis, 2011, n = paiens) 100% 80% 60% 40% 20% 0% Figure Uilizaion a midnigh inensive and specific ards (izis, 2011, n = paiens) 100% 80% 60% 40% 20% 0% Figure Uilizaion during opening hours (izis, 2011, n = paiens) 41

43 Proporion of hours ih overuilizaion Proporion of hours ih overuuilizaion Proporion of hours ih overuilizaion As can be seen in he graphs he uilizaion differs a lo per ard. Wha sands ou is he small uilizaion a midnigh a ards W A2 and W A2K. These ards are mean for shor say paiens and are herefore also open during he nigh, bu apparenly only a fe beds are used hen. Also he uilizaion on ards SZ K2 and W EHH is lo compared o he oher ards. The reason for he lo uilizaion a W EHH is ha i is an emergency ard for Cardiology and Lung medicine Overuilizaion per ard/ Overuilizaion on cerain momens of he day The curren performance for overuilizaion is he number of hours on hich here is overuilizaion compared o he number of hours on hich he uilizaion is larger han zero. This is comparable o he probabiliy of overuilizaion on a ard, hich ill be used in he ool. Figure 3.9 and 3.10 sho he performance for clinical ards and Figure 3.11 for day care ards. 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% Figure Overuilizaion general ards (izis, 2011, n = paiens) 12% 10% 8% 6% 4% 2% 0% Figure Overuilizaion inensive and specific ards (izis, 2011, n = paiens) 50% 40% 30% 20% 10% 0% Figure Overuilizaion day care ards (izis, 2011, n = paiens) 42

44 0:00:00 2:00:00 4:00:00 6:00:00 8:00:00 10:00:00 12:00:00 14:00:00 16:00:00 18:00:00 20:00:00 22:00:00 For general ards he proporion of hours ih overuilizaion is relaively small, ih one excepion, ard H0 on locaion Sophia. This is he ard for psychiary paiens. The reason for his is no knon. For inensive and specific ards he performance differs, bu here are no exreme ouliers. For day care ards he proporion of hours ih overuilizaion is relaively large, excep for ard K2 on locaion Sophia. For K2 his corresponds o he small uilizaion. When average uilizaion is lo i is unlikely ha he proporion of hours ih overuilizaion is large. The large probabiliy of overuilizaion on he oher day care ards is also explainable. I is possible for mos day care paiens o say in a chair insead of a bed a he end of heir admission period. These chairs are no considered in he calculaions. In advance e expeced ha ards ih a large average uilizaion also have a large proporion of hours ih overuilizaion. For inensive and day care ards his paern is visible in he graphs. Hoever, for general ards his is no he case, his probably has o do ih he variabiliy in he number of beds. The nex secion describes his aspec. One of he expeced reasons for overuilizaion on cerain momens of he day is he fac ha some ne paiens are already admied hile ohers have no ye been discharged. Figure 3.12 proves his idea: he peak of admissions happens before he peak of discharges. This resuls in a peak in he number of used beds in he morning (Figure 3.13) Admissions day care ards Admissions clinical ards Discharges day care ards Discharges clinical ards Figure Average number of admissions and discharges per hour (source: izis, 2011, n = paiens) 43

45 MAPD of number of beds used a midnigh MAPD of number of beds used a midnigh Number of used beds Day care ards Clinical ards Figure Average number of beds per hour (source: izis, 2011, n = paiens) Variabiliy in number of used beds over he eek Figure 3.14, 3.15 and 3.16 sho he variabiliy in he number of used beds over he days. I appears ha ards ih a very small average uilizaion, ofen have a large variabiliy in ha uilizaion, like for example W A2K, W EHH and SZ K2. This is logical, because hen a large variabiliy is a given, he only ay o deal ih his ihou creaing oo much overuilizaion is keeping he average uilizaion lo. The graphs also sho ha day care ards have a relaively large variabiliy in he number of used beds during he eek. 70% 60% 50% 40% 30% 20% 10% 0% Figure 3.14 Variabiliy in he number of used beds a midnigh general ards (izis, 2011, n =45083 paiens) 180% 160% 140% 120% 100% 80% 60% 40% 20% 0% Figure 3.15 Variabiliy in he number of used beds a midnigh inensive and specific ards (izis, 2011, n = paiens) 44

46 Number of used beds MAPD of number of beds used beeen 7 am and 6 pm 80% 60% 40% 20% 0% Figure 3.16 Variabiliy in he number of used beds during opening hours (izis, 2011, n = paiens) In he previous secion for some ards he probabiliy of overuilizaion could no be explained by a large uilizaion, like for example SZ B5G and W EHH. No i appears ha hese ards have a relaively large variabiliy in he number of used beds. So e can conclude ha causes of overuilizaion are a high average uilizaion and a large variabiliy in ha uilizaion. The variabiliy is deermined by calculaing he deviaion from he average uilizaion over a hole year. Hoever, his deviaion is no random; here is a paern during he eek (Figure 3.17). In he eekend feer beds are occupied, bu also during eekdays he number of used beds is no consan Day care ards Clinical ards Figure Average number of beds per eekday (source: izis, 2011, n = paiens) Variabiliy in number of admissions over he eek The nex indicaor is he variabiliy in he number of admissions over he eek. Figure 3.18, 3.19 and 3.20 sho he curren performance of his indicaor. When you compare Figure 3.16 ih Figure 3.20 i becomes clear ha for day care ards here is he same paern in he variabiliy in he number of beds as in he admissions. This is logical because paiens on hese ards leave before he end of he day, so he number of used beds is (almos) equal o he number of admissions. 45

47 MAPD of number of admissions per day MAPD of number of admissions per day MAPD of number of admissions per day 100% 80% 60% 40% 20% 0% Figure 3.18 Variabiliy in he number of admissions over he eek general ards (izis, 2011, n = paiens) 120% 100% 80% 60% 40% 20% 0% Figure 3.19 Variabiliy in he number of admissions over he eek inensive and specific ards (izis, 2011, n = paiens) 70% 60% 50% 40% 30% 20% 10% 0% Figure 3.20 Variabiliy in he number of admissions over he eek day care ards (izis, 2011, n = paiens) Anoher aspec ha sands ou is he high variabiliy for he inensive care unis. A probable reason for his is ha hese ards are scheduled by planners of muliple specialisms. If a ard is scheduled by one planner, i is easier o already consider he number of admissions for he ard. For he ICs here is no conrol for his a all. Also ard B3 on locaion Weezenlanden is an oulier, he reason for his is no knon. 46

48 Average MAPD of ime beeen admissions Average MAPD of ime beeen admissions Average MAPD of ime beeen admissions Variabiliy in number of admissions during he day For he variabiliy in he number of admissions during he day, e only considered days ih more han one admission. On days ih one admission i is no possible o disribue he admissions beer. Figures 3.21, 3.22 and 3.23 sho he resuls. The variabiliy is especially large for he day care and shor say ards, W A2 and W A2K. Hoever, here are no exreme ouliers. 100% 80% 60% 40% 20% 0% Figure 3.21 Variabiliy in he number of admissions during he day general ards (izis, 2011, n = paiens) 100% 80% 60% 40% 20% 0% Figure 3.22 Variabiliy in he number of admissions during he day inensive and specific ards (izis, 2011, n = paiens) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Figure 3.23 Variabiliy in he number of admissions during he day day care ards (izis, 2011, n = paiens) 47

49 Sandard seasonaliy index The reason ha his indicaor as menioned by managers of he ards is ha ofen a lo of admissions are scheduled a he same ime. Figure 3.12 already shos ha especially beeen 7 and 8 am a lo of paiens arrive A consan paern in he amoun of ork over he eek To calculae he amoun of ork for a ard e need informaion abou inensiy of care per ype of reamen and e need he care poins for admissions and discharges. These numbers should be deermined by saff of he ards. For mos ards hese numbers are no ye available. Hoever, he same numbers are also used in he inensiy of care ool. To ards already use his ool, so for hese ards i is possible o calculae he curren performance of his indicaor. These ards are SZ B5G and he Urology par of W A5. As he sandard seasonaliy indices e choose he averages per day of he eek over he hole year. Figure 3.24 shos hese indices for boh ards. The mean absolue deviaion from he sandard seasonaliy index for Urology a W A5 is 0,114. The MAPD is he same, because he average seasonaliy is 1. This is no exremely large, bu he amoun of ork also does no exacly follo a consan paern. For SZ B5G he MAPD is even larger, namely 0,215. So his ard follos even less a consan paern SZ B5G W A5 Urology Figure Sandard seasonaliy indices amoun of ork (izis, 2011, n = 2771 paiens) Variabiliy in amoun of ork during he day hile considering oher asks a he ards For ard SZ B5G he average MAPD in he amoun of ork during he day is 51,2%. This is very high. I means ha he ork is no evenly disribued over he day. For Urology on ard W A5 he MAPD is 23,5%, so on ha ard he ork is beer disribued over he day Saff availabiliy/produciviy Because Isala Klinieken does no have a saff scheduling program, i is no possible o calculae he saff availabiliy and produciviy for all ards. Hoever, he inensiy of care ool uses he basic saff schedule. So for he ards ha use his ool e can calculae he curren performance for his indicaor. For Urology on W A5 he average saff produciviy during opening hours in 2011 as 4,18 paiens per nurse. The mean absolue percenage deviaion in his produciviy is 21,4%. So he saff produciviy is large, bu here is also a large variabiliy in his produciviy. 48

50 Ward SZ B5G, Gynecology has an average saff produciviy of 2,94 paiens per nurse. This is much smaller han he produciviy of Urology, bu i is difficul o compare hese numbers, because i highly depends on he complexiy of paiens a a ard. The mean absolue percenage deviaion of he produciviy a SZ B5G is 25,9%. This is jus a lile bi larger han he variabiliy a he Urology ard. This chaper described he performance measures ha should be included in he ool. The lis, hich is deermined based on lieraure and inervies and a survey ihin he hospial, is displayed in Figure 3.5. This chaper also discussed a ay o measure hese performance indicaors and he curren performance based on hese measures. From he curren performance, i becomes clear ha improvemens migh be helpful. The ool developed in his sudy can conribue o his. For he ool o be effecive he ay of presening he performance measures is very imporan. The nex chaper ill focus on ha aspec. 49

51 Chaper 4 Tool expecaions From he previous chaper e kno hich performance measures are imporan for he ards. The nex quesion is ho hese measures should be presened in a ool. The admissions planners are imporan in his, because hey ill use he ool. Hoever, also he deparmen Paien Logisics is involved, because hey iniiaed he research. Secion 4.1 discusses he opinion of his deparmen. The opinion of he admissions planners is discussed in Secion 4.2. Finally in Secion 4.3 all hese aspecs are combined in a concepual model for he ool. 4.1 Deparmen Paien Logisics The deparmen Paien Logisics consiss of one employee, Bernd van den Akker. During an inervie ih him e discussed muliple aspecs abou he ool. Appendix E gives a ranslaion of he design of he inervies ih he admissions planners, during he inervie ih Bernd van den Akker he same design as used; only he firs eigh quesions ere skipped. Ineracion ih curren scheduling process The curren scheduling program focuses on a perspecive per paien. The ool should complemen his ih a ard perspecive. I should give an overvie insead of looking a individual paiens. I ould be ideal if he ool could generae an opimal schedule and he admissions planner only has o check heher he schedule is feasible. Hoever, i is no possible for a compuer o generae an opimal schedule because of many consrains. Therefore, i is beer if he ool shos he qualiy of a schedule hen he schedule is ready and before he paiens are noified abou he admissions. The momen ha a schedule is ready differs per deparmen. This depends on he scheduling horizon (van den Akker, 2012b). The ool should be inegraed ih he curren scheduling program. Hoever, his ill no be possible immediaely. Firs i should be proven ha i orks ouside he program (van den Akker, 2012b). Unil no e assumed he main goal of he ool should be o give he admissions planner insighs abou he curren schedule. Hoever, during he inervie i as menioned ha i should also be possible o use previous schedules as inpu. In his ay i is possible o see heher here is a rend in he qualiy of scheduling. Then you can also compare he expeced qualiy o he realized qualiy, hich indicaes ha influence he schedule has on he realizaion (van den Akker, 2012b). Layou The nex aspec discussed during he inervie is he layou of he ool. According o Bernd van den Akker (2012b) he ool should sho o abs, namely resuls and seings. The seings ab should for example sho he ards for hich he planner schedules, he eighing facors, he scheduling horizon ec. The resuls should be displayed in differen layers. The highes layer is he oal qualiy grade. The layer belo is a grade per performance measure, because ih only one oal grade he planner does no kno ho o improve i. Per performance measure here should also be a graph over ime ih norms. For he qualiy grades here should also be a norm. The norms ill differ per specialism and he grades canno be compared beeen specialisms, because he complexiy of scheduling differs a lo (van den Akker, 2012b). 50

52 Some planners schedule for muliple ards, if hese ards have a combined saff schedule he ool should combine he performance measures for hese ards. If his is no he case separae graphs and sub grades should be given for he ards (van den Akker, 2012b). Nex o hese layers he resuls ab should sho he rend of previous qualiy grades and he rend in difference beeen expeced qualiy and realizaion. This las aspec does no help o schedule beer, bu i does help he planner o rus he ool (van den Akker, 2012b). Ineracion beeen specialisms Some ards are used by muliple specialisms and scheduling for hese ards is herefore done by muliple planners. In he ne hospial his ill happen even more. According o Bernd van den Akker (2012b) scheduling hese ards should be done by one admissions planner or by muliple planners ha make he schedule ogeher. Therefore, e do no have o ake his ino accoun for he ool. Exreme examples of ards ha are used by muliple specialisms are day care and shor say ards. Also for hese ards i ould be bes if he planners ork ogeher o make he schedule (van den Akker, 2012b). Oher aspecs Finally i as menioned ha he ool should be easy o undersand, i should no cos much ime o use and i should be very flexible regarding he inpu needed for he calculaions (van den Akker, 2012b). 4.2 Admissions planners In Isala Klinieken a lo of people are involved in scheduling admissions. The surgical specialisms ofen have a planner or a planning eam ha schedules he admissions. For mos non-surgical specialisms he admissions are scheduled by he secrearies of he policlinic, because for hese paiens he admissions should ofen ake place ihin a fe days. Finally, for some day care ards he admissions are scheduled by a secreary of he ard. To ge a complee overvie of ha admissions planners expec of he ool, e approached planners from all hese differen areas for an inervie. We inervieed he planners from Inernal Medicine, Urology, Orhopedics, Gynecology, Cardiology and General Surgery (Smulders, 2012; Boerendans & Vos, 2012; Ooserveld, 2012; van den Berg, 2012; Schue, 2012; Huenesein & de Graaf, 2012; Vogelzang, 2012). Appendix E gives a ranslaion of he inervie design for hese inervies. General idea abou he research During he inervies i became clear ha i ill no be possible for all he specialisms o use he ool in he same ay. I ill differ per specialism for hich purposes he ool ill be used. The main purposes recognized during he inervies are: Making improvemens o a fuure schedule This is only possible if he schedule is made far enough in advance and if he appoinmens are no ye announced o he paiens. For his purpose, forecass are needed for he paiens ha are no ye scheduled, like for example he emergency paiens. Check aferards For he specialisms ha do no schedule far enough in advance, he ool can be used o indicae improvemens for he nex period and for monioring by managemen. For his i is 51

53 necessary ha he ool also orks for a complee schedule, he resuls should be visible for managemen and i should be possible o compare he resuls ih previous periods. Demonsraing o managemen ha he problems are Some admissions planners are no able o change much abou he effecs for he ards, ofen because hey are bounded by decisions on a higher level, like for example he OR block schedule or he division of specialiss over he blocks. The ool could help he planners o demonsrae hese problems, insead of jus basing i on a feeling. Using hese expecaions from he ool ill resul in less noise han using he realizaion on he ard. To demonsrae ha he problem is, i is necessary ha he ool gives specific and clear numbers per indicaor. During he inervie ih he planners of a day care ard i became clear ha hey no use anoher scheduling program han oher planners. Hoever, i is no necessary o focus on his, because i is he inenion ha all planners ill use he same scheduling program in he fuure. Ineracion ih curren scheduling process The nex subjec discussed during he inervies is he ineracion ih he curren scheduling process. One aspec ha is menioned ofen is ha i should be linked o he curren scheduling program, izis. Ho i should be linked is no clear. I could be compleely inegraed, bu a buon ha opens he ool is also sufficien. A leas i should use acual daa from he scheduling program. The planners prefer o use he ool hen one paien is scheduled. Hoever, for his already a lo is possible in he curren scheduling program. This informaion is ofen no used, bu i is available. Therefore, e focus on a ool ha is used afer a firs version of he schedule is available. This is recognized by mos of he planners. They menion ha he scheduling period for hich he ool can be used should no be oo shor, because han i is difficul o ge an overvie and o make changes. A final aspec menioned is ha he planners should be able o use he ool far enough ahead. This is useful for he ards, bu i is also he only ay in hich you sill can change somehing in he appoinmens ih paiens. Layou The main aspec menioned abou he layou of he ool is ha i should no become oo complex. The ideas abou ho o keep i simple differ. Examples are using colors and shoing graphs. All planners prefer separae graphs or numbers per performance measure and per ard. Mos of hem an o see a norm for he performance measures, bu only if his norm is no fixed in advance. No all planners agree o his. Displaying a norm gives he feeling ha you have o achieve somehing and ha you are judged. The ool should also sho he rend of he performance, such ha improvemens become visible and you can learn from i. Wih hese rends i can also be used as managemen informaion. Ineracion beeen specialisms In he curren siuaion some specialisms share he same ard. Ho he planners deal ih his differs a lo. For some ards agreemens are made abou he number of beds per specialism. If one specialism exceeds his number i should be discussed ih a cenral planner. Oher ards have 52

54 compleely separae beds and saff for each specialism. A final soluion is o jus schedule separaely and check righ before he schedule is execued if i fis. All hese ays of orking are no seen as an ideal siuaion. Soluions menioned are a cenral scheduling office ha sill has some decenralized planners, a number of specialism specific beds and some combined beds and finally, a predeermined number of beds per specialisms, bu his number mus differ per day. Oher aspecs During he inervies some oher, more pracical, aspecs ere menioned. I ould be useful if he resuls ould be prinable and donloadable o Excel. In ha ay you can also discuss i aferards and conrol for managemen becomes easier. The ool should give individual resuls, so per planner or eam. I should no be possible o see he resuls of oher deparmens. Flexibiliy is also very imporan. An example ha as given is he number of available beds. This number is smaller during reducion periods. A final hing menioned is he aspecs on hich he planners are judged by managemen and by specialiss. Compleely occupying he OR blocks has prioriy. Wihou changing his i is no possible for he planners o focus more on he effecs on he ards. This aspec as menioned by all he planners e inervieed. 4.3 Concepual model The aspecs menioned abou he ool in he previous secions can be combined ino a concepual model of he ool. The figures in Appendix F sho a concep version of he layou of he ool. The firs figure shos he resuls ab. This ab gives a oal number for he qualiy of he schedule and per indicaor i shos he performance and an explanaion in a graph. The second figure shos he layou of he second ab, hich concerns rends. For he oal qualiy and per indicaor i shos he rend in earlier calculaed resuls and in he realized performance. The final figure gives he layou of he seings ab. Here all he aspecs ha can be changed ill be displayed. Nex o his layou, he previous secions give some oher requiremens for he ool, hich are summarized belo: 1. Flexibiliy The ool should be flexible. I should be possible ha seings differ per specialism and per ard, bu hey should also be flexible over ime. Examples of such seings are he ards per specialism, he eighing facors for he differen indicaors, he norms per indicaor, he number of available beds and he scheduling horizon. 2. Simple The ool should no be oo complex. This means ha i is easy o undersand for he admissions planners, bu also ha i does no require much ime o use i. 3. Feasible for a fuure schedule and for schedules already performed I should be possible o calculae he qualiy of a schedule for over a fe eeks, so for a schedule ha is no complee because of emergencies ec. Hoever, also a previous schedule should be possible as inpu for he ool. 53

55 4. Uses daa from izis The ool should use he daa from he curren scheduling program, izis. 5. Individual resuls The ool should give resuls per planner or planning eam. According o one of he planners i should no be possible o see he resuls of ohers. Hoever, he deparmen Paien Logisics does no agree o his, i ould be good if planners could compare heir performance o ha of ohers. 6. Resuls visible for managemen The resuls should also be visible for he managemen of a specialism, such ha i can be used as managemen informaion. 7. Abiliy o donload daa o Excel I should also be possible o donload he daa o Excel. This has o do ih he previous requiremen. If i is possible o donload he daa o Excel, i can be used as managemen informaion even more. 8. Prinable resuls Admissions planners also an o use he informaion during progress meeings. For his i ould be useful if hey could prin he resuls. 9. Applicable in he ne hospial Because Isala Klinieken moves o a ne building, i is imporan ha he ool ill also be applicable here. To reach his i is should be easy o change he curren ards ino he ne larger ards in he ne hospial. These ne larger ards ill be used by muliple specialisms. Hoever, e do no focus on ha, because here ill probably a cenral planning eam per ard. This chaper described ho he ool should presen he performance measures deermined in Chaper 3. Appendix F shos figures of he concepual model and above a lis of oher requiremens is given. This describes he ideal siuaion for he proposed ool. We only develop par of his ool and fuure research is necessary o deermine if all hese aspecs described are needed o make he ool useful in he hospial. The nex sep is o forecas he daa needed for he performance indicaors. The nex chaper sars ih his by looking for exising models from lieraure. 54

56 55

57 Chaper 5 Exising models from lieraure No he performance indicaors and he expecaions of he ool are clear, i is ime o sar hinking abou he model ha has o forecas he daa necessary. We could of course consruc a model compleely from scrach and do everyhing ourselves, bu hy reinven he heel? Maybe here are already some models ha generae insighs in he effecs of an admissions schedule. A leas models exiss ha combine admissions schedules ih he processes a he ards, hose e can use as a saring poin. We sar his lieraure search ih some surveys in he area of healh care logisics. These surveys are described in Secion 5.1. From hese surveys e composed a lis of aricles ha could be ineresing in his research. These aricles are discussed in Secion Surveys We use hree surveys abou healh care logisics ha menion he ineracion beeen admissions or ORs and ards. The firs one is an aricle of Vanberkel e al. (2009). They sae ha aricles ofen focus on a single deparmen and hey give an overvie of aricles ha discuss he relaionship beeen differen deparmens. In oal hey found 88 aricles ha describe such a relaionship. The relaionship beeen he OR and he ard is found in 29 of he aricles. The nex survey is from Cardoen e al. (2010). The aricle focuses on lieraure abou operaing room planning and scheduling. The differen aricles are analyzed in differen ays. One ay is o look a he performance crieria. Ten aricles discuss he performance for ards. The final survey is rien by Guerriero and Guido (2011). This aricle discusses ho operaional research can be used in planning and scheduling for he ORs. Hereby, he focus is on mahemaical models. A disincion is made beeen aricles ha propose a model on a acical level and aricles ih a model on an operaional level. On a acical level hey find six aricles ha have leveling of ards and bed occupancy as an objecive and on he operaional level here are o aricles ha have he number of beds used as a crierion. 5.2 Aricles Alhough his research focuses on he operaional level, mos aricles abou he ineracion beeen admissions schedules and ards focus on higher hierarchical levels. There is one aricle ha focuses on he sraegic level and fifeen aricles abou he acical level. The models in hese aricles ill no be immediaely applicable for our research. Hoever, e could use i, or par of i, as a saring poin for our model. There are five aricles ha do focus on he operaional level. The aricles use differen approaches for analyzing he ineracion beeen admissions schedules and ards; his gives ideas for approaches in our model. Some aricles generae a schedule hile ohers jus evaluae hem. The goal of his research is no o generae a schedule, bu aspecs from research projecs ha do his could sill be useful. Table 5.1 gives an overvie of he aricles. I indicaes per aricle hich approach is used, heher a schedule is generaed, a shor descripion of he subjec and heher i useful for our model. Alhough all hese aricles discuss he ineracion beeen admissions schedules and ards, only a fe conain aspecs e can use in his research. The main reason ha some aricles are no applicable is heir focus. Some of he aricles focus on opimizaion (Beliën & Demeulemeeser, 2007; Beliën & Demeulemeeser, 2008; Cardoen, Demeulemeeser, & Beliën, 2009; Bekker & Koeleman, 56

58 2011; Sanibáñez, Begen, & Akins, 2007). In ha forecasing he effecs of a schedule is also necessary, bu doing his in an innovaive ay is no he main focus. Oher aricles focus on muliple consequences of an admissions schedule (Adan & Vissers, 2002; Adan, Bekkers, Dellaer, Vissers, & Yu, 2009; Van Oosrum, Van Houdenhoven, Hurink, Hans, Wullink, & Kazemier, 2006; Beliën, Demeulemeeser, & Cardoen, 2006). Therefore hey ofen use a simple, sraighforard calculaion is used per aspec. Nex o he focus of he aricles also he approach may cause ha is no applicable here. In some of he aricles simulaion is used (Griffihs, Price-Lloyd, Smihies, & Williams, 2005; Kim & Horoiz, 2002; Cochran & Bhari, 2006). I is possible o forecas he effecs of an admissions schedule by simulaion. Hoever, like menioned in Chaper 4, one of he consrains for he ool is ha i should give he resuls fas. Wih simulaion his is no possible. Final drabacks of some of he aricles are ha he explanaion is no exensive enough (Adan, Bekkers, Dellaer, Jeune, & Vissers, 2011; Sanibáñez, Begen, & Akins, 2007; Cochran & Bhari, 2006) or ha i does no fi he siuaion of our problem (Elkhuizen, Bor, Smeenk, Klazinga, & Bakker, 2007; de Bruin, Bekker, van Zanen, & Koole, 2010). Five of he aricles menioned in he able are useful for his research. We no discuss hese aricles in more deail. The aricle of Harrison e al. (2005) describes a simulaion model ha forecass he mean and he variabiliy of he uilizaion on a hospial level. The forecas is used o predic he ranges for beds needed in he fuure and o es if a change in he number of beds needed is an acual change or if i is jus par of he normal flucuaions. They ry o find a heoreical disribuion for he empirical daa of admissions and discharges. For he admissions a Poisson process ih differen mean arrival raes for each day of he eek is used. The disribuion for he discharges is more complex. For his muliple sages are defined. In every sage here are cerain probabiliies ha a paien is discharged, ha he paien says in his sage and ha he paien goes o he nex sage. The las sages correspond o he paiens ha have he longes lengh of say. In our model e could use his idea o predic hen a paien is discharged. Nex o his e could invesigae he idea of a Poisson process ih a differen mean for each day of he eek for he emergency paiens. Harris (1986) also gives a simulaion model. This model is mean for decisions on a acical level. The inpu for he model is a imeable for he ORs. This gives per OR block ho many major, inermediae and minor surgeries should be performed. Also he bed limis and he disribuion for he lengh of say per surgery ype are used as inpu. The oupu of he model is he bed requiremens, he occupancy and he number of cancellaions. To find he disribuion for he lengh of say hisorical daa and he inpu of he user is used o deermine he minimum, mean and maximum lengh of say. These are convered ino a cumulaive disribuion. We could use his idea for our model. The model of Vanberkel e al. (2011a) gives he disribuion for he number of used beds based on he MSS. The inpus for he model are he MSS and empirical disribuions for he number of paiens in an OR block and he lengh of say per specialy. The model consiss of hree seps, deermining he disribuion for a single block, for one MSS cycle and he seady sae disribuions for coninuous cycles. This final disribuion can be used o deermine ho many beds are needed hen you accep a cerain probabiliy of no having enough beds. Muliple sudens supplemened his model (Bosch, 2011; Smeenk, 2011; Vlijm, 2011; Vollebreg, 2011). We can use he research of Smeenk (2011) for our ool. He adaped he model such ha i gives disribuion of he number of used beds per hour insead of per day. The main difference is ha in our siuaion here is much more cerainy abou he number of paiens and he lengh of say, because he reamens for paiens are knon. 57

59 Table Overvie aricles Aricle Approach Generae schedule? 58 Subjec Sraegic level Harrison e al. (2005) Simulaion No Correcly forecasing he variabiliy in he occupancy level of beds Tacical level Adan and Vissers (2002) ILP Yes Admission profile o reach arge uilizaion of resources Adan e al. (2009) ILP Yes Include sochasic lengh of say in model Adan and Vissers (2002) Bekker and Koeleman (2011) Beliën e al. (2006) Queuing and Quadraic programming Visualizaion by sofare sysem No No Deermine admissions quoa and analyzing he impac of variabiliy in admissions on he bed capaciy Visualizaion of impac of MSS on demand for muliple resources Beliën and Demeulemeeser (2007) ILP Yes Muliple models o consruc MSS o minimize expeced oal bed shorage Beliën and ILP, column Yes Selec surgery schedule ha minimizes Demeulemeeser (2008) generaion required number of nurses Cochran and Bhari Queuing and No Making clear he complex siuaion of (2006) simulaion bed planning. De Bruin e al. (2010) Queuing heory No Evaluaing curren size of ards using he Erlang Loss model Elkhuizen e al. (2007) Spreadshee No Capaciy model ha gives insigh in he calculaions required nursing saff per shif Harris (1986) Simulaion No Predic he daily bed requiremens and cancellaions based on a cyclic imeable of operaions Kim and Horoiz (2002) Simulaion No Analyzing he effecs of a daily quoa sysem and reserving beds for elecive paiens on he performance of ICU Sanibáñez (2007) ILP Yes Schedule surgical blocks for specialies considering amongs oher hings possurgical resource consrains Van Oosrum e al. ILP, column Yes MSS o maximize OR uilizaion and (2006) generaion level beds Vanberkel e al. (2011a) Analyical, No Compuing donsream orkload saisical model disribuion as a funcion of an MSS Vanberkel e al. (2011b) Qualiaive Yes Use model Vanberkel e al. (2011a) o schedule evaluae proposals for a ne MSS generaion Operaional level Adan e al. (2011) MILP and No Translaion of acical plans ino decision rules operaional plans Cardoen e al. (2009) Column Yes Opimizing sequence of surgeries generaion based on muliple objecives Griffihs e al. (2005) Kusers and Groo (1996) Liig and Isken (2007) Simulaion model Analyical model Analyical model Useful for our model? Yes No No No No No No No No No Yes No No No Yes No Calculae he number of nurses needed per shif such ha ha-if analyses can be performed No No Predicion of resource availabiliy Yes No Shor-erm occupancy predicion Yes No No No

60 Kusers and Groo (1996) give a predicion of he availabiliy of he resources beds, nursing saff and ORs. Only he firs o migh be ineresing for our model. The predicions are used as decision suppor for he admissions planners. Each of he resources is spli ino differen disribuions ha can be added up. For he number of beds his is for example aiing lis admissions, emergency admissions, discharges from paiens ha are already in he hospial, discharges from scheduled admissions and discharges from emergency admissions. Per aspec he mean and variaion are calculaed. Afer implemening he model i appeared ha he variabiliy in occupancy of he OR decreased, bu he variabiliy in occupancy of beds increased. From his research especially he idea of spliing he calculaions in differen pars ha can be added, migh be ineresing for our model. The aricle of Liig e al. (2007) also gives predicions for he bed occupancy levels. Like he aricle of Kusers and Groo (1996), his aricle splis he admission and discharge flos in muliple aspecs ha can be added. Nex o ha i especially focuses on geing he correc daa for he calculaions. From his lieraure revie e can conclude ha here is already much research performed abou he ineracion beeen admissions schedules and ards. Hoever, mos of hese aricles focus on he acical level. I appears ha forecasing ha ill happen on he ards based on an admissions schedule on an operaional level has no ye been he subjec of many aricles. In he nex chaper e combine his lieraure revie and he conclusions from he previous o chapers ino our on ool. 59

61 Combinaion of scheduled and nonscheduled paiens Yes No Chaper 6 - Tool for insighs abou he ards Previous chapers discussed some aspecs ha are needed o consruc he final ool. This chaper combines his informaion and gives a descripion of he ool. Secion 6.1 builds upon he ideas for forecas models given in Chaper 5 such ha hese can be used in our siuaion. As can be seen in he concepual model in Chaper 4, he ool should also indicae a raing for he oal qualiy of an admissions schedule. For his o be possible, e need o kno ho o summarize he numbers of he differen indicaors ino one raing for he oal qualiy. Secion 6.2 describes his inegraion of he indicaors. Secion 6.3 gives a descripion of he ool, of ho i is developed and i discusses is verificaion and validaion. 6.1 Forecass Table 3.10 gave a lis of variables ha should be forecased o be able o calculae he performance measures. In his secion e explain per aspec ho he ool forecass i. Some forecass are a combinaion of calculaions for day care paiens and clinical paiens. This disincion needs o be made because he discharge process differs. Clinical paiens alays say a leas one nigh and he discharge ime does no depend on he lengh of say of he paien, hile day care paiens only say a fe hours and herefore he discharge ime depends on he lengh of say. A disincion beeen day care and clinical paiens is necessary if discharges are involved in he calculaions. Anoher difference beeen he forecass is ha some are based on a combinaion of scheduled and nonscheduled paiens. This disincion beeen scheduled and non-scheduled paiens is no necessary if he forecas depends on resuls of oher forecass or if e do no ake ino accoun non-scheduled admissions. Table 6.1 shos hich forecass are calculaed in hich ay. Appendix G gives an overvie of he noaions e use in he formulas in his secion. Table 6.1 Division forecass No Expeced imes beeen scheduled admissions (6.1.1) Expeced number of admissions (6.1.2) Combinaion of day care and clinical paiens Yes Expeced number of used beds per hour (6.1.5) Expeced number of paiens per hour, for each reason of admission and for each number of days already in he hospial (6.1.6) Disribuion of he number of used beds per hour (6.1.3) Expeced number of discharges (6.1.4) Expeced imes beeen scheduled admissions Because he exac arrival imes of non-scheduled admissions are difficul o forecas, e only consider he scheduled admissions hen calculaing he expeced imes beeen admissions. The imes of admissions can be deermined from he daa from he scheduling program. A problem ih his is ha if all admissions are scheduled a he same ime, here ould be no variabiliy in he imes beeen he admissions. Hoever, his ould no be a represenaion of he qualiy for he ards; if all paiens ould arrive a for example 10 am i is no ideal for he ard. To avoid his e use he sar and he end of he day as dummy admissions. For clinical ards i should be indicaed in he seings beeen hich hours of he day scheduled admissions could ake place. These are hen used as he sar and end of he day. 60

62 6.1.2 Expeced number of admissions We need o kno he expeced number of admissions per hour, E ], and per day, E a ]. In his [ a m, is he ard, m is he day in he scheduling horizon and is he imeslo (in his case he hour). Boh of hese forecass can be calculaed by adding he scheduled and non-scheduled admissions. The number of scheduled admissions per hour, E ], can be deermined from he daa of he [ sa m, scheduling program. The number of scheduled admissions per day is he sum of he number per hour. We assume, ihou loss of generaliy, ha here is a consan paern in he expeced number of non-scheduled admissions over he eek. Therefore, e deermine per eekday r he expeced number of non-scheduled admissions, E [ nsa r ], based on hisorical daa. We also assume, ihou loss of generaliy, ha he disribuion of he admissions over he day does no depend on he day. So e divide he expeced number of non-scheduled admissions over he hours in a ay ha is equal for all days, perca. This disribuion can be deermined based on hisorical daa. The formulas for he expeced number of admissions per hour and per day are hen as follos. In his Q is he las day of he scheduling horizon and T is he number of ime slos per day. [ am, ] E[ sam, ] E E[ nsa ]* perca for m=0,,q; =0,,T-1; r [ m E[ a m T ] 1 0 E[ sa m, ] E[ nsa r ] for m=0,,q; Disribuion of he number of used beds per hour This secion describes he disribuion of he number of used beds per hour. Smeenk (2011) already reformulaed he acical model of Vanberkel e al. (2011a) such ha i deermines a disribuion per hour. Therefore, e use he model of Smeenk (2011) as a saring poin for our calculaions. The main difference beeen our model and he model of Smeenk is ha in our case here is no recurring cycle and herefore here are already some paiens in he ard a he momen e sar calculaing. Anoher aspec ha differs is ha e have more informaion abou he paiens. We kno ho many scheduled paiens ill be admied each day and e kno more abou he discharge probabiliies of hose paiens because heir reamen ypes are knon. A final difference is ha Smeenk (2011) differeniaes beeen paiens ha are admied on he day of surgery and paiens ha are admied he day before. We do no make ha disincion. The calculaions for he disribuion of he number of used beds per hour are relaively complex and exensive. Therefore e only explain he general idea here, Appendix I describes he calculaions in more deail and gives he exac formulas. The basic idea of he calculaions is displayed in Figure 6.1. The oal disribuion of he number of used beds per hour and per day is calculaed by combining he separae disribuions for clinical and day care paiens. Boh of hese disribuions are calculaed by combining he disribuions for scheduled and non-scheduled paiens. For each of hese disribuions e group he paiens based on heir ime of admission. For clinical paiens his means grouping he paiens per admission day and for day care paiens per admission hour. A he momen of calculaion here are already some 61

63 paiens presen a he ard. Day care paien ha are already presen ill no be here anymore he nex day so e do no include hem. Hoever, clinical paiens, scheduled and non-scheduled, migh sill be here he nex day. We rea hese paiens as a separae group and include hem in he calculaions for he scheduled clinical paiens. Adding he disribuions for scheduled and non-scheduled paiens and grouping he paiens based on heir ime of admission is also done in he model of Smeenk (2011). Hoever, making a disincion beeen clinical and day care paiens and considering paiens already presen is no included in ha model. We no discuss he seps needed o calculae he disribuions per group of paiens. Toal disribuion Clinical paiens Day care paiens Scheduled paiens Non-scheduled paiens Scheduled paiens Non-scheduled paiens Paiens already presen Paiens admied day 0 Paiens admied hour 0 a day 0 Paiens admied hour 0 a day 0 Paiens admied day 0 Paiens admied day 1 Paiens admied hour 1 a day 0 Paiens admied hour 1 a day 0 Paiens admied day 1 Ec. Ec. Ec. Ec. Figure 6.1 Overvie calculaions disribuions of he number of used beds per hour Calculaion seps for groups of scheduled clinical paiens 1. Disribuion for he firs day, paiens already presen Unil he momen of calculaion here is cerainy abou he number of paiens ha are already presen. Afer ha every paien has a cerain probabiliy p o be discharged during a imeslo and a probabiliy 1-p of saying. If here are k paiens a he ard, he probabiliy of x paiens a he ard during he nex period can be calculaed using a binomial disribuion, k kx x ( p) (1 p). We kno he disribuion for he number of paiens a he ard during he x previous imeslo. We can use his disribuion ogeher ih he binomial formula o calculae he disribuion for he number of paiens a he ard during his imeslo. 62

64 2. Disribuion for he day of admission, paiens no ye admied Clinical paiens alays say one nigh, so for he day of admission you kno ih cerainy ho many paiens here ill be on he ard. 3. Disribuion for he oher days 3.1. Disribuion for he beginning of he day Disribuion for he firs day, paiens already presen Here again e can use a binomial disribuion like explained in he firs sep Disribuion for he day afer admission, paiens no ye admied Because no clinical paiens are discharged on he day of admission he number of paiens a he beginning of he day afer admission is knon ih cerainy Oher days For his a binomial disribuion can be used as explained earlier. The difference is ha e no use a probabiliy ha a paien is discharged during a hole day insead of during one imeslo Disribuion for he res of he day We use he disribuion for he beginning of he day as a saring poin and calculae he disribuions for he res of he day again ih a binomial disribuion using a probabiliy ha a paien is discharged during a imeslo. Calculaion seps for groups of non-scheduled clinical paiens The disribuion for he non-scheduled paiens does no depend on he exac dae. Hoever, e do assume ha i depends on he eekday. Therefore, e perform he seps described belo per eekday and aferards rerie i o he daes in he planning horizon. 1. Disribuion for he day of admission 1.1. Disribuion for he beginning of he day A he beginning of he day of admission i is cerain ha here are no paiens a he ard Disribuion for he res of he day We use a Poisson disribuion for he arrival process of non-scheduled paiens. On he day of admission no paiens are discharged, so his Poisson disribuion is sufficien for he calculaions. 2. Disribuion for he oher days 2.1. Disribuion for he beginning of he day Disribuion for he firs day afer admission Because no paiens are discharged during he day of admission e can again use he Poisson disribuion Disribuion for he oher days These calculaions are similar o hose for scheduled paiens. We use a binomial disribuion ih a probabiliy ha a paien is discharged during a cerain day Disribuion for he res of he day This is also similar o he scheduled paiens. We use a binomial disribuion ih a probabiliy ha a paien is discharged during a ime slo. 63

65 Calculaion seps for groups of scheduled day care paiens The calculaions for he groups of day care paiens differ from ha of he clinical paiens because no e use imeslos insead of days. For he clinical paiens hese days are divided ino imeslos, here e do no divide he imeslos ino smaller pars. This makes he calculaions less complex. 1. Disribuion for he hour of admission We assume ha day care paiens alays say a he hospial a leas one hour. Therefore e kno ih cerainy ho many paiens are a he ard during he hour of admission. 2. Disribuion for oher hours For he oher hours e use a binomial disribuion ih a probabiliy ha a paien is discharged during a ime slo. Calculaion seps groups of non-scheduled day care paiens For he non-scheduled day care paiens e perform he seps per eekday. So aferards e have o rerie his for he days in he planning horizon. 1. Disribuion for he hour of admission For he arriving process of non-scheduled paiens e use a Poisson disribuion. 2. Disribuion for oher hours For he oher hours e again use a binomial disribuion ih a probabiliy a paien is discharged during a ime slo. For he binomial disribuions e need he probabiliies ha a paien is discharged during a cerain ime slo or day. These probabiliies are based on he lengh of say disribuions per reamen ype and for he clinical paiens on a consan paern in hich discharges are disribued over he day Expeced number of discharges Lieraure gives o opions o deermine he expeced number of discharges. The firs one is described in he maser hesis of Smeenk (2011) and he second opion is he muli sage discharge model of Harrison e al. (2005). To deermine he disribuion for he number of used beds per hour, Secion 6.1.3, e used he model of Smeenk (2011). To keep he oupu and calculaions consisen e also use he ideas from his model for he discharges. Alhough e use he same idea as in he previous secion, hese calculaions are much easier because e need expeced values insead of disribuions. Here e also only explain he general idea of he calculaions. Appendix J describes i in more deail and gives he exac formulas. The general idea is displayed in Figure 6.2. The discharges of clinical paiens are firs calculaed per day. The discharges of clinical paiens per hour are based on his number and on a consan paern of ho he discharges are disribued over he day. For day care paiens i is he oher ay around. Firs he discharges are calculaed per hour and hen hese are added for he discharges per day. Again he paiens are grouped based on heir admission day or hour. The expeced numbers of discharges per group of paiens are based on he disribuion for he lengh of say and he expeced number of paiens in ha group. 64

66 Discharges per day Discharges per hour Clinical paiens Day care paiens (summaion of discharges per hour) Clinical paiens (discharges per day imes a percenage for ha hour) Day care paiens Scheduled paiens Non-scheduled paiens Scheduled paiens Non-scheduled paiens Paiens already presen Paiens admied day 0 Paiens admied hour 0 a day 0 Paiens admied hour 0 a day 0 Paiens admied day 0 Paiens admied day 1 Paiens admied hour 1 a day 0 Paiens admied hour 1 a day 0 Paiens admied day 1 Ec. Ec. Ec. Ec. Figure 6.2 Overvie calculaions expeced number of discharges Expeced number of used beds per hour The expeced number of used beds per hour can be calculaed based on he numbers from he previous secions. The number of used beds during his hour, E ], is he number of used beds [ b m, during he previous hour plus he paiens ha arrived during ha hour, he expeced number of admissions E ], minus he paiens ha lef, he expeced number of discharges E ]. [ a m, [ d m, There is a small difference in he calculaions for clinical and day care paiens. For clinical paiens you need o consider he paiens ha are already in he hospial a he beginning of he day. For day care paiens he number of paiens a he beginning of he day is zero. The formulas look like shon belo. As a saring poin e use he number of used beds a he momen he calculaions are performed. E[ bm, 1 ] E[ am, 1 ] E[ d m, 1 ] 0 E[ bm, ] for m=0,,q; =0,,T-1; E[ bm 1, T 1 ] E[ am 1, T 1 ] E[ d m1, T 1 ] 0 E[ b' 0 ] [ ] [ ' ] m, 1 E am, 1 E d m, 1 E[ b' m, ] for m=0,,q; =0,,T-1;

67 6.1.6 Expeced number of paiens per hour, for each reason of admission and for each number of days already in he hospial This secion is abou he expeced number of paiens per hour, for each reason of admission, j, and for each number of days already in he hospial, v. Day care paiens leave before he end of he day, so for hese paiens he aspec of he number of days already in he hospial does no have o be considered. The formulas are hen almos he same as in he previous secion. The only difference is ha no for he number of admissions and, j, j discharges e also need o consider he reason of admission, E [ ] and E [ d' ]. Hoever, i migh be difficul o forecas he number of non-scheduled paiens per reason of admission. Therefore, e rea he non-scheduled paiens as a separae reason of admission. The formula for, j he expeced number of day care paiens per hour and per reason for admission, E [ ' ], is: a m, m, b m,, j, j, j, j E[ b' 0, 1 ] [, 1 ] [ ', 1 ] m E am E d m E[ b' m, ] for m=0,,q; =0,,T-1; ; j 0 0 For clinical ards he calculaions are more difficul, in ha case paiens say a leas one nigh in he hospial, so e have o consider he number of days a paien is in he hospial. The paiens ha are in he hospial a a cerain momen of ime and are here for he same number of days are all admied on he same day. A he beginning of he admission day he number of paiens of his group is zero. The res of ha day paiens are admied and because i concerns clinical paiens, no discharges ake place. The nex days no paiens are admied anymore, on hese days discharges ake place. Jus like for he day care ards e no need o kno he expeced number of admissions, j per reason of admission, E [ ]. We even need o kno he discharges per reason of admission a m,, j, v and per number of days he paiens are in he hospial, E [ ]. To make i no oo complex e rea he non-scheduled paiens as a separae reason of admission. The formula for he expeced number of clinical paiens per hour, per reason for admission and per number of days already in he, j, v hospial, E [ ], is: b m, 0 0, v 0, j, v, j E[ b, ] E[ a ], j, v m m, 1 0, v 0 E[ bm, ], j, v1, j, v1 for m=0,,q; =0,,T-1; ; j; v E[ bm 1, T 1 ] E[ d m1, T 1 ] 0, v 0, j, v, j, v E[ bm, 1 ] E[ d m, 1 ] 0, v Inegraing performance measures Wih he forecass from he previous secion all daa ha is needed o calculae he performance measures is available. Hoever, e also need o combine hese performance measures o ge an aggregae number for he qualiy of he schedule. Combining he performance measures is called muli crieria analysis. There are differen approaches for his, bu mos of hem require alernaives o be knon. In our siuaion his is no he case, e do no have o make a decision beeen muliple alernaives; e jus an o kno he value of one alernaive. The mehod for muli crieria analysis ha corresponds o his is consrucing a single d m, 66

68 aggregaed objecive funcion. Muliple objecives, shon belo. x j, are combined ino one funcion, y, like y f ( x 1,..., x n ) Mehods ha are ofen used for an aggregaed objecive funcion are eighed sum and eighed produc. The formulas for hese are as follos. In his is he eigh for he objecive j. j eighed eighed sum n j1 n j1 produc j n x j j1 j ( x ) j j A problem of he eighed produc model is ha he oupu is no linear. For he ool i is imporan ha he oupu can easily be inerpreed as a number for he qualiy. This is easier ih a linear objecive funcion. For his reason e choose o use he eighed sum model. For his model e need o rerie he performance measures such ha hey are on he same scale. For his e use value funcions. This means ha he bes possible amoun for a cerain performance measure ges he value one and he ors possible amoun ges he value zero. The funcion for he amoun in beeen needs o be deermined by he decision maker. Hoever, his can make i very complex. Therefore e assume a linear value funcion for all performance measures. The bes and ors possible amouns sill need o be deermined. The bes amouns are sraighforard for all performance measures. For he performance measures probabiliy of overuilizaion and for all measures concerning variabiliy he bes possible amoun is zero. The oher performance measures are uilizaion and saff produciviy, hese need o be according o a cerain norm. For he objecive funcion e use he deviaion from his norm. Therefore he bes possible amoun here is also zero. The ors possible amouns are more difficul o deermine. This ill be differen per specialism. To deermine his e use hisorical daa. During he implemenaion for a specialism e consider he realizaion of he performance measures for he pas year and deermine he ors realizaion during ha year. Finally also he eighs for he performance measures need o be deermined. This is very subjecive and depends on he opinion of muliple sakeholders. Per performance measure he user of he ool should indicae o hich imporance caegory i belongs. Table 6.2 shos he caegories ih is corresponding eighs. Using hese eighs means ha a performance measure hich is no a all imporan ill no be aken ino accoun in he oal qualiy and for example a measure ha is indicaed as very imporan is ice as imporan as a measure ih he indicaion neural. 6.3 Tool for insighs abou he ards No all informaion ha is needed o develop he ool is available. Secion describes ho e developed he ool. The second secion gives he descripion of he proposed ool. Finally Secion discusses is verificaion and validaion. 67

69 Table 6.2 Imporance caegories Imporance Weigh No a all imporan 0 No imporan 1 Neural 2 Imporan 3 Very imporan Tool developmen To avoid misakes and unnecessary seps e used a sofare developmen mehod. Muliple mehods exis for sofare developmen. The mehod e use is Exreme Programming. Exreme Programming is an ieraive mehod for small o medium sized eams (Beck, 2000). The advanage of such an ieraive mehod is ha no all requiremens need o be knon in advance, he projec can evolve over ime. Exreme Programming is based on a se of principles. The lis belo gives he mos relevan principles for his projec (Beck, 2000): Small releases The hole projec should be divided ino subpars, called sories. During he projec already small orking pars should be delivered. Simple design In a previous chaper e deermined ha he ool should be kep simple. This does no only yield for he display, also for he coding behind i. Tesing Differen pars of he projec can only be compleed hen hey run a cerain number of ess. This esing should be done by he programmer (uni esing) and by he cusomer (funcional esing). Refacoring A ne version of he sysem should be made by adaping he exising design of he sysem. Afer adaping i should sill run all he ess. Cusomer involvemen In Exreme Programming he cusomer should be involved a lo during he developmen process. To creae small releases he projec should be divided ino muliple sories. The sories on he highes level are called releases (Beck, 1999). These releases can again be divided ino sories, hich are called ieraions (Beck, 1999). To complee an ieraion muliple asks are needed. The cusomers ha should be involved in he developmen are he deparmen Paien Logisics and he admissions planners. Involving admissions planners of all specialisms ould become oo complex. Therefore e only include he planner of Urology in his process. The cusomers are involved in esing he ieraions and releases. Nex o ha, he deparmen Paien Logisics is involved in selecing he sories for he releases and ieraions. In his ay e make sure ha he mos imporan aspecs are compleed firs. Our ool can be divided ino hree main pars, he resuls, he rends and he seings. The seings and he resuls are very dependen on each oher. So hey canno be seen as differen releases. Hoever, he resuls and seings ogeher can be developed ihou he rends ab. We chose o 68

70 focus on he release resuls/seings firs. This is he only possible sequence because he rends depend on he resuls. While looking a he available daa from he scheduling program i becomes clear ha informaion abou fuure schedules is sored differenly han informaion abou schedules ha are already performed. We choose he fuure schedules as he sub release e focus on firs. These decisions are made ogeher ih he deparmen Paien Logisics. The resuls and seings for schedules in he fuure can be divided ino ieraions by looking a he differen indicaors. Because of ime consrains e are no able o perform all hese ieraions. Togeher ih he deparmen Paien Logisics e decided o focus on he indicaors admissions over he eek and overuilizaion. The ard of Urology is a clinical ard, bu someimes also day care paiens are admied. Therefore he calculaion of he forecass should be suiable for he combinaion of clinical and day care paiens. Figure 6.3 gives an overvie of he releases and ieraions jus menioned. Each ieraion can be divided ino muliple asks. Tool Resuls/ seings Trends Fuure schedules Schedules already performed Toal qualiy Uilizaion Admissions over he eek Admissions over he day Overuilizaion Work amoun over he eek Work amoun over he day Saff Figure 6.3 Releases and ieraions Because he ool ill no be complee a he end of his research, i is imporan ha he deparmen Paien Logisics is able o coninue developing he ool. For his reason e program he ool in VBA in Excel Tool descripion We developed par of he proposed ool like described in he previous secion. Figure 6.4 and Figure 6.5 sho ho his ool looks like. The ex in he figures is in Duch, because his version of he ool is 69

71 acually used by he admissions planner of Urology. The ool can be used as follos. Firs Excel documens need o be donloaded from he scheduling program, izis. These are documens ih informaion abou he scheduled paiens and abou he paiens ha are already a he ard. Nex he ool can be opened and some seings migh be changed, like he scheduling horizon. When he admissions planner presses he buon Bereken ( Calculae ), he ool reads he daa from he donloaded Excel documens, calculaes he variables needed and shos i in he chars. Afer calculaion he buon Grafieken versuren ( Send chars ) becomes visible. Pressing his buons opens an message ih an Excel documen ih he chars as an aachmen. This makes i possible for he admissions planner o send he chars o for example he ard or managemen. If he ool is closed he calculaed daa is saved in he Excel shee, such ha i can be used for validaion, hich is described in he nex secion. Unil no e used he ool o ge he daa for he ard of Urology for one eek in advance. When e do no save sub resuls for verificaion he calculaion ime is only a fe seconds. Urology has a relaively small ard. For specialisms ih larger ards he calculaion ime ill be longer Tool verificaion and validaion To make sure he ool shos he correc informaion o he planners e need o verify and validae he ool. Because e only programmed par of he ool, e only verify and validae his par. Firs e verify he ool o make sure he calculaions in he ool correspond o he mahemaical formulas discussed earlier in his chaper. This as already one of he aenion poins hen esing of he sofare developmen. Addiional verificaion is done by calculaing (sub) resuls by hand (using Excel) and comparing his o he oupu of he ool. For he admissions over he eek he verificaion is relaively simple, bu for he overuilizaion i is more complex. Therefore, e save as much sub resuls as possible in he Excel shees. For some sub resuls e simplified he inpu of he model. During he verificaion some small changes ere needed, bu a he end he sub resuls ere as expeced hen looking a he mahemaical model. The nex sep is o validae he ool, his means checking heher he oupu of he ool corresponds o he realizaion. The problem ih his is ha he realizaion can be influenced by oher facors han he admissions schedule, for example by decisions made a he OR or he ard. We also validae he ool in anoher ay, namely discussing i ih he admissions planners and he ard, his is called face validiy. Validaion is already parly done by he admissions planner hen esing he ool. Hoever, i is imporan ha he oupu of he ool represens he experiences a he ards. For he validaion of he admissions over he eek e compare he oupu of he ool ih he acual number of admissions over he eek. We used he ool for a fe eeks for Urology. The oupu of he ool and he realizaion is shon in Figure 6.6. From his i seems ha he paern in he expeced number of admissions is similar o ha in he acual number of admissions. This is also recognized by he employees a he ard. The deviaions are mainly due o non-scheduled paiens and no shos. The mean absolue percenage error for he expeced number of admissions is 30%. To be more cerain abou he correcness more daa is needed. 70

72 Figure 6.4 Prin screen proposed ool (seings ab) Figure 6.5 Prin screen proposed ool (resuls ab) 71

73 Number of admissions Number of admissions Oupu ool Realizaion Figure 6.6 Validaion expeced number of admissions over he eek Figure 6.7 shos he probabiliy of overuilizaion for he Urology ard for he period during hich e used he ool. The verical lines are he momens on hich ne forecass are calculaed. On hese momens he probabiliy of overuilizaion decreases. This is logical because on hese momens here is more cerainy abou he number of paiens a he ard hen hen you have o predic i a eek ahead. The employees a he ard recognized he paern in he probabiliy of overuilizaion. A he end of he eek i is ofen mos difficul o arrange everyhing. The oupu of he ool shos ha during he day he larges probabiliy of overuilizaion is ofen a he beginning of he afernoon, his is also recognized by he ard. The probabiliy of overuilizaion canno be compared o he realizaion daa as easily as he number of admissions. In he realizaion he probabiliy of overuilizaion is jus 0% or 100%. To validae i e divide he scale of probabiliy of overuilizaion in five inervals, 0%-20% ec., and calculae per inerval during hich percenage of he hours here as overuilizaion. We expec ha in he inervals ih higher probabiliies, overuilizaion ould occur more ofen. The resuls are shon in Figure 6.8. For he firs four inervals our expecaions seem o be correc. Hoever, for probabiliies of overuilizaion beeen 80% and 100% overuilizaion occurs less ofen. An explanaion for his migh be ha hen here is a large probabiliy of overuilizaion, he ards recognize his and make changes o he schedules o preven overuilizaion. Wih a smaller probabiliy of overuilizaion he pressure o change he schedule migh no be large enough. The period for hich e used he ool is relaively small. To be more cerain abou he validiy of he oupu of he ool more daa is necessary. 72

74 Proporion of hours ih overuilizaion Probabiliy of overuilizaion Probabiliy of overuilizaion 100% 80% 60% 40% 20% 0% 100% 80% 60% 40% 20% 0% Figure 6.7 Probabiliy of overuilizaion Urology 50% 40% 30% 20% 10% 0% 0%-20% 20%-40% 40%-60% 60%-80% 80%-100% Forecased probabiliy of overuilizaion Figure 6.8 Validaion probabiliy of overuilizaion In his chaper e described ho o forecas he daa needed o calculae he performance measures. Nex o ha e discussed ho o combine he performance measures ino a raing for he oal qualiy using he eighed sum mehod. And finally e developed par of he ool using Exreme Programming. From he verificaion and validaion i became clear ha he ool generaes useful oupu for he admissions planner. Alhough e already developed par of he ool his should of course no be he end of he projec. Evenually a complee ool should be used by all admissions planners in he hospial. The nex chaper discusses ho o reach his. 73

75 Chaper 7 - Implemenaion Alhough he ool is no ye ready o be implemened e do describe an implemenaion plan for he ool. Secion 7.1 describes he seps ha should be aken afer his research and Secion 7.2 discusses he sakeholders ha are involved. 7.1 Implemenaion seps This secion describes he seps ha are needed during he implemenaion. 1. Develop he ool The firs sep should be o complee he ool. This means programming he remaining releases and ieraions displayed in Figure 6.1. This developmen should be done according o he sofare developmen mehod described in Secion Tes ool ihin one specialism Tesing he ool as already par of he developmen phase. Hoever, also afer ha he ool should be used ihin one specialism for some ime before coninuing he implemenaion. This period could especially be used o collec enough informaion for he nex sep in he implemenaion. 3. Convince hospial managemen Suppor from he managemen of he hospial is very imporan o successfully implemen he ool in he hole hospial. This suppor is needed o ge resources for he res of he implemenaion. Managemen should be convinced ha he ool has enough advanages for he hospial. 4. Creae urgency Currenly, he main focus hen scheduling admissions is compleely filling he operaing rooms. The goal of he ool is o le he admissions planners also consider he ards. Hoever, if hey ill be judged on he qualiy of heir schedule for he operaing rooms here ill be no urgency for hem o use he ool. This change in focus should be iniiaed on a sraegic level. Hospial managemen should suppor i and communicae i o he res of he organizaion. 5. Develop he ool in curren sofare We programmed he firs version of he ool in VBA in Excel. Hoever, his is no an appropriae programming environmen hen implemening an applicaion ino he hole organizaion. The bes soluion ould be o inegrae he ool in he curren applicaions of he hospial. 6. Implemen he ool for all specialisms The nex sep is o implemen i in he hole organizaion. This should no be done for he hole organizaion a he same ime, bu in such a ay ha enough suppor can be given o he users. This implemenaion sep consiss of muliple sub seps. 1. Convince managemen Also ihin one specialism suppor of managemen is very imporan. Managemen should make sure ha he focus is more on ards, like explained in Secion Inform employees In he previous sep e already menioned managemen, bu also oher employees should be informed abou he implemenaion of he ool. For example employees on he ards and specialiss, hey are no direcly involved in using he ool, bu hey could experience he consequences hen admissions planners schedule differenly hen using he ool. 74

76 Lo Influence High 3. Make he ool suiable The ool is developed in such a ay ha i is applicable for mos of he specialisms. Hoever, per specialism specific seings migh be needed and hey migh use i in a differen ay. 4. Train users The admissions planners ill be he people ha acually use he ool. They should be rained for his. The informaion ha he ool gives is ne o hem, so he raining should no only focus on using he ool, bu also on inerpreing he oupu. Nex o he admissions planners, managemen could also ge valuable informaion from he ool. Therefore hey should also be informed abou he ay in hich hey can use he oupu. 5. Use he ool The final sep in he implemenaion per specialism is o acually le he admissions planners use he ool. I ould be bes no o le hem use all indicaors in he ool a once. Only shoing a fe of he indicaors a firs les hem ge used o he ne ay of orking. 7. Monioring Afer he ool is implemened in he hole organizaion he projec is no complee. The final sep is monioring. During his sep problems in he usage of he ool should be solved. I also should be made sure ha everyone uses he ool in he righ ay. Finally he ool should coninuously be improved. We discuss some ideas for his in Secion 8.3 in he nex chaper. 7.2 Sakeholder analysis During he implemenaion seps discussed in he previous secion muliple sakeholders are involved. Figure 7.1 displays hese sakeholders in a frameork hich is based on he model of Michell e al. (1997). We esimaed per sakeholder heir involvemen, heir influence and heir opinion abou his projec. The posiion of he sakeholders ihin his frameork indicaes ho o cope ih hem during he projec. Hospial managemen Managemen specialism Specialiss Keep saisfied Admissions planners Manage closely Negaive Neural Posiive Minimal invesmen Keep informed Wards ICT Lo Involvemen High Figure 7.1 Sakeholder analysis The implemenaion is he final sep in he projec. In he nex chaper e look back ih some conclusions and a discussion, bu e also look ahead; ha are opics for fuure research? 75

77 Chaper 8 - Conclusion and recommendaions In he beginning of his research e deermined he main objecive for he projec. No i is ime o reflec on his. Secion 8.1 gives he main conclusions from our research. The discussion in Secion 8.2 describes he value of he research, bu also is shorcomings. Finally, Secion 8.3 discusses our recommendaions for Isala Klinieken. 8.1 Conclusions In Chaper 1 of his repor e formulaed he objecive of his research as: Generaing insighs in he effecs of operaional offline scheduling decisions made for hospial admissions on he capaciy planning processes a he ards. During he research i became clear ha hese insighs should be given by using a ool ha shos he admissions planners he qualiy of heir schedule. I requires hree seps o ge from an admissions schedule o he insighs for he planners. Firs muliple aspecs of he ards need o be forecased based on he schedule. Nex he forecass should be summarized ino performance measures. Finally he performance measures need o be displayed o he admissions planners. This research as divided according o hese required seps. We sared ih he performance measures. Based on inervies and a survey on all hierarchical levels ihin Isala Klinieken e consruced a lis of performance measures hich are imporan for he ards. This resuled in a long lis of performance measures. Therefore, e reduced his lis by using evaluaion crieria and by looking a he imporance based on he inervies and he survey. Finally e can conclude ha he performance measures ha should be shon o he admissions planners are: Average uilizaion per ard Variabiliy in he number of used beds over he eek Saff produciviy Variabiliy in amoun of ork during he day Consan paern in amoun of ork over he eek Variabiliy in admissions during he day Variabiliy in admissions over he eek Saff availabiliy Overuilizaion per ard Overuilizaion on cerain momens of he day To deermine ho hese performance measures should be displayed o he admissions planners, e inervieed muliple planners and he deparmen Paien Logisics. I became clear ha he ool should summarize he performance measures ino a raing for he oal qualiy of a schedule. Nex o his he ool should sho he rends of he oal qualiy and of he performance measures. Anoher imporan requiremen for he ool is ha i should be flexible. This is ranslaed ino a possibiliy for admissions planners o change muliple seings in he ool. To deermine he performance measures from an admissions schedule, muliple forecass are needed. The mos complex forecass are he expeced number of discharges and he disribuion for he number of used beds per hour. We based he calculaions of hese forecass on he model of Smeenk (2011). 76

78 Because of ime consrains e ere only able o develop par of he ool, namely he resuls for he admissions over he eek and he overuilizaion for fuure schedules. We did his developmen according o he sofare developmen mehod Exreme Programming. During he developmen he ool as esed by he admissions planner of Urology. From he verificaion and validaion i became clear ha he ool generaes useful oupu for he admissions planner. Therefore, e can conclude ha i is possible o generae insighs in he effecs of scheduling decisions made for hospial admissions on he capaciy planning processes a he ards. 8.2 Discussion Wih he proposed ool from his research Isala Klinieken has an indicaion of he qualiy of an admissions schedule on he operaional offline level. This is a ne ay of orking in hospials. Currenly people mosly look a realizaion daa o evaluae he qualiy of processes. There are some developmens in forecasing on he operaional online level, bu no ye on he offline level. Like concluded in he lieraure revie in Chaper 5, research abou he ineracion of ORs and ards ofen focuses on he acical level. So also in ha field his poin of vie is relaively ne. We performed our research for Isala Klinieken and herefore e only used informaion from his hospial. Because of his e canno conclude somehing abou he exernal validiy of he research, more research in muliple hospials is needed for his. If e ould need o perform his research again, e ould probably do some hings differenly. Firs, i ould be beer o design he inervies and survey in Chaper 3 in a more srucured ay. For he inervies ih he planners e did make a srucured design, herefore afer hese inervies e ere more cerain e go all he informaion e needed. Nex o his during he inervies and he survey e asked he respondens some quesions ha already indicaed cerain ansers. For example in he survey e focused on he amoun of ork, because his aspec as menioned a lo in earlier conversaions. Amoun of ork as an imporan aspec according o he respondens, bu he resuls migh be differen if e did no focus on his aspec ha much. Anoher discussion poin is he ay e evaluaed he performance measures in Secion 3.4. This evaluaion as done by o people. A lo of measures needed o be evaluaed based on muliple crieria. This makes i difficul o have a good overvie. Therefore, he agreemen as very small a firs and he discussion consised for a large par of explaining he measures and crieria. I ould have been beer o spend more ime on evaluaing he measures. Neverheless, leing muliple people evaluae he performance measures as relevan, e believe ha i improved he resuls. 8.3 Recommendaions In his secion e describe our recommendaions for Isala Klinieken. We give recommendaions in hree areas. The firs subsecion describes he main recommendaions resuling from his research. During he projec e ere also involved in some oher areas relaed o our research. Secion gives some recommendaions based on his. Finally Secion discusses he possibiliies for fuure research Recommendaions from research Based on our research e have four recommendaions for Isala Klinieken. The firs o are sraighforard; e recommend Isala Klinieken o coninue developing he proposed ool and o implemen he ool like discussed in Chaper 7. I could evenually improve he processes a he 77

79 ards, bu firs of all i helps recognize he problems in he ineracion beeen he admissions schedule and he ards. Our nex recommendaion is o change he focus ihin Isala Klinieken more o he ards. During he research i became clear ha a lo of people in he organizaion recognize he need for insighs for he admissions planners abou he ards. This indicaes ha here are opporuniies for his projec in he organizaion. Hoever, Isala Klinieken is jus a he sar of his process. One of he main barriers in his is ha he OR deparmen is seen as he mos imporan resource. The admissions planners are judged based on he qualiy of heir schedule for he OR. For he proposed ool o be effecive his focus in he organizaion needs o change. I should become possible for he admissions planners o ake ino accoun he ards in heir decisions. Finally o be able o give admissions planners relevan insighs in he processes a he ards i is mos pracical if he admissions for one ard are scheduled by one admissions planner or a leas one eam. In he curren siuaion his is he case for a lo of ards, bu in he ne hospial he ards ill be shared beeen specialisms. Also for ards ha are used by almos all specialisms, like day care ards, i is imporan ha here is some coordinaion in he scheduling of admissions Oher recommendaions In he firs chaper of his repor e described he ineracion beeen he admissions schedule and he ards on all hierarchical levels. In his research e focused on he operaional offline level, bu based on he descripion in he firs chaper e can also give some recommendaions abou he oher levels. In Isala Klinieken a lo of informaion is available abou he effecs of he decisions for he ORs on he ards. We recommend o acually use his informaion for saff scheduling a he ards. On he sraegic level here is informaion abou producion agreemens, on he acical level he acical planning ool predics he effecs of he OR block schedule on he ards and on he operaional level Isala Klinieken has he inensiy of care ool. During he inervies i became clear ha beeen he OR block schedule and he admissions schedule here is anoher level in he organizaion, namely he schedule for he specialiss. We recommend o also use his level o influence he processes a he ards and o give he ards more informaion for saff scheduling Fuure research The firs opic for fuure research is ho o change he focus in Isala Klinieken from he OR deparmen o he ards. In Secion e already menioned his aspec. Ho his should be done is no knon, more research is needed for his. Secondly, fuure research is needed abou insighs in oher resources han he ards. In his research e mainly focused on insighs in he ards. Hoever, admissions planners have o ake ino accoun more resources, like for example he IC, he ORs and he X-ray. During some of he inervies i as menioned ha he planners should also ge insighs ino hese oher resources. The planner of he specialism Cardiology already akes ino accoun all needed resources per admission. More research is necessary o find ou heher i ould resul in improvemens if his is done in he hole organizaion. 78

80 The hird fuure research aspec is he correlaion beeen he performance measures. In Chaper 3 e deermined he performance measures ha should be shon o he admissions planners. In his research e did no consider correlaion beeen hese measures because of o reasons. Firs, e don kno heher he correlaion sill exiss if he planner uses he measures o change he schedule. Secondly, he correlaion should also be recognized by he admissions planner. If he planner is convinced ha he measures are independen, i is beer o sho hem boh. For hese reasons i is beer o look a correlaion afer he proposed ool is used for some ime. Afer ha i should be deermined, ogeher ih he planner, if some performance measures can be removed. For he proposed ool iself more research is needed o improve he inpu of he ool. The ool uses amongs oher hings he expeced number of non-scheduled paiens and a disribuion for he lengh of say as inpu. For he non-scheduled paiens e no look a hisorical daa abou emergency paiens. Hoever, non-scheduled paiens are no only emergency paiens and he number also depends on he ime unil he admission day. The longer he remaining ime unil he admission day, he more paiens ill be added o he schedule. The disribuion for he lengh of say per reamen ype is no deermined based on hisorical daa. Hoever, some reamens only consis of a fe paiens per year. In ha case his disribuion is no reliable. Also e do no consider he expeced lengh of say indicaed by he ards. This could give more informaion ih more cerainy abou he lengh of say. More research is needed o improve hese o ypes of inpu of he ool. Mos of he inpu needed for he ool is based on hisorical daa. We also recommend linking he ool ih he daa arehouse, such ha he inpu can coninuously be adaped according o ne hisorical daa. A final opic for fuure research is he ay o inegrae he performance measures ino a raing for he oal qualiy of a schedule. We described a mehod for his in Secion 6.2. Hoever, his mehod appears o be very dependen on amongs oher hings he ard and he lengh of he scheduling horizon. This makes i difficul o compare his number beeen ards and over ime. More research is needed o make his raing for oal qualiy more consisen. 79

81 Bibliography.isala.nl/overisala. (2012). Rerieved February 27, 2012, from Isala: hp://.isala.nl/overisala/profiel/pages/defaul.aspx.isalabou.nl. (2012). Rerieved February 27, 2012, from Isala bou: hp://.isalabou.nl/isalabou/pages/defaul.aspx Adan, I., & Vissers, J. (2002). Paien mix opimisaion in hospial admission planning: a case sudy. Inernaional Journal of Operaions & Producion Managemen, 22(4), Adan, I., Bekkers, J., Dellaer, N., Jeune, J., & Vissers, J. (2011). Improving operaional effeciveness of acical maser plans for emergency and elecive paiens under sochasic demand and capaciaed resources. European Journal of Operaional Research, 213(1), Adan, I., Bekkers, J., Dellaer, N., Vissers, J., & Yu, X. (2009). Paien mix opimisaion and sochasic resource requiremens: A case sudy in cardiohoracic surgery planning. Healh care managemen science, 12(2), Aiken, L. H., Clarke, S. P., Sloane, D. M., Sochalski, J., & Silber, J. H. (2002). Hospial Nurse Saffing and Paien Moraliy, Nurse Burnou, and Job Dissaisfacion. The Journal of he American Medical Associaion, 288(16), Beck, K. (1999). Embracing change ih exreme programming. Compuer, 32(10), Beck, K. (2000). Exreme programming explained; embrace change. Addison Wesley Longman, Inc. Bekker, R., & Koeleman, P. M. (2011). Scheduling admissions and reducing variabiliy in bed demand. Healh care maangemen science, 14(3), Beliën, J., & Demeulemeeser, E. (2007). Building cyclic maser surgery schedules ih leveled resuling bed occupancy. European Journal of Operaional Research, 176, Beliën, J., & Demeulemeeser, E. (2008). A branch-and-price approach for inegraing nurse and surgery scheduling. European Journal of Operaional Research, 189(3), Beliën, J., Demeulemeeser, E., & Cardoen, B. (2006). Visualizing he Demand for Various Resources as a Funcion of he Maser Surgery Schedule: A Case Sudy. Journal of medical sysems, 30(5), Boerendans, G., & Vos, J. (2012, May 21). (A. Hooijsma, Inervieer) Zolle. Bosch, J. (2011). Beer uilisaion of he OR ih less beds. Broers, E. (2012, April 3). (A. Hooijsma, Inervieer) Zolle. Cardoen, B., Demeulemeeser, E., & Beliën, J. (2009). Sequencing surgical cases in a day-care environmen: An exac branch-and-price approach. Compuers & Operaions Research, 36(9), Cardoen, B., Demeulemeeser, E., & Beliën, J. (2010). Operaing room planning and scheduling: A lieraure revie. Elsevier, 201(3), Cochran, J. K., & Bhari, A. (2006). Sochasic bed balancing of an obserics hospial. Healh care managemen science, 9(1), Cohen, J. (1960). A coefficien of agreemen for nominal scales. Educaional and Psychological Measuremen, 20(1), de Bruin, A., Bekker, R., van Zanen, L., & Koole, G. (2010). Dimensioning hospial ards using he Erlang loss model. Annals of Operaions Research, 178(1),

82 Doran, G. T. (1981). There's a S.M.A.R.T. ay o rie managemen's goals and objecives. Managemen Revie, 70(11), Elbrech, G. J., & Seenbergen, R. (2012, April 13). (A. Hooijsma, Inervieer) Zolle. Elkhuizen, S. G., Bor, G., Smeenk, M., Klazinga, N. S., & Bakker, P. J. (2007). Capaciy managemen of nursing saff as a vehicle for organizaional improvemen. BMC Healh Services Research, 7(1), Fasoli, D. R., Fincke, B. G., & Haddock, K. S. (2011). Going Beyond Paien Classificaion Sysems o Creae an Evidence-Based Saffing Mehodology. The Journal of Nursing Adminisraion, 41(10), Gorard, S. (2005). Revisiing a 90-year-old debae: he advanages of he mean deviaion. Briish Jourlan of Educaional Sudies, 53(4), Griffihs, J., Price-Lloyd, N., Smihies, M., & Williams, J. (2005). Modelling he requiremen for supplemenary nurses in an inensive care uni. Journal of he Operaional Research Sociey(56), Guerriero, F., & Guido, R. (2011). Operaional research in he managemen of he operaing heare: a survey. Healh care managemen science, 14(1), Hannink, A. (2012, March 28). (A. Hooijsma, Inervieer) Zolle. Hans, E. W., van Houdenhoven, M., & Hulshof, P. J. H. (2011). A Frameork for Healh Care Planning and Conrol. Rerieved from hp://purl.uene.nl/publicaions/76144 Harris, R. (1986). Hospial bed requiremens planning. European Journal of Operaional Research, 25(1), Harrison, G. W., Shafer, A., & Mackay, M. (2005). Modelling Variabiliy in Hospial Bed Occupancy. Healh Care Managemen Science, 8(4), Hings, W. (2012, April 4). (A. Hooijsma, Inervieer) Zolle. Huenesein, W., & de Graaf, J. (2012, June 7). (A. Hooijsma, Inervieer) Zolle. Hurs, K. (2010). Evaluaing he srenghs and eaknesses of NHS orkforce planning mehods. Nursing Times, 106(40). Kim, S.-C., & Horoiz, I. (2002). Scheduling hospial services:he efficacy of elecive-surgery quoas. The Inernaional Journal of Managemen Science, 30(5), Klappe, A. (2012, March 29). (A. Hooijsma, Inervieer) Zolle. Kusers, R. J., & Groo, P. M. (1996). Modelling resource availabiliy in general hospials, Design and implemenaion of a decision suppor model. European Journal of Operaional Research, 88(3), Landis, J. R., & Koch, G. G. (1977). The Measuremen of Observer Agreemen for Caegorical Daa. Biomerics, 33(1), Li, L., & Benon, W. (1996). Performance measuremen crieria in healh care organizaions: Revie and fuure research direcions. European Journal of Operaional Research, 93(3), Liig, S. J., & Isken, M. W. (2007). Shor erm hospial occupancy predicion. Healh care managemen science, 10(1), Livak, E., Buerhaus, P. I., Davidoff, F., Long, M. C., McManus, M. L., & Berick, D. M. (2005). Managing Unnecessary Variabiliy in Paien Demand o Reduce Nursing Sress and Improve Paien Safey. Join Commission Journal on Qualiy and Paien Safey, 31(6),

83 Michell, R. K., Agle, B. R., & Wood, D. J. (1997). Toard a Theory of Sakeholder Idenificaion and Salience: Defining he Pinciple of Who and Wha Really Couns. The Academy of Managemen Revie, 22(4), Nielsen, J., & Landauer, T. K. (1993). A mahemaical model of he finding of usabiliy problems. INTERACT '93 and CHI '93 conference (pp ). Ne York: ACM Ne York. Ooserveld, J. (2012, May 10). (A. Hooijsma, Inervieer) Zolle. Sanibáñez, P., Begen, M., & Akins, D. (2007). Surgical block scheduling in a sysem of hospials: an applicaion o resource and ai lis managemen in a Briish Columbia healh auhoriy. Healh care managemen science, 10(3), Schue, S. (2012, May 22). (A. Hooijsma, Inervieer) Zolle. Shabin, A., & Mahbod, M. A. (2007). Prioriizaion of key performance indicaors: An inegraion of analyical hierarchy process and goal seing. Inernaional Journal of Produciviy and Performance Managemen, 56(3), Smeenk, H. (2011). Predicing bed census of nursing ard from hour o hour. Amserdam. Smulders, J. (2012, May 9). (A. Hooijsma, Inervieer) Zolle. van Apeldoorn, A. (2012, April 11). (A. Hooijsma, Inervieer) Zolle. van den Akker, B. (2012a, May 7). Zolle. van den Akker, B. (2012b, May 2). (A. Hooijsma, Inervieer) Zolle. van den Berg, B. (2012, May 16). (A. Hooijsma, Inervieer) Zolle. van Hoorn, A., & Wend, I. (2008). Benchmarking: een kesie van leren. Houen: Drukkerij Badoux. Van Oosrum, J. M., Van Houdenhoven, M., Hurink, J., Hans, E. W., Wullink, G., & Kazemier, G. (2006). A maser surgical scheduling approach for cyclic scheduling in operaing room deparmens. OR specrum, 30(2), Vanberkel, P. T., Boucherie, R. J., Hans, E. W., Hurink, J. L., & Livak, N. (2009). A Survey of Healh Care Models ha Encompass Muliple Deparemens. Deparmen of Applied Mahemaics, Universiy of Tene, Enschede. ISSN Vanberkel, P. T., Boucherie, R. J., Hans, E. W., Hurink, J. L., van Len, W. A., & van Haren, W. W. (2011b). Accouning for Inpaien Wards When Developing Maser Surgical Schedules. Aneshesia & Analgesia, 112(6), Vanberkel, P., Boucherie, R., Hans, E. W., Hurink, J., Len, W. v., & Haren, W. v. (2011a). An exac approach for relaing recovering surgical paien orkload o he maser surgical schedule. Journal of he Operaional Research Sociey, 62, Veraar, S. (2012, April 12). (A. Hooijsma, Inervieer) Zolle. Vlijm, R. (2011). Relaing he maser surgery schedule o he orkload a he nursing ards. Zolle. Vogelzang, M. (2012, June 14). (A. Hooijsma, Inervieer) Zolle. Vollebreg, R. (2011). Breaking don he alls beeen OR and ard. 82

84 83

85 Appendices Appendix A OR block schedule Locaion Weezenlanden OR1 OR2 OR4 OR5 OR6 OR7 M A M A M A M A M A M A Monday ORT ORT ORT ORT URO URO ENT ENT ORT ORT JAW JAW Tuesday ORT ORT ORT ORT URO URO ENT ENT ORT ORT JAW JAW Wednesday ORT ORT ORT ORT URO URO ENT ENT URO URO JAW JAW Thursday ORT ORT ORT ORT URO URO ENT ENT DS DS JAW JAW Friday ORT ORT URO URO ENT ENT Monday ORT ORT ORT ORT URO URO ENT ENT ORT ORT URO URO Tuesday ORT ORT ORT ORT URO URO ENT ENT ORT ORT JAW JAW Wednesday ORT ORT ORT ORT URO URO ENT ENT ORT ORT URO URO Thursday ORT ORT ORT ORT URO URO ENT ENT DS DS JAW JAW Friday ORT ORT URO URO ENT ENT JAW JAW OR8 OR9 OR10 OR11 Day care OR 1 Day care OR 2 M A M A M A M A M A M A Monday GY GY THO THO THO THO THO THO DS DS OPH OPH Tuesday THO THO THO THO THO THO THO THO ORT OPH ORT OPH Wednesday THO THO THO THO THO THO THO THO OPH OPH OPH OPH Thursday THO THO THO THO THO THO THO THO ORT ORT OPH OPH Friday THO THO THO THO THO THO OPH ORT OPH ORT Monday GY GY THO THO THO THO THO THO DS DS OPH OPH Tuesday THO THO THO THO THO THO THO THO ORT OPH ORT OPH Wednesday THO THO THO THO THO THO THO THO OPH OPH OPH OPH Thursday THO THO THO THO THO THO THO THO ORT ORT OPH OPH Friday THO THO THO THO THO THO OPH ORT OPH ORT Legend GY Gynecology JAW Ja surgery ENT Ear, nose and hroa surgery DS Special denal surgery ORT Orhopedics THO Thoracic surgery URO Urology OPH Ophhalmology Empy/flexible slos 84

86 Locaion Sophia OR1 OR2 OR3 OR4 OR5 M A M A M A M A M A Monday NE NE GS GS EM EM GS GS NE NE Tuesday GS GS GS GS EM EM GS GS NE NE Wednesday GS GS GS GS EM EM GS GS NE NE Thursday NE NE GS GS EM EM GS GS NE NE Friday NE NE GS GS EM EM GS GS NE NE OR6 OR7 OR8 Day care OR 1 Day care OR 2 M A M A M A M A M A Monday GS GS GY GY PS PS GS GS GS GS Tuesday GY GY GY GY PS PS PS PS GS GS Wednesday PS PS GY GY PS PS ENT ENT ENT ENT Thursday GS GS GY GY GS GS GY GY GS GS Friday PS GY GY GY PS PS PS/ENT PS/MRI PS/ENT PS/MRI Legend EM GS GY NE PS ENT Emergency OR General surgery Gynecology Neurosurgery Plasic surgery Ear, nose and hroa surgery Empy/flexible slos 85

87 Appendix B Inervie design acical level A he beginning of he inervie e gave a shor inroducion ino he research and he goal of he inervie. Afer ha he folloing quesions ere asked: 1) Wha is, afer lisening o his explanaion, you firs idea abou his research? And ha could you conribue o i? 2) Wha are he asks and decisions for he ards in hich you are involved? 3) Wha are he asks and decisions for he admissions schedules in hich you are involved? 4) On hich aspecs do you judge he ards? 5) In lieraure a disincion is made beeen financial performance and qualiy performance, and beeen inernal and exernal performance (see able belo ih examples). Considering his frameork, are here any oher aspecs on hich you judge he ards? Financial Qualiy Inernal Uilizaion Level bed uilizaion Exernal Liquidiy Cancellaions 6) Wha ould be he ideal siuaion for he ard from your poin of vie? 7) Wha should he admissions planners ake ino accoun hen scheduling he admissions? 8) Which informaion should hey see for his? 9) Does your specialism share ards ih oher specialisms? (only for managers of specialisms) a) Ho does he ineracion go? b) Wha role should he ool play in his ineracion beeen specialisms? 10) In he ne hospial some hings ill change, also for he ards. Are here aspecs ha e should ake ino accoun hen developing he ool o make sure he ool is also useful in he ne hospial? 11) Wha should be he resuls afer implemening he ool (he final goal)? 12) Are here aspecs hich are no ye discussed, bu hich are imporan o ake ino accoun during his research? 86

88 Appendix C Inervie design operaional level A he beginning of he inervie e gave a shor inroducion ino he research and he goal of he inervie. Afer ha he folloing quesions ere asked: 1) Wha is, afer lisening o his explanaion, you firs idea abou his research? And ha could you conribue o i? 2) Wha are he asks and decisions for he ards in hich you are involved? 3) Which aspecs are imporan for a good admissions schedule for he ards? 4) In lieraure a disincion is made beeen financial performance and qualiy performance, and beeen inernal and exernal performance (see able belo ih examples). Considering his frameork, are here any oher aspecs on hich you judge he ards? Financial Qualiy Inernal Uilizaion Level bed uilizaion Exernal Liquidiy Cancellaions 5) Wha ould be he ideal siuaion for he ard from your poin of vie? 6) Wha should he admissions planners ake ino accoun hen scheduling he admissions? 7) Which informaion should hey see for his? 8) Is here a paern in hen emergencies arrive (during he day and during he eek) for some ards? And is his differen per ard? 9) Wha deermines he amoun of ork on a ard? 10) Wha ould be he bes ay o deermine inensiy of care? 11) Should he amoun of ork be leveled over he eek? 12) Should he amoun of ork be leveled during he day? 13) In he ne hospial some hings ill change, also for he ards. Are here aspecs ha e should ake ino accoun hen developing he ool o make sure he ool is also useful in he ne hospial? 14) Wha should be he resuls afer implemening he ool (he final goal)? 15) Are here aspecs hich are no ye discussed, bu hich are imporan o ake ino accoun during his research? 87

89 Appendix D Survey quesions For my sudy Indusrial Engineering and Managemen I perform a research projec for he deparmen Paien Logisics of Isala Klinieken (my supervisor is Bernd van den Akker). The research is abou he effecs of an admissions schedule on he ards. An admissions planner schedules he admissions for a specialism hile looking especially a he reamens (for example he OR). The ards adap o his schedule. The number of (scheduled) paiens on a ard differs a lo. A reason for his is ha admissions planners a his momen do no have insighs ino he effecs of heir schedule on he ards, le alone ha hey could ake i ino accoun. The goal of my research is o change his. The firs quesion in his is: hich insighs should he admissions planners have o be able o ake he ards ino accoun? This especially depends on ha you as saff of he ards hink is a good schedule. By filling ou his survey you can help me in his. Thank you very much for your cooperaion and good luck in compleing he survey! Sincerely, Annemaaike Hooijsma (a.hooijsma@isala.nl) General A hich ard do you ork (muliple ansers possible)? o SZ-A1 o o WL-B6 o Oher namely: For hich specialism do you ork a his ard (muliple ansers possible)? o Aneshesiology o o Urology o Oher namely: Is his ard only mean for day care paiens? o Yes o No 88

90 Schedule previous eek If you hink back o he previous eek, on hich days did you hink he admissions schedule as good (muliple ansers possible)? And hy as he schedule on hese days good? o Monday o o Sunday o There as no day ih a good schedule Reasons (a he nex quesion you can explain he days ih a poor schedule): On hich day did you hink he admissions schedule as no good (muliple ansers possible)? And hy? o Monday o o Sunday o There as no day ih a poor schedule Reasons: Imporan aspecs Belo a lis is displayed of aspecs ha could be imporan for a good schedule for a ard (hen looking a scheduled admissions). Could you indicae per aspec ho imporan his is for your ard? No all aspecs ill be compleely feasible, bu he admissions planner could ake i ino accoun, ha makes a leas an improvemen possible. No a all imporan No imporan Neural Imporan Very imporan Sufficien beds for he scheduled paiens o o o o o Sufficien beds reserved for emergencies o o o o o The paiens (from your specialism) on he righ ard o o o o o A consan amoun of ork during he day o o o o o The same amoun of ork on differen days o o o o o A predicable amoun of ork o o o o o 89

91 Menion a leas one oher aspec ha is imporan for a good admissions schedule looking from he perspecive of a nurse. Menion a leas one oher aspec ha is imporan for a good admissions schedule looking from he perspecive of he paien. Mos ards in Isala Klinieken have a KPI dashboard and/or a day board. On hese board amongs oher hings indicaors are displayed hich he ard uses for conrol. Are on he dashboard and/or day board of your ard indicaors displayed hich should also be knon by he admissions planner, such ha he/she can ake hem ino accoun hen scheduling he admissions? If yes, ha are hese indicaors? o No o Yes, namely: Emergencies Indicae if he folloing saemens are rue or false. A paern is recognizable in hen emergencies arrive during he day. So on cerain momens of a day ofen a lo of emergencies arrive. o True o False A paern is recognizable in hen emergencies arrive during he eek. o True o False Amoun of ork In previous quesions amoun of ork is already menioned a fe imes. Hoever, ha is amoun of ork on a ard exacly? Make a choice beeen he folloing o saemens: o The number of used beds says enough abou he amoun of ork o Also he inensiy of care of he paiens in he beds is imporan for deermining he amoun of ork 90

92 If e an o deermine he amoun of ork based on he inensiy of care of he paiens, here has o be a ay o deermine his inensiy of care. Of course he inensiy of care ill differ per paien and per nurse, bu i is imporan o give an esimaion of i. The nex quesions are abou his. Make a choice beeen he folloing hree saemens: o Inensiy of care of a paien is almos consan, so one number for he inensiy of care per paien is sufficien o I should be possible o indicae he inensiy of care of a paien per day o I should be possible o indicae he inensiy of care of a paien per hour Indicae if he folloing saemen is rue or false: The inensiy of care of a paien also depends on he reamen phase (for example jus before or afer a surgery) o True o False Make a choice beeen he folloing o saemens: o Inensiy of care based on he ype of reamen gives a good indicaion o Inensiy of care differs a lo per paien hile he reamen migh be he same. So he inensiy of care should be based on paien characerisics. (only visible if on he previous quesion he second anser as given) You menion ha inensiy of care should be based on paien characerisics. Which paien characerisics, ha are knon in advance, give enough informaion of inensiy of care (muliple ansers possible)? o ASA classificaion (deermined by he anesheis) o ADL scale (aciviies of daily living scale) o Oher namely: Nex o (inensiy of care of) he paiens ha are presen a he ard, here migh also be some oher aspecs ha lead o amoun of ork. Indicae heher he folloing aspecs lead o ha much amoun of ork ha he admissions planners should have insighs in i (muliple ansers possible). o Number of admissions o Number of discharges o Oher namely: 91

93 Indicae heher he folloing saemen is rue or false: If on a cerain day he number of discharges and admissions is small, i is no a problem if here are a lo of paiens a he ard. o True o False Is i possible o compare he differen aspecs of amoun of ork? For example, a discharge on average akes ice as much ime as an admission. Or a discharge akes on average hree imes as much ime as an average paien ha you have o care for during a day. o Yes o No Amoun of ork on differen days The folloing quesions are abou a consan amoun of ork on differen days, so for example on Tuesday here is he same amoun of ork as on Monday. Wheher he amoun of ork should also be equally disribued over he day, ill be discussed laer. When ansering hese quesions, assume ha amoun of ork ill be measured like you indicaed during he previous quesions. Make a choice beeen he folloing saemens: o I ould be ideal if he amoun of ork ould be equal for all (eek)days o Amoun of ork doesn have o be equal for all days, as long as here is consan paern in he amoun of ork during he eek (so on Monday alays he same amoun of ork as on oher Mondays ec.) o Amoun of ork doesn have o be equal for all days or according o consan paern. (only visible for non-day care ards and if on he previous quesion he firs anser as given) You indicae ha he amoun of ork should be equal on differen days, bu hich days do you mean? o Amoun of ork should be consan on eekdays, bu no on eekend days o Amoun of ork should be consan on boh eekdays and eekend days. Hoever, i is no a problem if amoun of ork is smaller during he eekend. o Amoun of ork should be consan for all days (so during he eekend he same amoun of ork as on eekdays) 92

94 (only visible for non-day care ards and if on he firs quesion of his subsecion he firs or second anser as given) Imagine, during a eek a 0:00 here is alays he same number of paiens (ih he same inensiy of care) on your ard, nex o ha every day has he same number of admissions and discharges. Would his mean ha he amoun of ork is equal for all days? If no, hy? o Yes o No, because Amoun of ork during he day Previous quesioned discussed a consan amoun of ork on differen days. This does no mean ha he amoun of ork is also equally disribued over he day. The nex quesion ill deal ih ha. Make a choice beeen he nex hree saemens: o Amoun of ork should also be consan during he day. o Amoun of ork does no have o be consan during he day, as long as i follos a consan paern (for example alays a high amoun of ork beeen 10 and 11 am) o I is no necessary ha he admissions planner ries o ge he amoun of ork consan or according o a consan paern during he day, his ill go auomaically. (only visible for non-day care ards and if on he previous quesion he firs or second anser as given) Make a choice beeen he folloing o saemens: o Amoun of ork should be consan during dayime (for example beeen 8:00 and 17:00), bu i is no a problem if i is smaller during he nigh. o Amoun of ork should be consan during he hole day (24 hours) (only visible for non-day care ards and if on he firs quesion of his subsecion he firs anser as given) Make a choice beeen he folloing o saemens: o A consan amoun of ork during he day is only necessary for eekdays o A consan amoun of ork during he day is necessary for all days (only visible for non-day care ards and if on he firs quesion of his subsecion he second anser as given) Make a choice beeen he folloing o saemens: o A consan paern for he amoun of ork during he day is only necessary for eekdays o A consan paern for he amoun of ork during he day is necessary for all days 93

95 Oher commens Are here any aspecs ha ere no menioned in his survey, bu hich are imporan o consider during his research? Thank you for compleing his survey! You can close he survey by pressing he buon Send. If you have any quesions or commens, please le me kno (a.hooijsma@isala.nl). 94

96 Appendix E Inervie design admissions planners A he beginning of he inervie e gave a shor inroducion ino he research and he goal of he inervie. Also e inroduced he performance indicaors menioned during he previous inervies and survey. Afer ha he folloing quesions ere asked: 1. Could you shorly explain ho you make an admissions schedule? 2. Ho far ahead do you make he schedule? 3. Does he schedule change a lo afer i is ready, for example because of emergencies? 4. Wha is, afer lisening o his explanaion, your firs idea abou his research? 5. Which aspecs of he ards do you already see in he curren scheduling program? 6. Do you also consider hese aspecs hen making he schedule? 7. Which aspecs abou he ards ould you prefer o see nex o he ones already menioned? 8. If all he performance measures ha are menioned ould be shon in he ool, i could become very unclear. Only hose measures hich are really useful for you as a planner should be displayed. Could you rank he menioned performance measures based on heir imporance? 9. Should he ool be combined ih he curren scheduling program? And ho? 10. A hich momen in he scheduling process should he ool be used? 11. Imagine ha you open he ool, ha ould you an o see? Ho ould i look like on he screen? 12. Ho should he performance measures be presened? 13. Should he ool also give norms? 14. Should he ool also give he performance of previous periods? 15. Do you schedule admissions for muliple ards? If yes, ho should he ool display his? 16. Do oher specialisms also schedule for he ards for hich you schedule? a. Ho is his ineracion going no? b. In he ne hospial his ill happen even more because of he larger ards. Ho should his be arranged? Imagine you are making he schedule before anoher specialism, ho could you ake his oher specialism ino accoun? 17. Are here aspecs hich are no ye discussed, bu hich are imporan o ake ino accoun during his research? 95

97 Appendix F Concepual model 96

98 97

99 98

Public Works Department Project Charter

Public Works Department Project Charter Public Works Deparmen Projec Charer Projec Name: Kingshighway Signalizaion Improvemen Projec Deparmen: _Public Works Division: _Engineering Projec Number: _CMAQ-7302(630) Accoun Number: _412-500-501-873-101

More information

Scheduling Activities for the Patrol Boat Force

Scheduling Activities for the Patrol Boat Force Scheduling Aciviies for he Parol Boa Force Ping Cao, Greg Searle DSTO Defence Sysems Analysis Division Mark Horn, Houyuan Jiang, Phil Kilby CSIRO Mahemaical and Informaion Sciences July 2003 Repor Documenaion

More information

10/31/15. WellStar Health System. Effectiveness of an Evidence-Based Practice Nurse Training Program. Background/Significance.

10/31/15. WellStar Health System. Effectiveness of an Evidence-Based Practice Nurse Training Program. Background/Significance. Effeciveness of an Evidence-Based Pracice Nurse Training Program WellSar Healh Sysem -For-Profi 5-Hospial Sysem ~ 1294 Beds Inegraed Healh Sysem: 2 Healh Parks, 1 Nursing Home, 2 Hospices, 8 Urgen Care

More information

The Growth in Home-Based Wage and Salary Employment in the United States, : How Much and Why?

The Growth in Home-Based Wage and Salary Employment in the United States, : How Much and Why? The Growh in Home-Based Wage and Salary Employmen in he Unied Saes, 1980-2000: How Much and Why? Gerald S. Oeinger! Deparmen of Economics Universiy of Texas a Ausin oeinge@eco.exas.edu April 2006! The

More information

Advanced Practice Nursing Master s Degree Option: Nurse-Midwifery (49 credits)

Advanced Practice Nursing Master s Degree Option: Nurse-Midwifery (49 credits) Advanced Pracice Nursing Maser s Degree Opion: Nurse-Midwifery (49 credis) The Marquee Universiy Nurse-Midwifery Program (NMP) is designed o prepare scholarly, clinically specialized nurses who will funcion

More information

IS LAMB PROMOTION WORKING?

IS LAMB PROMOTION WORKING? IS LAMB PROMOTION WORKING? Oral Capps, Jr. Gary W. Williams* TAMRC Commodiy Marke Research Repor No. CM-01-07 November 2007 * Capps is Professor of Agriculural Economics and Souhwes Dairy Markeing Endowed

More information

,II II' GUIDELINES FOR THE DEVELOPMENT OF SMALL SCALE RURAL WATER SUPPLY AND SANITATION PROJECTS IN ETHIOPIA. l 'S /\11) CR5 *******

,II II' GUIDELINES FOR THE DEVELOPMENT OF SMALL SCALE RURAL WATER SUPPLY AND SANITATION PROJECTS IN ETHIOPIA. l 'S /\11) CR5 ******* l 'S /\11) CR5 *******,II II' CATHO IC RELIEF SERVICES GUIDELINES FOR THE DEVELOPMENT OF SMALL SCALE RURAL WATER SUPPLY AND SANITATION PROJECTS IN ETHIOPIA GUIDELINES FOR THE DEVELOPMENT OF SMALL-SCALE

More information

Regulation Conditions Influence on the Thermic and Exergic Efficiency Factors Size with the Heat Exchanger of a Coiled Type

Regulation Conditions Influence on the Thermic and Exergic Efficiency Factors Size with the Heat Exchanger of a Coiled Type Journal of Energy and Power Engineering 10 (016) 87-91 doi: 10.1765/1934-8975/016.05.003 D DAVID PUBLISHING Regulaion Condiions Influence on he Thermic and Exergic Efficiency Facors Size wih he Hea Exchanger

More information

EXPLAIN THE CULTURAL NEEDS OF EMPLOYEES, ISLAMIC AZAD UNIVERSITY

EXPLAIN THE CULTURAL NEEDS OF EMPLOYEES, ISLAMIC AZAD UNIVERSITY Arabian Journal Business and Managemen Review (OMAN Chaper) Vol., No.8, March. 0 EXPLAIN THE CULTURAL NEEDS OF EMPLOYEES, ISLAMIC AZAD UNIVERSITY Ali Siada PhD *Hasan Ali Bakhiar Nasrabadi PhD** Mohadeseh

More information

The effect of personalized nursing intervention based on CGA on the quality of life for elderly patients with chronic diseases.

The effect of personalized nursing intervention based on CGA on the quality of life for elderly patients with chronic diseases. Biomedical Research 2018; 29 (9): 1779-1783 ISSN 0970-938X www.biomedres.info The effec of ized nursing based on CGA on he qualiy of for elderly paiens wih chronic diseases. Fen Gong 1, Yan-ling Hu 2*

More information

Wanda Scott. Norquest College NFDN Assignment 2. Nursing Process / Care Plan. Cindy Hoyme

Wanda Scott. Norquest College NFDN Assignment 2. Nursing Process / Care Plan. Cindy Hoyme Running Head: ASSIGNMENT 2: NURSING PROCESS / CARE PLAN 1 Wanda Sco Norques College NFDN 1001 Assignmen 2 Nursing Process / Care Plan Cindy Hoyme Inroducion This is a Nursing Care Plan o pracice criical

More information

MILLENNIUM STRATEGIC PLAN A History of Forward Thinking

MILLENNIUM STRATEGIC PLAN A History of Forward Thinking MILLENNIUM STRATEGIC PLAN 2001-2010 A Hisory of Forward Thinking 4721/Millenium Plan 5/7/07 10:54 AM Page i BOARD OF TRUSTEES Armando M. Codina Chairman Besy S. Akins Paul L. Cejas Vice Chairman Rafael

More information

Department of Economics Working Paper 2016:14

Department of Economics Working Paper 2016:14 Deparmen of Economics Working Paper 2016:14 The maching process: Search or mismach? Nils Gofries and Karolina Sadin Deparmen of Economics Working paper 2016:14 Uppsala Universiy November 2016 P.O. Box

More information

Partnership With a Graduate Nursing Program: Voices of the Community

Partnership With a Graduate Nursing Program: Voices of the Community Universiy of Nebraska Omaha DigialCommons@UNO Disseraion and Thesis Barbara A. Holland Collecion for Service Learning and Communiy Engagemen (SLCE) 2003 Parnership Wih a Graduae Nursing Program: Voices

More information

NAVY COMMAND, CONTROL, AND COMMUNICATIONS - AN INTRODUCTION

NAVY COMMAND, CONTROL, AND COMMUNICATIONS - AN INTRODUCTION JOHN W. NEWLAND and H. GREGORY TORNATORE NAVY COMMAND, CONTROL, AND COMMUNCATONS - AN NTRODUCTON n his aricle we review he meaning of command and conrol in erms of a sysems-ype conrol process and he Navy's

More information

Influence of Stress and Nursing Leadership on Job Satisfaction of Pediatric Intensive Care Unit Nurses

Influence of Stress and Nursing Leadership on Job Satisfaction of Pediatric Intensive Care Unit Nurses Marquee Universiy e-publicaions@marquee College of Nursing Faculy Research and Publicaions Nursing, College of 9-1-2000 Influence of Sress and Nursing Leadership on Job Saisfacion of Pediaric Inensive

More information

FACADE IMPROVEMENTS INCENTIVE PROGRAM for EXISTING COMMERICAL BUILDINGS

FACADE IMPROVEMENTS INCENTIVE PROGRAM for EXISTING COMMERICAL BUILDINGS GAINESVILLE COMMUNITY REDEVELOPMENT AGENCY EASTSIDE COMMUNITY REDEVELOPMENT AREA FACADE IMPROVEMENTS INCENTIVE PROGRAM for EXISTING COMMERICAL BUILDINGS 802 NW 5h Avenue Suie 200 Gainesville, FL 32601

More information

1998 Wage and Vacancy Survey of Medical Laboratories Wages Up Slightly, Vacancies on Upswing

1998 Wage and Vacancy Survey of Medical Laboratories Wages Up Slightly, Vacancies on Upswing SPECIAL REPORT Barbara M. Casleberry, PhD, MT(ASCP) Laurie L Wargelin, MA 1998 Wage and Vacancy Survey of Medical Laboraories Wages Up Slighly, Vacancies on Upswing Dr Casleberry is vice presiden, ASCP

More information

An empirical analysis of the factors impacting discount rates: Evidence from the U.S. Marine Corps "i:t. Nayantara Hensel a,*, Martin Deichert b

An empirical analysis of the factors impacting discount rates: Evidence from the U.S. Marine Corps i:t. Nayantara Hensel a,*, Martin Deichert b ELSEVIER Review of Financial Economics 17 (2008) 62-78 RFE An empirical analysis of he facors impacing discoun raes: Evidence from he U.S. Marine Corps "i: Nayanara Hensel a,*, Marin Deicher b a Graduae

More information

Lonestar Dustoff Co. sets the standard for MEDEVAC crews in Iraq

Lonestar Dustoff Co. sets the standard for MEDEVAC crews in Iraq EAGLE February 2009 THE OFFICIAL MAGAZINE OF THE COMBAT AVIATION BRIGADE, 4TH INFANTRY DIVISION Iron Eagle Brigade parners wih IAF for missions over Baghdad skies Lonesar Dusoff Co. ses he sandard for

More information

NURSING. Caring for Populations: Transforming Undergraduate Education R U T G E R S S C H O O L O F

NURSING. Caring for Populations: Transforming Undergraduate Education R U T G E R S S C H O O L O F OFFICE OF THE DEAN School of Nursing Rugers, The Sae Universiy of New Jersey 180 Universiy Avenue Newark, NJ 07102 Nonprofi Organizaion U.S. Posage P A I D Permi No. 5287 Newark, NJ 07102 R U T G E R S

More information

Help Wanted in New Zealand: The ANZ Bank Job Advertisement Series

Help Wanted in New Zealand: The ANZ Bank Job Advertisement Series UNIVERSITY OF WAIKATO Hamilon New Zealand Help Waned in New Zealand: The ANZ Bank Job Adverisemen Series Brian Silversone Deparmen of Economics Working Paper in Economics 3/4 November 4 Paper Presened

More information

MarApr 2013 Vol 41, No. 2

MarApr 2013 Vol 41, No. 2 RACER FLYER Inel for he Family Securiy Forces prepares for Disurbance Conrol See DCT, Page 5 MARAPR 2013 L. Col. Kyle Noel Commander, 113 ASOS COMMANDER S COMMENTS Welcome from he 113h Air Suppor Operaions

More information

Engineers Camp Pendleton Day

Engineers Camp Pendleton Day Sociey of American Miliary Engineers Camp Pendleon Day CDR Glenn Shephard Assisan Operaions Officer, Naval Faciliies Engineering Command Souhwes June 25, 2009 251Jun 09 Naval Faciliies Engineering Command

More information

1. Call to Action / Tag Line: Join the nation thank a Vietnam veteran!

1. Call to Action / Tag Line: Join the nation thank a Vietnam veteran! Conens: 1. Call o Acion / Tag Lines 2. Key Messages 3. Imporan Supporing Facs 4. Background 5. Quoable Quoes (POTUS, SD, former POW) 6. Clarificaion of Terminology 7. TABs 1. Call o Acion / Tag Line: Join

More information

Standard Operating Procedures for Sterilization Services in Camps

Standard Operating Procedures for Sterilization Services in Camps Sandard Operaing Procedures for Serilizaion Services in Camps Family Planning Division Minisry of Healh and Family Welfare Governmen of India March 2008 Sandard Operaing Procedures for Serilizaion Services

More information

Adopted 2/15/18. Copy to be provided.

Adopted 2/15/18. Copy to be provided. Adoped 2/15/18. Copy o be provided. Projec Summary 1. Projec Tile: Blacksone & McKinley Transi Oriened Developmen 2. Projec descripion: briefly describe he projec seing, and is relaionship o exising and

More information

CORPORATE RISK REGISTER & ASSURANCE FRAMEWORK

CORPORATE RISK REGISTER & ASSURANCE FRAMEWORK CORPORATE RISK REGISTER & ASSURANCE FRAEWORK BRIEFING NOTE PREPARED FOR TRUST BOARD 7 SEPTEBER 2017 There are 26 risks on he as anagemen Team eeing on 10 h Augus 2017. Noe 771 is pending de-escalaion whils

More information

T H E U N I T E D S TAT E S O F A M E R I C A VIETNAM WAR COMMEMORATION

T H E U N I T E D S TAT E S O F A M E R I C A VIETNAM WAR COMMEMORATION T H E U N I T E D S TAT E S O F A M E R I C A VIETNAM WAR COMMEMORATION Conens: 1. Call o Acion / Tag Lines 2. Key Messages 3. Imporan Supporing Facs 4. Background 5. Quoable Quoes (POTUS, SD, former POW)

More information

errrs R-*tT+.1'.4 $f t-iaq Office of the ACC:iionai Director zf-d]-q c? ftiilr TADOmOilA! OIRECI $R *rcnmu*ril 'c') -{qir, { c{qeir

errrs R-*tT+.1'.4 $f t-iaq Office of the ACC:iionai Director zf-d]-q c? ftiilr TADOmOilA! OIRECI $R *rcnmu*ril 'c') -{qir, { c{qeir N - \ ENNAI \ The following Group'c'Poss are vacan in cghs chennai' lneresed candidaes may submi heir applicaion in he prescribed Proforma available inhewebsie"cghs chennain.nic,in"wihin onemonh ofdaeof

More information

MONTHLY STATISTICS OP REGISTERED UNEMPLOYMENT IN THEEUROPEAN COMMUNITY. January 1977

MONTHLY STATISTICS OP REGISTERED UNEMPLOYMENT IN THEEUROPEAN COMMUNITY. January 1977 DER EUROPÄISCHEN GEMEIS ESE GEMEENSCHAPPEN - D NAUTESEUROPEENNES - S' AFTEN - ISTITUTO STAT'ST. TATISTlSKE DEPARTEMENT F 'CAL OFFICE OF THE EUROPE UE COMUNITÀ EUROPEE - 8u. íuropaeiske FAELLESSKABI lunites

More information

COMMENCEMENT CEREMONY. Friday, May 18, :45 a.m. Bowman Gray Stadium

COMMENCEMENT CEREMONY. Friday, May 18, :45 a.m. Bowman Gray Stadium COMMENCEMENT CEREMONY Friday, May 18, 2018 8:45 a.m. Bowman Gray Sadium WINSTON-SALEM STATE UNIVERSITY COMMENCEMENT CEREMONY Friday, May 18, 2018 g 8:45 a.m. Bowman Gray Sadium g 1250 Souh Marin Luher

More information

Medical Benefits Member Guide

Medical Benefits Member Guide 20 17 Medical Benefis Member Guide SEPTEMBER 1, 2016 AUGUST 31, 2017 THIS GUIDE IS FOR: Acive employees Reirees who are no eligible for Medicare www.myuhc.com/hs www.healhselecofexas.com/cdhs (866) 336-9371

More information

course descriptions. COURSES G EGenEd Subjects and Prefixes INDEX

course descriptions. COURSES G EGenEd Subjects and Prefixes INDEX Course Descripions Course descripions appear on he following pages alphabeically by course prefix. The hree-leer prefix idenifies subjec area; he hree digis ha follow idenify he paricular course. Generally,

More information

UNCLASSIFIED Paper No. 2-3

UNCLASSIFIED Paper No. 2-3 Paper No. -3 An Opimal -Dv Guidance Law for Inerceping a Boosing Targe Absrac Lawrence C. Ng*, Eric Breifeller, and Arno G. Ledebuhr Lawrence Livermore Naional Laboraory P.O. Box 808, L-78, Livermore,

More information

EIT_Food. CLC West Networking Event. Maarten t Van Oort Director CLC West. Leuven I 25 January 2018

EIT_Food. CLC West Networking Event. Maarten t Van Oort Director CLC West. Leuven I 25 January 2018 EIT Food CLC Wes Neworking Even Maaren Van Oor Direcor CLC Wes Leuven I 25 January 2018 EIT_Food #foodinnovaion EIT Food CLC Wes Neworking Even Maaren Van Oor Direcor CLC Wes 25 January 2018 I Leuven Belgium

More information

2018 ANZICS NEW ZEALAND REGIONAL ASM 4 6 APRIL 2018 THE HILTON HOTEL, AUCKLAND, NZ HANDBOOK

2018 ANZICS NEW ZEALAND REGIONAL ASM 4 6 APRIL 2018 THE HILTON HOTEL, AUCKLAND, NZ  HANDBOOK ! 2018 ANZICS NEW ZEALAND REGIONAL ASM 4 6 APRIL 2018 THE HILTON HOTEL, AUCKLAND, NZ WWW.ANZICS2018.NZ HANDBOOK TABLE OF CONTENTS 3 Convenor's Welcome 4 Thanks To Our Sponsor 5 Auckland Map 6 Our World

More information

A ccess. n e w h a m p s h i r e. Guiding Questions. Living with Disability in the Granite State

A ccess. n e w h a m p s h i r e. Guiding Questions. Living with Disability in the Granite State A ccess vol. 1 issue 2 n e h a m p s h i r e M A Y 2 0 0 8 Living ith Disability in the Granite State policy brief: the changing dymanics of hospital care for mental illness & substance use in ne hampshire

More information

Patient Perceptions and Social Impact. Preliminary Results of the Bristol MRC Study

Patient Perceptions and Social Impact. Preliminary Results of the Bristol MRC Study Eye (1991) 5, 373-378 Patient Perceptions and Social Impact. Preliminary Results of the Bristol MRC Study K. J. LOE, D. A. GREGORY, R. I. JEFFERY, D. L. EASTY Bristol Summary One hundred and nine inpatients

More information

Headquarters Tennessee State Guard

Headquarters Tennessee State Guard Headquarters ennessee State Guard NSG AS OF 1 JULY 2012 HEADQUARERS - NSG MAJ JIM LEE & 1SG LARRY GILES Headquarters NSG PERSA NSG UNI MANNING OAL PERSONNEL Prior Non-Prior Male Female Officers 35 +1 24-1

More information

Oregon Trauma Systems Summary and Hospital Resource Criteria

Oregon Trauma Systems Summary and Hospital Resource Criteria Appendix D Oregon Trauma Systems Summary and Hospital Resource Criteria OREGON TRAUMA SYSTEM SUMMARY Emergency Medical Services Section State Health Division OCTOBER 1989 In 1985 the Oregon Legislature

More information

"Fair Shake' Plan THE ORIGINAL HOUSE P~ESENTS THE WAKE FOREST SPECIAL ON ANYTHING ON THE MENU

Fair Shake' Plan THE ORIGINAL HOUSE P~ESENTS THE WAKE FOREST SPECIAL ON ANYTHING ON THE MENU Commiee OLD GOLD AND BLACK Tuesday, March 4, 1969 PAGE THREE Seeking Answer To Problem Of Shrinking Lake By WLLAM MLLER STAFF WRTER The Reynolda Lake Commi. ee will consider in March he recommendaions

More information

Resource allocation in an outpatient department. Part one. M.M. Kats J.H. Quik

Resource allocation in an outpatient department. Part one. M.M. Kats J.H. Quik Resource allocation in an outpatient department Part one by M.M. Kats J.H. Quik Deventer Hospital Resource allocation in an outpatient department A system for allocating generic rooms for the outpatient

More information

How to deal with Emergency at the Operating Room

How to deal with Emergency at the Operating Room How to deal with Emergency at the Operating Room Research Paper Business Analytics Author: Freerk Alons Supervisor: Dr. R. Bekker VU University Amsterdam Faculty of Science Master Business Mathematics

More information

The Joint Training System

The Joint Training System READINESS AND AN AUOMAIO Joseph C. Barto, III Lieutenant Colonel, United States Army Joint Warfighting Center Fort Monroe, VA 23651 atrick D. Clark Commander, United States Navy Joint Warfighting Center

More information

Decision support system for the operating room rescheduling problem

Decision support system for the operating room rescheduling problem Health Care Manag Sci DOI 10.1007/s10729-012-9202-2 Decision support system for the operating room rescheduling problem J. Theresia van Essen Johann L. Hurink Woutske Hartholt Bernd J. van den Akker Received:

More information

DOES INTERNATIONAL OUTSOURCING REALLY LOWER WORKERS INCOME? ***

DOES INTERNATIONAL OUTSOURCING REALLY LOWER WORKERS INCOME? *** Forthcoming in Journal of abor Research DOES INTERNATIONA OUTSOURCING REAY OWER WORKERS INCOE? Jan König and Erkki Koskela December 00 Abstract: We analyze the impact of international outsourcing on income,

More information

Weiss v Bretton Woods Condominium II 2014 NY Slip Op 33905(U) October 2, 2014 Supreme Court, Nassau County Docket Number: Judge: Arthur M.

Weiss v Bretton Woods Condominium II 2014 NY Slip Op 33905(U) October 2, 2014 Supreme Court, Nassau County Docket Number: Judge: Arthur M. Weiss v Breon Woods Condominium II 2014 NY Slip Op 33905(U) Ocober 2, 2014 Supreme Cour, Nassau Couny Docke Number: 7878-12 Judge: Arhur M. Diamond Cases posed wih a "30000" idenifier, i.e., 2013 NY Slip

More information

At ND, greater percentage of accepted students enroll

At ND, greater percentage of accepted students enroll Welcome Class of '87 VOL. XVl, NO. l h: imkp:nlkn suden newspaper Sl'rYing nore Jamc and sain mary's Saurday, Augus 27, 1983 Freshman orienaion begins A ND, greaer percenage of acceped sudens enroll ByPATSAN

More information

How can the MST hospital reduce the variability in bed utilization at the nursing wards, while the OR capacity will be used in an efficient way?

How can the MST hospital reduce the variability in bed utilization at the nursing wards, while the OR capacity will be used in an efficient way? July, 2017 3 Management summary Health care costs are rising, the Dutch population is aging, and the government and health care insurers are cutting costs. These are only a few of the current developments

More information

Petra H. Vrieler BSc. May Master s Thesis. Industrial Engineering and Management. Health Care Technology and Management. University of Twente

Petra H. Vrieler BSc. May Master s Thesis. Industrial Engineering and Management. Health Care Technology and Management. University of Twente Petra H. Vrieler BSc May 29 Master s Thesis Industrial Engineering and Management Health Care Technology and Management University of Twente Enschede, the Netherlands Management on the Ambulatory Surgery

More information

Hospital Bed Occupancy Prediction

Hospital Bed Occupancy Prediction Vrije Universiteit Amsterdam Master Thesis Business Analytics Hospital Bed Occupancy Prediction Developing and Implementing a predictive analytics decision support tool to relate Operation Room usage to

More information

Pressure Ulcers and the Law. Cheryl Carter - Durban

Pressure Ulcers and the Law. Cheryl Carter - Durban Pressure Ulcers and the Law Cheryl Carter - Durban South African Context In 2009/2010, the Gauteng Department of Health and Social Development reportedly faced medical malpracice claims totalling R573

More information

Process optimization of the Preoperative Anesthesia Clinics at ZGT Almelo and Hengelo. R.C.A. Odijk

Process optimization of the Preoperative Anesthesia Clinics at ZGT Almelo and Hengelo. R.C.A. Odijk Process optimization of the Preoperative Anesthesia Clinics at ZGT Almelo and Hengelo R.C.A. Odijk 2012 OPTIMIZATION OF THE PRE ANESTHESIA CLINICS AT ZGT ALMELO AND HENGELO R.C.A. Odijk April 2012 Master

More information

THE USE OF SIMULATION TO DETERMINE MAXIMUM CAPACITY IN THE SURGICAL SUITE OPERATING ROOM. Sarah M. Ballard Michael E. Kuhl

THE USE OF SIMULATION TO DETERMINE MAXIMUM CAPACITY IN THE SURGICAL SUITE OPERATING ROOM. Sarah M. Ballard Michael E. Kuhl Proceedings of the 2006 Winter Simulation Conference L. F. Perrone, F. P. Wieland, J. Liu, B. G. Lawson, D. M. Nicol, and R. M. Fujimoto, eds. THE USE OF SIMULATION TO DETERMINE MAXIMUM CAPACITY IN THE

More information

MONTHLY STATISTICS OF REGISTERED UNEMPLOYMENT IN THE EUROPEAN COMMUNITY FEBRUARY 1977

MONTHLY STATISTICS OF REGISTERED UNEMPLOYMENT IN THE EUROPEAN COMMUNITY FEBRUARY 1977 A S E K D E R E U R O P E S E G E M E E N S C H A P P E N Q U E O E S C O M M U N A U R E S E U R O P E E N N E S C H E N G E M E N S C H A E N S U O S A S C O D Ü H A P P E N - D E S A S S K E D E P A

More information

The working papers of the Centre of Management Studies of IST (CEG-IST) are aimed at making known the results of research undertaken by its members.

The working papers of the Centre of Management Studies of IST (CEG-IST) are aimed at making known the results of research undertaken by its members. Os artigos de investigação do Centro de Estudos de Gestão do Instituto Superior Técnico (CEG-IST) destinam-se a divulgar os resultados da investigação realizada pelos seus membros. The orking papers of

More information

High tech, human touch:

High tech, human touch: High tech, human touch: Operations Research in the Operating Room and beyond Dr.ir. Erwin W. Hans Associate prof. Operations Management and Process Optimization in Healthcare dep. Operational Methods for

More information

A SURGICAL PROCEDURE TYPE SCHEDULE FOR GENERAL SURGERY

A SURGICAL PROCEDURE TYPE SCHEDULE FOR GENERAL SURGERY A SURGICAL PROCEDURE TYPE SCHEDULE FOR GENERAL SURGERY A robust tactical surgery scheduling approach to manage elective- and semi-urgent patient uncertainty Author: R.F. Jacobs Examination Committee: Prof.

More information

Chapter 1 INTRODUCTION TO THE ACS NSQIP PEDIATRIC. 1.1 Overview

Chapter 1 INTRODUCTION TO THE ACS NSQIP PEDIATRIC. 1.1 Overview Chapter 1 INTRODUCTION TO THE ACS NSQIP PEDIATRIC 1.1 Overview A highly visible and important issue facing the medical profession and the healthcare industry today is the quality of care provided to patients.

More information

INFECTION CONTROL PLAN- MAINTAINING COMPLIANCE WITH THE INFECTION CONTROL AND PREVENTION STANDARDS AND REGULATIONS: CMS CfC

INFECTION CONTROL PLAN- MAINTAINING COMPLIANCE WITH THE INFECTION CONTROL AND PREVENTION STANDARDS AND REGULATIONS: CMS CfC INFECTION CONTROL PLAN- MAINTAINING COMPLIANCE WITH THE INFECTION CONTROL AND PREVENTION STANDARDS AND REGULATIONS: CMS CfC 416.51 Lee Anne Blackwell, RN, BSN, EMBA, CNOR Vice President Clinical Services

More information

The Game Has Changed. Strategy For A Value Driven World. Steve Jenkins Senior Advisor. November 13, 2016

The Game Has Changed. Strategy For A Value Driven World. Steve Jenkins Senior Advisor. November 13, 2016 The Game Has Changed Strategy For A Value Driven World Steve Jenkins Senior Advisor November 13, 2016 Meet Sg2 Sg2, a Vizient company, is the health care industry s premier provider of market data and

More information

ADMINISTRATIVE APPROVALS -- (SPAD)

ADMINISTRATIVE APPROVALS -- (SPAD) ADMINISRAIVE APPROVALS -- (SPAD) * YOU NEED AN ADMINISRAIVE APPROVAL IF YOU INEND O: -Construct, expand, or locate a building or structure on a commercial, industrial, or residential multiple-family development

More information

Iil$IEl TI J ) NJCAA Champions. C,?fan',trt' _ - -..RCH 14, 1983

Iil$IEl TI J ) NJCAA Champions. C,?fan',trt' _ - -..RCH 14, 1983 il$el T J ) C,?fan',r' NJCAA Champions For he ninh ime, he Canon ATC ice hockey eiam has capured he NJCAA championship ile, and for Terry Marin i's he fifh ile in his coaching of he eam. Playing rn he

More information

University of Michigan Health System. Current State Analysis of the Main Adult Emergency Department

University of Michigan Health System. Current State Analysis of the Main Adult Emergency Department University of Michigan Health System Program and Operations Analysis Current State Analysis of the Main Adult Emergency Department Final Report To: Jeff Desmond MD, Clinical Operations Manager Emergency

More information

Nuclear Radiological Disaster Casualty Management (NRDCM): A Mul&- Na&onal Global Ini&a&ve

Nuclear Radiological Disaster Casualty Management (NRDCM): A Mul&- Na&onal Global Ini&a&ve Nuclear Radiological Disaster Casualty Management NRDCM): A Mul&- Na&onal Global Ini&a&ve Recent advances in NRDCM consequence management for nuclear disaster medicine assistance teams NDMATs) Na9onal

More information

Crowding at the Emergency Department of the Scheper Hospital Emmen

Crowding at the Emergency Department of the Scheper Hospital Emmen Crowding at the Emergency Department of the Scheper Hospital Emmen A system-wide imbalance analysis with staffing optimization and capacity planning to stay in the flow M.R. (Maurice) Darwinkel M.Sc. Thesis

More information

FFA Career Development Event: Telling the Beef Story Contest Toolkit

FFA Career Development Event: Telling the Beef Story Contest Toolkit The Beef Checkoff presents: FFA Career Development Event: Telling the Beef Story Contest Toolkit Resources and tools to engage in the growth and development of the beef ca le community. Revision date:

More information

Neurosurgery Clinic Analysis: Increasing Patient Throughput and Enhancing Patient Experience

Neurosurgery Clinic Analysis: Increasing Patient Throughput and Enhancing Patient Experience University of Michigan Health System Program and Operations Analysis Neurosurgery Clinic Analysis: Increasing Patient Throughput and Enhancing Patient Experience Final Report To: Stephen Napolitan, Assistant

More information

SOP no: A/003. To be revised in Date: February 2009

SOP no: A/003. To be revised in Date: February 2009 PGWC Standard Operating Procedure Category: Operating theatre activity data Recording and reporting of cancellation of elective theatre operations and/or procedures Compiled by: Theatre Efficiencies Task

More information

C-RAM: A Case Study in TRUE System-of-Systems Test and Evaluation

C-RAM: A Case Study in TRUE System-of-Systems Test and Evaluation NDIA 26 h Annual Naional T&E Confrnc San Digo CA C-RAM: A Cas Sudy in TRUE Sysm-of-Sysms Ts and Evaluaion Jim Bloodsworh PD C RAM Ts Branch Chif Jim.Bloodsworh@us.army.mil 256-774-6899 1 Nwork-Cnric Warfar

More information

CREATIVE PATIENT TRACERS. ObjecIves 6/16/17 THE FUNCTION OF EDUCATION IS TO TEACH. Discuss Use of PaIent Tracers. Explore Use of Tracer Tools

CREATIVE PATIENT TRACERS. ObjecIves 6/16/17 THE FUNCTION OF EDUCATION IS TO TEACH. Discuss Use of PaIent Tracers. Explore Use of Tracer Tools CREATIVE PATIENT TRACERS Mitzi Grey, BSN, RN- BC, Med mgrey@greyandco.com July 20, 2017 ObjecIves 1 Discuss Use of PaIent Tracers 2 Explore Use of Tracer Tools 3 Brainstorm ImplementaIon THE FUNCTION OF

More information

Training Workers with Disabilities Grant Program Announcement (GPA)

Training Workers with Disabilities Grant Program Announcement (GPA) BLUEPRINT FOR PROSPERITY JUNE 2014 Training Workers with Disabilities Grant Program Announcement (GPA) Grants of $5,000 to $100,000 Applications must be submitted no later than 11:59 p.m. on September

More information

Scheduling all patients within their desired access time by determining a reservation level for the consultation hours

Scheduling all patients within their desired access time by determining a reservation level for the consultation hours Scheduling all patients within their desired access time by determining a reservation level for the consultation hours I. Hof May 2017 Page 2 of 100 Scheduling all patients within their desired access

More information

Problem Statement. Focal Point: Widespread panic and confusion in the State of Hawaii, USA

Problem Statement. Focal Point: Widespread panic and confusion in the State of Hawaii, USA Root Cause Analysis Report RCA Example - Hawaii False Alarm (Revision 1) Problem Statement Report umber RCA Owner Chris Eckert Report Date 5/13/2018 RCA Facilitator Brian Hughes Focal Point: Widespread

More information

The incidence and nature of positive experiences in acute psychiatry: An exploratory study

The incidence and nature of positive experiences in acute psychiatry: An exploratory study The incidence and nature of positive experiences in acute psychiatry: An exploratory study Student W.A. Nugteren Student number 3850293 Status of paper Final version Date 21 August 2014 Institute Master

More information

Why Focus on Perioperative Services?

Why Focus on Perioperative Services? 1 Why Focus on Perioperative Services? 80% 60% 40% 20% 0% Perioperative Services are key to a hospital/system's success 68% % better performers revenue from perioperative services Perioperative Services

More information

nationals S n In Middle East:

nationals S n In Middle East: DUNIA BEAM PROJECT ERASMUS MUNDUS A SCHOLARSHIP SCHEME FOR EXCHANGE AND COOPERATION BETWEEN EUROPE AND MIDDLE EAST COUNTRIES JORDAN, LEBANON, PALESTIN AND SYRIA. S C H O L A RS H Ip s I ps For For s e

More information

University of Michigan Health System. Inpatient Cardiology Unit Analysis: Collect, Categorize and Quantify Delays for Procedures Final Report

University of Michigan Health System. Inpatient Cardiology Unit Analysis: Collect, Categorize and Quantify Delays for Procedures Final Report Project University of Michigan Health System Program and Operations Analysis Inpatient Cardiology Unit Analysis: Collect, Categorize and Quantify Delays for Procedures Final Report To: Dr. Robert Cody,

More information

Capacity Planning in the Healthcare Environment

Capacity Planning in the Healthcare Environment Capacity Planning in the Healthcare Environment At the Outpatient Clinic for Oncology of the University Medical Center Groningen Agaath Hermsen UMCG, Sector D, Oncology RUG, Faculty of Economic and Business,

More information

January What You Should Do. Background

January What You Should Do. Background in the news Nonprofit Organizations January 2015 IRS Issues Long Awaited Final Regula ons for Charitable Hospitals In this Issue: What You Should Do Background... 1 Final Regula ons Hospital Facili es

More information

Dordt Builds For The Future

Dordt Builds For The Future Dord Builds For The Fuure Noe: whowroe :c ORDT ::s::::::::n:o:e l::: 0) "';"';Ia\'''P,;;;(Q) Mavis Assink. his repor. is assis1anj: edior. a:::y will gree Dord sudens nex fall. The sudens will also noe

More information

Trust Board Meeting 02 March 2017

Trust Board Meeting 02 March 2017 Trust Board Meeting 02 March 2017 Title of the paper: Agenda item: 11/46 Guardian of Safe Working Quarterly Report (October 2016 January 2017) Lead Executive: Paul Da Gama, Director of Human Resources

More information

Welcome to Boston and Cambridge - AUA

Welcome to Boston and Cambridge - AUA Welcome o Boson and Cambridge - AUA 2013! for he 58h annual gahering of he Associaion of Universiy Archiecs. MIT and BC are very pleased o welcome you o our campuses and co-hos his year s conference. Our

More information

Quality& Liability Fall 2017 Midterm Scoring

Quality& Liability Fall 2017 Midterm Scoring Quality& Liability Fall 2017 Midterm Scoring The policies and procedures of a hospital provide: In the event the Medical Screening Examination does not reveal an Emergency Medical Condition: Patient

More information

COMMUNITY ENERGY STATE OF THE SECTOR

COMMUNITY ENERGY STATE OF THE SECTOR Published 2017 COMMUNITY ENERGY STATE OF THE SECTOR A study of community energy in England, Wales and Northern Ireland Infographic Booklet OVERVIEW 1 More than 222 Organisa ons With over 121 MW Energy

More information

WAITING ROOM SIGN IN I. CHIPPER

WAITING ROOM SIGN IN I. CHIPPER www. GoodByeFears.com I. CHIPPER WAITING ROOM SIGN IN On surgery day Chipper's first stop was the Outpatient Surgery waiting room. This is where Chipper signs in. When the nurse calls Chipper s name,

More information

Resilience Approach for Medical Residents

Resilience Approach for Medical Residents Resilience Approach for Medical Residents R.A. Bezemer and E.H. Bos TNO, P.O. Box 718, NL-2130 AS Hoofddorp, the Netherlands robert.bezemer@tno.nl Abstract. Medical residents are in a vulnerable position.

More information

University of Michigan Health System Analysis of Wait Times Through the Patient Preoperative Process. Final Report

University of Michigan Health System Analysis of Wait Times Through the Patient Preoperative Process. Final Report University of Michigan Health System Analysis of Wait Times Through the Patient Preoperative Process Final Report Submitted to: Ms. Angela Haley Ambulatory Care Manager, Department of Surgery 1540 E Medical

More information

The Ricann e Hadria n Initiativ e. for Communit y Organizing : a Project Report

The Ricann e Hadria n Initiativ e. for Communit y Organizing : a Project Report The Ricann e Hadria n Initiativ e for Communit y Organizing : a Project Report Lee Winkelman Boston, Massachusetts November 199 7 New Hampshire College Community Economic Development Advisor: Woullar d

More information

National Fire Academy Emmitsburg, MD. Sponsored in part through the generosity of Globe Manufacturing Company, LLC

National Fire Academy Emmitsburg, MD. Sponsored in part through the generosity of Globe Manufacturing Company, LLC National Fire Academy Emmitsburg, MD Sponsored in part through the generosity of Globe Manufacturing Company, LLC THREE STATE 2015 Course Offerings Course Name: Poli cs and the White Helmet Course Code:

More information

Summary of DNV Findings. Applies to All Physicians

Summary of DNV Findings. Applies to All Physicians Summary of DNV Findings DNV A few weeks ago, we had our annual visit from DNV to evaluate our compliance with Medicare Conditions of Participation. Amongst the non-conformities identified in that visit

More information

The Reason-for-Encounter mode of the ICPC: reliable, adequate, and feasible

The Reason-for-Encounter mode of the ICPC: reliable, adequate, and feasible Scand J Prim Health Care 1989; 7: 99-103 The Reason-for-Encounter mode of the ICPC: reliable, adequate, and feasible FRANS VAN DER HORST, JOB METSEMAKERS, FRANS VISSERS, GERHART SAENGER*, CEES DE GEUS

More information

Advance Directive: Understanding and honoring my future health care goals

Advance Directive: Understanding and honoring my future health care goals mycare Advance Directive: Understanding and honoring my future health care goals My Care, My Choices You might be healthy now, but what if you became very sick or injured in the future and couldn t speak

More information

THE SURGICAL CASE ASSIGNMENT AND SEQUENCING PROBLEM

THE SURGICAL CASE ASSIGNMENT AND SEQUENCING PROBLEM THE SURGICAL CASE ASSIGNMENT AND SEQUENCING PROBLEM A CASE STUDY Word count: 33.720 Anna Macken Student number : 01205262 Supervisor: Prof. dr. Broos Maenhout Master s Dissertation submitted to obtain

More information

M A Y F I E L D Creating Our Legacy

M A Y F I E L D Creating Our Legacy M A Y F I E L D Creating Our Legacy A Strategic Plan for Mayfield 2012 Historic Overvie Mayfield Kentucky is located in Graves County in the Jackson Purchase region of Kentucky. Established in 1821, the

More information

Care for Walk-in. Organizing a walk-in based Preoperative Assessment Clinic in University Medical Centre Utrecht

Care for Walk-in. Organizing a walk-in based Preoperative Assessment Clinic in University Medical Centre Utrecht Care for Walk-in Organizing a walk-in based Preoperative Assessment Clinic in University Medical Centre Utrecht Pieter Wolbers, MSc June 2009 A quantitative research into the preoparative process of University

More information

Asking Questions: Information Needs in a Surgical Intensive Care Unit

Asking Questions: Information Needs in a Surgical Intensive Care Unit Asking Questions: Information Needs in a Surgical Intensive Care Unit Madhu C. Reddy M.S. 1, Wanda Pratt Ph.D. 2, Paul Dourish Ph.D. 1, M. Michael Shabot M.D. 3 2 1 Information and Computer Science Department,

More information

Allocation of data mining resources: A system partitioning and hybrid control framework

Allocation of data mining resources: A system partitioning and hybrid control framework Allocation of data mining resources: A system partitioning and hybrid control framework John R. James 1 United States Military Academy Department of Electrical Engineering and Computer Science West Point,

More information

Review Process. Introduction. Reference materials. InterQual Procedures Criteria

Review Process. Introduction. Reference materials. InterQual Procedures Criteria InterQual Procedures Criteria Review Process Introduction As part of the InterQual Care Planning family of products, InterQual Procedures Criteria provide healthcare organizations with evidence-based clinical

More information