A PILOT STUDY ON DISTRICT HEALTH INFORMATION SOFTWARE 2: CHALLENGES AND LESSONS LEARNED IN A DEVELOPING COUNTRY: AN EXPERIENCE FROM ETHIOPIA Padikumar Thagasamy 1, Melaku Gebremichael 2, Metesot Kebede 2, Michael Sileshi 2, Noah Elias 2, Brook Tesfaye 2 1 Associate Professor, Departmet of Computer Sciece ad Iformatio Techology, College of Egieerig, Defese Uiversity, Federal Miistry of Natioal Defese, Debre Zeyit, Ethiopia 2 Policy ad Plaig Directorate, Federal Miistry of Health, Addis Ababa, Ethiopia ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - A properly fuctioig health iformatio system is cetral to achievig better health outcomes. Heceforward, strog health maagemet iformatio system is a backboe of strog health system. A properly fuctioig health maagemet iformatio system gets the right iformatio ito the right hads at the right time, eablig health data cosumers to make effective iformatio use. District health iformatio software 2 is ope source software for collectio, validatio, aalysis, ad presetatio of data tailored to maage itegrated health iformatio. This study examies lessos ad challeges of DHIS-2 to advace electroic health iformatio maagemet system i Ethiopia. A cross-sectioal pilot study o twety six health facilities foud i four purposefully selected regios of the coutry was coducted from February to November 2015. Data were collected usig iterview ad Likert scale questioaire. A cetral server was cofigured. The software was customized i view of the atioal health maagemet iformatio system. Traiig of traiees was give to users. The software was piloted o those twety six health facilities for three moths followig the traiig. It was foud that District Health Iformatio software cotributes i advacemet of existig health iformatio system primarily i terms of data utilizatio ad iformatio use. However, sice Ethiopia has a uique caledar, customizig reportig period to local caledar was a major challege. Key Words: District health Iformatio Software, DHIS-2, Pilot Study, Health Maagemet Iformatio System, Challege, Lessos Leared, Developig coutry, Ethiopia, Experiece 1. INTRODUCTION Utilizatio of precise, valid, timely ad credible data ad iformatio are the bases of decisio makig, policy ad plaig developmet [1, 2]. Strog health systems are cetral to achievig better health outcomes. Heceforward, strog Health Iformatio System (HIS) is a backboe of strog health system. A properly fuctioig HIS gets the right iformatio ito the right hads at the right time, eablig health data cosumers to make effective iformatio use [3]. Ethiopia adopted a oe pla, oe budget, ad oe report policy. I this regard, sice 2004 Ethiopia implemeted two types of electroic Health Maagemet Iformatio System (e-hmis) as itegral part of HIS providig fudametal iformatio for health system moitorig. Recet assessmet coducted by Ethiopia Federal Miistry of Health (FMOH) showed that despite itesive efforts to improve the efficiecy of e-hmis, data quality ad iformatio use had bee challeges for the past few years. I this circumstace, existig e-hmis could be icompetet to come across with high demad iformatio i order to make impressive gais ad pheomeo advacemets o major health outcomes i the ewly approachig Health Sector Trasformatio Pla (HSTP). District Health Iformatio Software (DHIS-2) is ope source software for collectio, validatio, aalysis, ad presetatio of data tailored to maage itegrated health iformatio. DHIS-2 has bee adopted for atioal ad program specific health iformatio system i aroud 47 coutries globally. It is a preferred health iformatio system especially as developig coutries try to trasform health care system by effective iformatio use as a critical process [4]. DHIS-2 i itself is a mirror that shows the picture of DHIS-2 to maagers ad policy makers, i district ad atioal level. Whatever iformatio trasparecy with assistace of proper ad efficiet iformatio techology, ad itegrity i its compoets, shows a better picture of the orgaizatio, orgaizatioal uits ad services delivery methods [2]. Cosequetly, this pilot study aimed to oversee likely opportuities ad challeges of DHIS-2 to advace existig e-hmis i Ethiopia. 2. METHODOLOGY 2016, IRJET Impact Factor value: 4.45 ISO 9001:2008 Certified Joural Page 1646
2.1 Pilot team compositio A atioal DHIS-2 core team comprisig differet skills ad backgrouds (Public health, Health Iformatio specialists, Epidemiologists, Iformatio ad Commuicatio Techology specialists, data maagers ad a project maager) was formed at the iitial stages of the project. Oe member of the team was appoited as the team leader. The purpose of the team was to coordiate the implemetatio of the DHIS-2 project by drivig the database desig process ad customizatio. Techical assistats from Health Iformatio Systems Program, Oslo Uiversity were borrowed to support the implemetatio. 2.2 Pilot Study Area, Desig ad Period Ethiopia, officially kow as the Federal Democratic Republic of Ethiopia, is a ladlocked coutry located i the Hor of Africa with over 91 millio ihabitats occupyig a total area of 1,100,000 km2 [5]. A crosssectioal study desig was coducted from February to November 2015. 2.3 Pilot Site Selectio Procedures Type of previously deployed e-hmis, socio-ecoomic status, geography, huma resource, iformatio techology ifrastructure, available historical data, ad admiistrative structure were used as pilot site selectio iclusio ad exclusio criterio. Accordigly, FMOH i cosultatio with Regioal health bureaus ad Zoal health departmets selected districts from the etire pool of available districts to clearly extract lessos ad challeges of DHIS-2 pilot implemetatio uder aforemetioed criterio. 2.4 Sample Size ad Samplig Techiques The pilot study was based o regios which were desiged to provide importat lessos for large scale implemetatio of DHIS-2 at atioal level. I geeral, based o iclusio criterio eleve regioal states (ie regioal ad two city admiistratio health bureaus) foud i Ethiopia were stratified ad a purposeful samplig method was deployed i order to select four regios. Accordigly, South Natios, Natioalities ad Peoples Regio (SNNPR), Oromia, Addis Ababa, ad Gambella regios were selected. Heceforth, followig pilot site selectio procedures Zoal health departmets foud i each of four selected regios carefully omiated their respective districts for DHIS-2 pilot implemetatio. All the 37 professioals who were workig i the selected districts (7 at Oromia, 12 at Addis Ababa, 8 at Gambella ad 10 at SNNPR) were icluded i the pilot study. Accordigly, seve day traiig of traiers was give o the objective of the pilot study ad o basic data etry, aalytics ad reportig procedures of the software for higher precisio ad accuracy. 2.5 Data Collectio Procedures, Processig ad Aalysis A five poit Likert scale questioaire was developed i order to assess attitudes, opiios, ad perceptios of users towards customized DHIS-2. Users choose from a rage of possible resposes to a specific questio; typically iclude Excellet,,,, ad Very poor. Data were also collected usig iterview of users. Variables primarily focus o six No-fuctioal requiremets of software egieerig. Data was umerically coded, trasformed, ad etered i to SPSS versio 20.0. Descriptive statistics was doe to describe the characteristics of users. Sequetial mixed method aalysis (quatitative data aalysis supplemeted by qualitative data aalysis) was coducted; as a result, fidigs were triagulated i order to fully explai ad make accoutable coclusios ad recommedatios from the pilot study. 3. RESULTS 3.1 Settig Up of a Cetral Server ad Customizatio of the DHIS-2 A cetral server was cofigured for the database. The advatage of usig oe cetral server i the coutry was because it was easier to provide techical support to oly oe cetral server istallatio. If stadaloe (offlie /olie) istallatios i each district, health facility, had bee chose, the oe eeded to provide techical support to all the hudreds of istallatios, whe ew versios eeded to be sychroized or whe somethig wet wrog. Due to poor Iteret coectivity ad iadequate capacity of the servers at the Miistry of Health headquarters, a cetral server usig cloud computig was set upped. The goal of cloud computig was cofigured to provide easy ad scalable access to well maaged computer servers ad other iformatio commuicatio techology services. DHIS-2 was customized to fit with Ethiopia health system based o a template of curretly fuctioig e- HMIS. All customizatio of the system was performed by the Ethiopia team uder the supervisio of Oslo Uiversity. The Ethiopia team demostrated great kowledge of iformatio systems ad the team members were quick learers. The customizatio 2016, IRJET Impact Factor value: 4.45 ISO 9001:2008 Certified Joural Page 1647
process wet smoothly demostratig the importace of kowledge trasfer ad capacity buildig. Service delivery, outpatiet ad ipatiet disease report types were desiged. All report types collect ad report data o mothly, quarterly ad yearly basis. A total of 122 idicators were icluded i the software. Age, sex, regio, ad other disaggregatio performace metrics were used for idicators based o type of program they moitor. Public Health Emergecy Maagemet (PHEM) module was also added to amplify the use of iformatio i protectig the atio from health ad health related hazards. PHEM module collects ad reports data o thirty three immediately reportable diseases ad seve weekly reportable diseases as part of public health emergecy preparedess ad maagemet. DHIS-2 was customized to eable data etry, exchage ad reportig to the ext higher level usig iteret. Durig weak or o iteret coectivity, data exchage is supported by exportig/ importig techiques usig simple secodary data storage devices such as flash disk ad memory sticks. Data flow was based o the health system hierarchy startig from health facilities to FMOH. Customized DHIS-2 was istalled o existig e-hmis computers foud i 26 health istitutios foud i four regioal health bureaus selected for the pilot study. Every HMIS officer at each health facility ad admiistrative structure has its ow privilege to access ad cosume data. 3.2 Users Satisfactio o Customized DHIS-2 Table 1, Table 2 ad Table 3 shows the overall mea for users satisfactio amog various key ofuctioal requiremets of software egieerig. Regardig iteroperability, 23(62.17) ad 28(75.68) of the users were agreed that DHIS-2 imports ad exports data i suitable formats. Although 21(56.76) of the users reported that DHIS-2 does t fuctio together with locally implemeted m-health. 20(54.05) of the users were agreed that DHIS-2 used low computig resources i terms of performace while 17(45.95) of them respoded that they are oly satisfied. Majority (70.27) of users were satisfied that DHIS-2 was easy to uderstad ad lear. 25(67.57) ad 16(43.24) of users were agreed that DHIS-2 has easy data etry ad aalysis respectively. 18(48.65) of users were satisfied that DHIS-2 had a feature to eter aual targets ad eligible for their specific district. 24(64.87) of users foud that customized DHIS-2 icluded all types of reports. Regardig security, 24(64.87) of users revealed that customized DHIS-2 icorporated strog autheticatio procedure i order to access data. Table-1: Users satisfactio o customized (Measure1) Measures Very Iteroperability Very Data Importig Data Exportig Supportability Performace 10 27.03 Very Low respose time Low computig resource Usability Easy Logi Very 1 2.70 Data etry ad Editig Data aggregatio ad compilatio Reportig Data aalysis Output Uderstadability Feedback mechaism Istall/ uistall Comprehesiveess Very Targets ad Eligible Icorporatig all report types Data trasfer Geographic Iformatio System Reliability Security Logi autheticatio Very 0 Very 0.00 Defie roles ad resposibilities of differet user accouts? Table-2: Users satisfactio o customized (Measure2) 2016, IRJET Impact Factor value: 4.45 ISO 9001:2008 Certified Joural Page 1648
Measures Iteroperability Data Importig 8 21.62 Data Exportig Supportability Performace 9 24.32 16 43.24 Low respose time 19 51.35 Low computig resource Usability Easy Logi 17 45.95 18 48.65 Data etry ad Editig 8 21.62 Data aggregatio ad compilatio 13 35.14 Reportig Data aalysis Output Uderstadability 11 29.73 21 56.76 20 54.05 26 70.27 Feedback mechaism 23 62.16 Measures Iteroperability Data Importig 6 16.22 3.62 Data Exportig Supportability Performace 7 18.92 3.95 2.16 Low respose time 2 5.41 3.38 Low computig resource Usability Easy Logi 4 10.81 3.65 4 10.81 3.27 Data etry ad Editig 6 16.22 3.76 Data aggregatio ad compilatio 5 13.51 3.41 Reportig Data aalysis Output Uderstadability 7 18.92 3.62 4 10.81 3.54 3 8.11 3.54 3 8.11 3.22 Feedback mechaism 2 5.41 3.43 Istall/ uistall 22 59.46 Istall/ uistall 3.14 Comprehesiveess Comprehesiveess Targets ad Eligible 18 48.65 Icorporatig all report types 13 35.14 Data trasfer 18 48.65 Geographic Iformatio System 15 40.54 Reliability Security 6 16.22 Logi autheticatio 13 35.14 Defie roles ad resposibilities of differet user accouts? 11 29.73 Targets ad Eligible 4 10.81 3.19 Icorporatig all report types 3 8.11 3.73 Data trasfer 5 13.51 3.65 Geographic Iformatio System 2 5.41 3.16 Reliability Security 11 29.73 4.14 Logi autheticatio 7 18.92 3.84 Defie roles ad resposibilities of differet user accouts? 5 13.51 3.84 Table-3: Users satisfactio o customized (Measure3) 4. DISCUSSION 2016, IRJET Impact Factor value: 4.45 ISO 9001:2008 Certified Joural Page 1649
Health iformatio system effective maagemet is oe of advaced ad valuable outcomes of health system maagemet. Therefore, oe of the greatest challeges for health system maagers is commitmet ad deep belief of implemetatio ad utilizatio of moder maagemet i health iformatio maagemet area [7, 8]. The result of the pilot study revealed importat challeges ad lessos of DHIS-2 i Ethiopia cotext. Accordig to fidigs, iteroperability, performace, usability, comprehesiveess, reliability, ad security criteria were met with eeds of users participated i the pilot study. Authors believe that HMIS officers at health facilities ad admiistrative levels should participate i proper traiig courses of iformatio use i decisio makig. I this cotext, similar study coducted i Ethiopia, Tazaia, Malawi, Botswaa ad Mogolia showed these challeges: - Customizatio of Ethiopia caledar - Health Iformatio experts did some other tasks; such as: service delivery to cliets [9] - High huma resource tur over [10] - Not eough computers for health data maagemet [10] - Iadequate access to DHIS skilled persoel [10] - Low health iformatio kowledge [11] - Istaces of DHIS-2 beig ot compatible with computers havig widows service pack 2 Operatig System - Challeges o Atiretroviral Treatmet data maagemet - Persoel resigatio durig DHIS activities, leads to improper DHIS implemetatio [11] - Not eough attetio paid to district iformatio techology protectio ad ifrastructure to esure DHIS software maiteace [12] - No appropriate policy for hardware, software, ad system developmet [12] The fidigs of our study also showed importat lessos that ifrastructure sigificace for proper implemetatio of DHIS, there is ot eough attetio paid to this subject i may other coutries. For example, there is ot primary ifrastructure for team workig, participatio i iformatio iterpretatio ad decisio makig. Furthermore, DHIS traiig ifrastructures is udesirable, first level data productio ad gatherig is iefficiet, there is o computerized report i urba ad rural health facilities, ad amely, there is ot eough iformatio for decisio makig i this level. The statistics tell us that the system is beig used. Geerally, due to the good attributes of the software, data maagers ad data etry staff readily welcomed it. They are describig it as beig user friedly ad very easy to use. The stadard reports ad the data visualizer are very popular tools for makig graphs ad reports. Users i districts ad health facilities also stated that they had easier access to their ow data, as well as data from the rest of the coutry from DHIS tha before. The argumet is that they ca ow access their data regardless of power-cuts (whe they used their laptop), viruses or computer problems, because the data was up there, always available, i the Iteret. Furthermore, they really appreciated the dyamic updates of data from aroud the coutry ad the messagig system i the DHIS, used for commuicatio betwee users ad the system support team to report bugs ad to get help whe havig problems. Just like Facebook as oe HMIS officer said. The pilot study revealed that user s participatio i all stages of DHIS desigig is very importat to comply with all users iformatio eeds [2, 13]. This study showed that there is user participatio i DHIS desigig stage. I 2005, oe study about DHIS i Keya showed that iformatio systems caot perform effectively uless users have eough kowledge of iformatio system desig, implemetatio ad applicatio ad to be empowered about iformatio culture, values ad its impact o effective health services delivery [14]. The preset pilot study also idetified that DHIS-2 provided simple aalysis ecouragig data use for decisio-makig. With ibuilt validatio process, data maagers are able to pick errors thus helpig i improvig data quality. A study coducted i Keya also idicated that DHIS has further [15]: - Improved the dissemiatio of public health iformatio ad facilitated public discourse ad dialogue aroud major public health threats - Stregtheed ability to moitor the icidece of public health threats ad timely respose - Improved the efficiecy of admiistrative systems i health care facilities. The preset pilot study also revealed that users were able to get istat access to their ow data, from aywhere ad at ay time. This greatly improved the sese of owership to the data ad the system ad thereby eabled importat learig-by-doig processes, sigificatly complemetig the formal traiig ad support. The study clearly showed that itegratig PHEM beefited the health system to capture public health emergecies before causig social disruptio ad ecoomic loss i additio to high degree of morbidity ad mortality. Oe user also added that Role of icorporatig PHEM i to DHIS-2 is very essetial ad priority of the health system that plays a key role i miimizig public health emergecy threats i Ethiopia. 2016, IRJET Impact Factor value: 4.45 ISO 9001:2008 Certified Joural Page 1650
5. CONCLUSIONS The orm whe implemetig coutry health iformatio system i Africa has bee to capture the data i stadaloe databases ad trasmit to the ext level through e-mail attachmets, electroically or physically o a memory sticks. Sigificat huma capacity o databases, data maagemet ad system support is eeded i order to maage a atioal health iformatio system based o umerous stadaloe database applicatios with fragile flows of data betwee them. Problems of data reportig, completeess ad the maiteace of umerous stadaloe applicatios across the coutry make it very complicated. Ethiopia made a bold move to adopt a web based system for data maagemet. The ability to implemet DHIS-2 olie o oe cetral server made it sigificatly easier to roll out the system coutrywide, as compared with the alterative of havig to support ad maitai separate istallatios i all districts, hospitals ad health facilities aroud the coutry. Pilot of DHIS-2 idicates that there are may areas beig improved. Accordig to our fidigs the majority of DHIS criteria compliace with HMIS officers is desirable. I geeral, the pilot study cotributes i advacemet of the existig HMIS, especially i terms of data utilizatio ad iformatio use. It is however importat that capacity buildig, relevat ifrastructure ad staffig levels be evaluated for the system before rollig out at atioal level. Also, we suggest that FMOH ad regioal health bureaus must work together before system implemetatio to icrease its adoptio. ACKNOWLEDGMENT We would like to ackowledge Health Iformatio System Program at Uiversity of Oslo, Norway for their ureserved techical guidace. We are also grateful to all peoples who have participated i the pilot study. REFERENCES [1] Heywood A, Rohde J. Usig iformatio for actio: a maual for health workers at facility level. South Africa: EQUITY Project. 2002. [2] Lippeveld T, Sauerbor R, Bodart C. Desig ad implemetatio of health iformatio systems: World Health Orgaizatio; 2000. [3]Orgaizatio WH. Everybody's busiess--stregtheig health systems to improve health outcomes: WHO's framework for actio. 2007. [4] Available from: www.dhis2.org. [5] Zewde B. A history of moder Ethiopia, 1855-1991: James Currey Publishers; 2001. [6] Gilvary A. New cocepts i iformatio scieces. Chapar Press; 2008. [7] Sarafi Zadeh A, Ali Paahi A. Maagemet iformatio systems. Mir publicatio i Persia Tehra. 2002. [8] Twaakyodo H, editor. DHIS Customizatio Tazaia experiece, 2003. Available o web site: http://www hispkerala org/coferece/presetatios/coutry 20presetatiio. [9] Muyepa A, editor. HIMS i Malawi: Achievemets, Challeges ad the Way Forward. Proceedig of the First Iteratioal DHIS Workshop; 2006. [10] Seid H. HISP Ethiopia: Curret Status ad Future Directios. [11] Chada O, Shogwe J. Evaluatio of DHIS pilot project i Botswaa. Gaboroe, Botswaa: Miistry of Health. 2006. [12] Barro P, Hamer L, Mahomed H, Braa J. District Health Iformatio System Guidelies. Europia Uio, Departmet of Health. 1998; 18. [13] Odhiambo-Otieo GW. Evaluatio criteria for district health maagemet iformatio systems: lessos from the Miistry of Health, Keya. Iteratioal joural of medical iformatics. 2005;74(1):31-8. [14] Maya A, Braa J, Øverlad LH, Titlestad OH, Mumo J, Nzioka C, editors. Natioal roll out of District Health Iformatio Software (DHIS 2) i Keya, 2011 Cetral server ad Cloud based ifrastructure. IST-Africa 2012 Coferece Proceedigs; 2012. BIOGRAPHIES Mr. Brook Tesfaye Tekle (BSc Computer Sciece Descriptio ad Iformatio Techology, MPH i Health Iformatics Specialty) is workig as a Health Moitorig ad Evaluatio Expert at Ethiopia Federal Miistry of Health. He has research iterests o Health Moitorig ad Evaluatio, Health Iformatio Systems, Electroic Medical Record, Mobile Health, Data Miig, Machie Learig, Artificial Itelligece, Big Data Aalytics, Health Data Maagemet, Health Iformatio Retrieval, Health Data Structure ad Algorithm, Tele-medicie. 2016, IRJET Impact Factor value: 4.45 ISO 9001:2008 Certified Joural Page 1651