Implementation of malnutrition screening and assessment by dietitians: malnutrition exists in acute and rehabilitation settings

Transcription

1 Implemetatio of malutritio screeig ad assessmet by dietitias: malutritio exists i acute ad rehabilitatio settigs Eleaor Beck, Craig Patch, Mariaa Milosavljevic, Shellie Maso, Corie White, Mady Carrie ad Kelly Lambert Abstract The prevalece of malutritio withi hospital settigs is a major cocer to all health care workers. The recet developmet of a simple screeig tool for use i such settigs has icreased the opportuity to idetify at-risk patiets i a reasoable time frame durig their admissio. This paper outlies the implemetatio of a routie utritio screeig ad assessmet, performed completely by dietitias, across both acute ad rehabilitatio settigs. Dietitias were able to scree, o average, 72% of eligible patiets, which esured timely dietetic itervetio. The routie malutritio screeig ad assessmet process highlighted differeces (P < 0.01) i the rates of malutritio betwee the acute wards (rage 7 to 14%) ad rehabilitatio ward (49%). Sigificat differeces betwee acute ad rehabilitatio patiets were also foud withi the majority of idividual diagostic groups, icludig all surgery, fractures, cardiovascular icidets ad respiratory illess (P < 0.01). The idetificatio of rates of malutritio betwee differet wards, diagoses ad istitutioal settigs provides dietetic maagers with a sophisticated tool that ca assist i the allocatio of dietetic resources. This operatioal framework for routie screeig of utritioally at-risk patiets i hospital, eables dietitias to develop patiet outcomes ad a effective utritio care model. (Aust J Nutr Diet 2001;58:92 97) Key words: malutritio, utritio assessmet, screeig, outcomes, rehabilitatio Itroductio Malutritio is prevalet i all hospitals (1). Malourished patiets experiece slower healig, icreased rates of complicatio ad icreased mortality (2,3). They stay i hospital loger (4), cost more to treat ad have higher readmissio rates (5,6). Nutritio itervetio has bee show to improve these outcome measures (7). Nutritio itervetio ca iclude utritio cousellig, review of dietary itake, supplemetatio (with high protei, high eergy driks), use of eteral formulas ad pareteral utritio. The critical step is to idetify at-risk patiets early i their hospital stay to implemet appropriate treatmet. The patiets must the be followed throughout their stay, with further educatio ad treatmet after discharge as required. Dietitias i our hospitals have traditioally see oly patiets who are o a special diet (e.g. for diabetes) or those who are referred. I this case, patiets may ot be referred to a dietitia util poor utritioal status impedes their recovery from surgery or illess. Iroically, research has show that patiets are just as likely to be malourished whether o a special diet or ot (8). Ferguso et al. (9) oted the absece of iformatio o implemetatio of malutritio screeig tools as a flaw i their use ad developmet. This paper describes the implemetatio of routie malutritio screeig ad assessmet of patiets i a hospital group with 310 acute ad 50 rehabilitatio beds. The aim of the implemetatio was to esure timely idetificatio of patiets requirig dietetic itervetio. The methodology varies from the first published use of the tool (10,11) i that a dietitia coducted the iitial screeig. This paper also outlies the diagostic-related groups likely to suffer malutritio. Methods Tools for screeig ad assessmet The study was coducted i the Departmet of Nutritio ad Dietetics at Wollogog ad Port Kembla Hospitals i the Illawarra Regio of New South Wales. A tool recetly developed by Australia dietitias was used for malutritio screeig (10). This tool, kow as the FBBC malutritio screeig tool, cosists of two simple questios regardig recet weight loss without itet ad dietary itake due to poor appetite. Patiets who are at risk of malutritio are idetified. Oce the risk is idetified, a full utritioal assessmet must be udertake to idetify the level of malutritio, if ay. Subjective global assessmet (SGA) was chose as the assessmet tool for this study. SGA is a simple assessmet that icludes questios o weight maiteace or loss, dietary itake ad gastroitestial symptoms such as ausea (12,13). A simple physical examiatio to review subcutaeous fat ad muscle wastig is also icluded. These subjective criteria are categorised ad patiets are scored as: A, well ourished; B, moderately malourished or at risk of malutritio; or C, severely malourished. Seve dietitias i the utritio departmet atteded a oe-day traiig course o use of the FBBC tool ad SGA. The course was coducted by the authors of the FBBC screeig tool. Further traiig was the carried out withi the hospital. Dietitias had the opportuity to perform utritio screeig (usig FBBC) ad the assessmet (usig SGA) i pairs ad small groups to Illawarra Area Health Service, ad Departmet of Biomedical Sciece, Uiversity of Wollogog, Wollogog, New South Wales E. Beck, BSc(Hos), DipNutrDiet, Cliical Supervisor i Dietetics Departmet of Nutritio ad Dietetics, Wollogog ad Port Kembla Hospitals, Wollogog, New South Wales C. Patch, BAppSc, DipNutrDiet, Seior Dietitia M. Milosavljevic, BSc(Hos), DipNutrDiet, Dietitia i Charge S. Maso, BSc, MSc(N&D), Seior Dietitia C. White, BSc, MSc(N&D), Dietitia M. Carrie, BSc, MSc(N&D), Dietitia K. Lambert, BSc, MSc(N&D), Dietitia Correspodece: E. Beck, Departmet of Biomedical Sciece, Uiversity of Wollogog, Wollogog, NSW eleaor_beck@uow.edu.au 92 Australia Joural of Nutritio ad Dietetics (2001) 58:2

2 esure they were familiar with the tools used. Review i pairs esured that cliical assessmets were cosistet withi the group. Policies were developed ad the etire process was outlied i a flow chart to aid implemetatio. I-service educatio o the process of screeig ad assessmet was coducted for ward ursig staff, urse uit maagers ad medical iters. Seior dietetic staff performed a five-day audit usig the FBBC tool to determie umbers of patiets likely to require further assessmet usig SGA. This was used to measure the chage i workload that might result from routie utritio screeig ad assessmet. The hospitals medical admiistratio was supportive of our cotact with all patiets. This iitial work i itroducig the process ad collectig raw umbers of patiets did ot require approval of a ethics committee ad was cosidered part of usual dietetic practice. I additio, itervetios after diagosis of malutritio did ot vary from stadard hospital procedures. Study populatio Screeig was implemeted across the two hospital sites where the departmet of utritio ad dietetics provided services: a 310-bed acute care facility ad a 50-bed rehabilitatio uit. The childre s ward, pre- ad post-atal wards, the ocology ward ad the critical care uits were excluded. I the case of the critical care uit, most patiets could ot commuicate with a dietitia to aswer the questios ad utritio itervetio was already required i almost all patiets regardless of screeig. I the ocology wards, it was deemed iappropriate to scree termially ill patiets or those udergoig palliative treatmets. The tool is ot validated for use i childre. Patiets sufferig demetia or other coditios where commuicatio was ot possible were ot screeed. Implemetatio Each morig, patiets admitted o the previous day (or over the weeked if screeig o a Moday) were idetified via ward lists obtaied from the hospital patiet admiistratio system. A dietitia idetified ewly admitted patiets o special diets. This allowed for review of the admissio diet whe the patiet was screeed (rather tha two separate visits to the patiet). For screeig, it is ot ecessary to read the patiet s medical otes but it is, of course, ecessary prior to utritio itervetio. However, the medical otes were read before screeig if the patiet was already o a special diet, or if the diagoses ad age would make utritio itervetio seem likely (e.g. admissios for bowel surgery or fractured eck of femur). After screeig, medical otes were examied for patiets for whom SGA was required. All malourished patiets received utritio care throughout their admissio ad patiets were offered follow-up i outpatiet cliics or by home visit by a domiciliary care dietitia. Nutritio itervetio icluded meu selectio with utritio staff (rather tha geeral kitche staff), provisio of commercial ad or domestic utritio supplemets, review of dietary itake i hospital by utritio staff ad ivestigatio of the home situatio by a dietitia. Educatio about appropriate itake was provided i all situatios. Eteral feeds were supplied if this was deemed appropriate by the medical ad utritio care team. If required, discouted home supplemets were supplied through the area health service home eteral utritio assistace scheme. Aalysis of screeig The umber of admissios ad umber of patiets screeed were documeted o summary sheets. These sheets also were used to record whether or ot a SGA was performed ad the score received (A, B or C). The total umber of malourished patiets was collated o a mothly basis. The diagosis was recorded from the admissio otes to allow review of at-risk groups. Records from patiets i rehabilitatio wards ad acute wards were collated separately. Rates of malutritio were calculated by ward type ad diagosis. Aalysis of meas was used to examie differeces i proportios of malourished patiets betwee the differet wards with α-values less tha 0.05 defied as sigificat. (Aalysis of meas examies how far the rate of malutritio o a ward deviates from the mea rate of all patiets ad is more sesitive tha a χ 2 test for detectig extreme deviatios from the average) (14). Chi-square tests were used to aalyse differeces i the rates of malutritio betwee acute ad rehabilitatio settigs for each of the diagoses, with α-values less tha 0.05 defied as sigificat. Data were aalysed usig JMP statistical software (JMP, versio 3.04, 1999, SAS Istitute Ic, Cary, NC). Results Rates of screeig ad assessmet There were 7129 documeted ew admissios to wards where screeig was i place i the 14 moths betwee July 1998 (first data collectio) ad August Of these admissios, 5149 patiets were screeed. This represeted approximately 72% of all patiets eligible for screeig i this period. Complete data relatig to all wards were available for 11 moths (September 1998 to July 1999) ad relatig to diagosis types for eight moths (Jauary to August 1999). Reasos listed by dietitias o the summary forms for ot screeig patiets icluded: patiet discharged; patiet receivig palliative care or ot for resuscitatio; demetia; ad patiet previously screeed for this admissio. SGA ad idetificatio of malourished patiets Twety per cet ( = 1004) of the 5149 patiets screeed required further ivestigatio usig SGA. That is, they were classified as at risk of malutritio usig the FBBC screeig tool. Twelve per cet ( = 634) of all patiets screeed (or three-fifths of those show at risk) were idetified as malourished after ivestigatio usig SGA (Table 1). Levels of detectio of malutritio varied i differet wards. Results for eleve moths (the period for which full data relatig to wards were complete) are show i Table 2. The highest rates of malutritio were detected i rehabilitatio wards. Aalysis of meas showed sigificat differeces betwee wards, with rates i rehabilitatio ad ocology wards sigificatly higher tha the other wards (P < 0.01). Although screeig o Australia Joural of Nutritio ad Dietetics (2001) 58:2 93

3 Table 1. Summary of malutritio screeig usig FBBC tool (10) ad utritioal assessmet usig subjective global assessmet (SGA) (a) (12,13). Results show umbers of patiets for 14 moths of implemetatio from July 1998 to August 1999 Patiets admitted 7129 Total patiets screeed 5149 Patiets requirig SGA 1004 (20) Patiets with SGA score A 370 Patiets with SGA score B 394 Patiets with SGA score C 240 Patiets malourished, SGA score B or C 634 (12) (a) SGA: A, well ourished; B, moderately malourished or at risk of malutritio; C, severely malourished. ocology wards was ot icluded i our routie implemetatio, a sample of patiets screeed (durig a twomoth period ad excludig palliative patiets) showed these sigificatly higher rates. Iterestigly, rates i the surgical, orthopaedic ad coroary care wards were sigificatly lower (P < 0.01). I the acute care facility, rates of malutritio varied widely depedig o the diagosis as recorded by the dietitia. Although there was also variatio i rehabilitatio patiets depedig o diagosis, the rates of malutritio were uiformly high (with the exceptio of patiets admitted for back pai). Table 3 shows these data for eight moths from Jauary 1999 to August 1999 (the period for which data relatig to diagosis were complete). The rates of malutritio were statistically differet betwee the settigs i the followig groups of patiets: geeral surgery, fractures, cardiac, real, gastro-itestial surgery, elective orthopaedic procedures, fractured eck of femur, respiratory illess, cerebral vascular accidet, ad other medical (all P < 0.01) ad sigificat for diagosis of eurological disorders ad cofusio (both P <0.05). Discussio Dietitias ad the screeig process Malutritio is ofte igored ad idetificatio of these patiets ca be difficult due to the lack of a uiform measuremet techique (15). Dietitias at both our hospitals are ow aware of every ew patiet o their ward withi a maximum of 24 hours post-admissio (or 72 hours for patiets admitted o Friday afteroo or the weeked). Nutritio itervetio is therefore more timely. The screeig process has bee icorporated withi existig staff establishmet by a review of work practices. That is, treatmet of malourished patiets is give priority over routie review of patiets who have received previous dietary cousellig for chroic coditios. Other istitutios i Australia ad overseas have implemeted various screeig tools for malutritio, but it has ot always bee dietitias who have performed the screeig. Nursig, clerical ad utritio techical staff, have all bee used (9 11,16). We believe that dietitias are ideally positioed to perform this role. They have advaced skills i complete utritioal assessmet ad i oe visit ca perform screeig ad assessmet, ad the iitiate treatmet if required. They ca also perform routie review of pre-existig coditios requirig dietary moitorig (e.g. diabetes) all i a sigle cosultatio. I additio, if a patiet caot be screeed for a particular reaso (for example, if the patiet suffers from demetia), the dietitia has still idetified a at-risk patiet ad ca review him or her throughout the admissio. This decisio requires the cliical judgemet ad assessmet of a dietitia. Nightigale ad Reeves (17) showed dietitias to be more kowledgeable i assessmet ad maagemet of uderutritio whe compared to doctors, ursig staff ad medical studets. Screeig umbers Ideally all patiets should be screeed for malutritio. Our rate of 72% may be explaied by the difficulty i accessig patiets with a short legth of stay. Although patiets who were ot screeed have ot bee ivestigated i this study, it was assumed that patiets sufferig malutritio are likely to have loger legths of stay (5). A post-discharge study of people with a short legth of stay, may be appropriate to outlie their utritioal status clearly. Aother reaso patiets were ot screeed was exclusio of patiets iappropriate for screeig. For example, patiets with demetia were ot screeed (18) but were recorded iitially as eligible patiets as a dietitia eeded to see them to discer that they were ieligible to participate i screeig. However, the fact that dietitias did idetify these patiets Table 2. Ward area Rates of malutritio usig FBBC tool (10) ad subjective global assessmet (SGA) (a) (12,13) of utritioal status by ward type (September 1998 to July 1999) Total patiets screeed Patiets requirig SGA Total patiets with SGA score A Total patiets with SGA score B Total patiets with SGA score C Patiets (b) malourished Medical (19) (14) Real, medical (24) (13) Surgical (14) (10) Orthopaedic (14) (11) Coroary care (10) (7) Rehabilitatio (86) (49)** Ocology (c) (71) (45)** ** Sigificatly higher rates tha other wards usig aalysis of meas (α <0.01). (a) SGA: A, well ourished; B, moderately malourished or at risk of malutritio; C, severely malourished. (b) SGA score B or C. (c) Sample results oly from two moths of screeig. 94 Australia Joural of Nutritio ad Dietetics (2001) 58:2

4 through admissio lists is positive. Patiets with demetia could still receive appropriate utritio care i hospital ad have their home situatio examied. Numbers of malourished patiets The overall rate of malutritio foud i this study (12%) is similar to other rates reported i the literature. I Caada, Azad et al. (18) foud 15% of patiets malourished i a study of 152 patiets assessed withi 72 hours of admissio to a tertiary care hospital. Ferguso et al. (10) sampled 408 patiets i a Australia hospital ad foud a rate of malutritio of 15%, which is similar to that i the Caadia patiets ad that foud i this study. I cotrast, Covisky et al., usig SGA (2), foud higher rates of malutritio with approximately 40% of patiets moderately or severely malourished. However, this group s sample of 369 patiets were all at least 70 years of age (mea 81 years). Ufortuately, age was ot recorded i our study, but our populatio icluded all age groups, except paediatrics ad post-atal, ad is likely to be youger tha that of Covisky et al. Acute versus rehabilitatio settigs Although the overall rate of malutritio is ot high compared to some other studies, high rates o particular wards require further ivestigatio. Comparig the overall percetage of malutritio i acute ad rehabilitatio wards shows that, regardless of origial diagosis, patiets i rehabilitatio are far more likely to be malourished. This is a critical area of ivestigatio for subsequet work. Icreased dietetic services ad, perhaps, altered food service requiremets should be directed to this area. Dietetic staffig caot be based o bed umbers aloe. Specifically, although a dietitia may care for a smaller umber of hospital beds, the types of patiets may ecessitate much greater utritioal itervetio. Similarly, sigificatly lower rates of malutritio i certai wards (surgical, orthopaedic, ad coroary care) may demostrate that routie utritio screeig is of less beefit o these wards. Dietitias may have a substatial workload i other areas of utritio itervetio i these wards but utritio support for uderutritio will likely be a small part of their role. The high rate of malutritio o rehabilitatio wards also emphasises the eed to ivestigate the admissio status of these patiets. May rehabilitatio patiets trasfer from acute wards ad may or may ot have received utritioal support i the acute settig before trasfer. Other rehabilitatio patiets may have bee admitted from home. Aother reaso for skewed results may be that the admittig diagosis is ot the reaso the patiet requires rehabilitatio. Patiets are categorised by diagosis at admissio by clerical staff but this diagosis may chage Table 3. Number of patiets by diagoses with malutritio [usig subjective global assessmet, SGA (12,13)] i the acute settig ( = 2298) ad rehabilitatio settig ( = 316) from Jauary to August 1999 Acute Rehabilitatio Diagosis at admissio Geeral surgery (icludig woud maagemet) Fractures, excludig fractured eck of femur, vertebrae Patiets screeed () Patiets with SGA score B or C (a) Patiets screeed () Patiets with SGA score B or C (a) (5) 9 7 (78)** (4) (80)** Cardiac (4) 9 9 (100)** Real (14) 7 6 (86)** Gastroitestial surgery (20) (94)** Elective orthopaedic procedures, e.g. total (2) 10 3 (30)** kee replacemet Fractured eck of femur (12) (81)** Respiratory illess 75 9 (12) 5 5 (100)** Back pai, back surgery or crushed vertebrae 53 1 (2) 9 0 (0) Cacer or haematological disorder (48) (59) Cerebral vascular accidet 41 0 (0) (52)** Neurological, icludig surgery 25 2 (8) 6 4 (67)* Cofusio 8 2 (25) (80)* Weight loss for ivestigatio 7 5 (71) 2 2 (100) Above or below kee amputatios 6 5 (83) 14 9 (64) Other medical (b) (7) (34)** Ukow or ot recorded 73 8 (11) 15 0 (0) Head or brai ijury (c) (c) 8 4 (50) Total (9) (59) * P < 0.05 sigificatly differet from acute patiets. ** P < 0.01 sigificatly differet from acute patiets. (a) SGA score of B or C meas the patiet is malourished. (b) Other medical icludes: viral illess, arthritis, diabetes for stabilisatio, electrolyte imbalaces of ukow origi, acopia, ucoscious, uriary tract ifectios, psychiatric coditios icludig overdose, falls for ivestigatio ad pai maagemet. (c), ot recorded i acute settig. Australia Joural of Nutritio ad Dietetics (2001) 58:2 95

5 throughout admissio. For example, a failed femoral popliteal bypass may require loger admissio with treatmet for ifectio, evetual amputatio of limb ad subsequet rehabilitatio. The origial diagosis may ot be the most appropriate diagosis to idetify groups at risk of malutritio. Regardless of utritioal status at admissio, it is importat to track patiets with log legths of stay to idicate if there is a clear eed for utritio itervetio if complicatios arise throughout the admissio. Future studies will examie how this process may best be achieved. Diagostic groups Our fidigs regardig the type of diagosis for malourished patiets are cosistet with other studies (4,5). Chima et al. (5) characterised patiets accordig to broad diagostic groups based o diagosis o admissio to medical wards. Patiets with gastroitestial disease were sigificatly more likely to be malourished tha the geeral sample. Patiets with gastroitestial disease i the acute settig (regardless of requiremet for surgery) were more likely to be malourished i our study also. High umbers of malourished patiets i the respiratory illess ad cacer groups i our study also compare with the results of other studies (4). Improvig patiet outcomes Withi our istitutios, all patiets idetified as malourished o admissio are give appropriate cousellig ad care while i hospital, icludig provisio of utritio supplemets. A previous study has show the effectiveess of such utritio itervetio for i-patiets presetig with umerous coditios, both medical ad surgical (7). A primary goal i measurig outcomes of malutritio itervetios must be to show improvemet i the utritioal status of the patiet. A simple measuremet of this could ivolve repeatig the SGA after the itervetio. Patiets i hospital for log periods of time ca have repeat SGAs performed at desigated times ad, as metioed, all patiets with a log legth of stay who are ot malourished o admissio could have screeig after a give time. The challege will remai of how best to review patiets i the commuity. Decreasig legth of stay i a acute ipatiet settig meas poor utritioal status caot be addressed fully i a ipatiet settig ad pre- ad post-admissio services may be required (5). Withi our istitutio, uiformity of documetatio of utritio care, icludig documetatio of educatio provided, patiet goals ad outcomes egotiated, ad strategies to achieve these facilitates review ad follow-up by fellow dietitias. At discharge, the ward dietitia offers outpatiet followup to all malourished patiets. The domiciliary care dietitias have commeced approximately bimothly reviews of all malourished patiets who require home visits. However, there are logistical difficulties, which make 100% commuity follow-up impossible. These problems iclude dietitia umbers, patiet refusal ad cotact difficulties. Future work is required i this area to esure review of these patiets ad trackig of itervetios i relatio to utritioal status, readmissio rates ad other health outcomes. The primary beefit of routie malutritio screeig is the improvemet i idividual patiet health outcomes, but malutritio screeig ca also be fiacially justified (9) uder a diagostic-related group fudig system. Codig of malutritio as a co-morbidity documets the requiremet for utritio itervetio i patiet care ad hece dietetic positios ca be justified, maitaied ad, perhaps, ehaced. Coclusio At our hospitals, dietetic work practices ad hospital referral processes have bee challeged ad altered to esure timely dietetic itervetio i malourished patiets. Differeces i levels of malutritio betwee diagostic groups do exist ad idetificatio of the specific groups most likely to require dietetic itervetio meas that resources ca be diverted to areas of greatest eed. Further ivestigatio of differeces betwee various acute ad rehabilitatio wards is essetial. Fially, for future work, auditig of medical records of our patiet groups will allow ivestigatio of itervetios ad related outcomes. Ackowledgmets We would like to thak all the dietitias i the departmet of utritio ad dietetics. Their ability to embrace ew ideas ad provide critical feedback esured the success of the project. The support of the medical ad ursig staff at both Wollogog ad Port Kembla Hospitals is greatly appreciated. Refereces 1. McWhirter JP, Peigto CR. Icidece ad recogitio of malutritio i hospital. BMJ 1994;308: Covisky KE, Marti GE, Beyth RJ, Justice AC, Sehgal AR, Ladefeld CS. The relatioship betwee cliical assessmets of utritioal status ad adverse outcomes i older hospitalised medical patiets. J Am Geriatr Soc 1999;47: Naber TH, Schermer T, de Bree A, Nustelig K, Eggik L, Kruimel JW, et al. Prevalece of malutritio i osurgical hospitalised patiets ad its associatio with disease complicatios. Am J Cli Nutr 1997;66: Messer RL, Stephes N, Wheeler WE, Hawes MC. Effect of admissio utritioal status o legth of hospital stay. Gastro Nurs 1991;13: Chima CS, Barco K, Dewitt MLA, Maeda M, Tera JC, Mulle KD. Relatioship of utritioal status to legth of stay, hospital costs, ad discharge status of patiets hospitalised i the medicie service. J Am Diet Assoc 1997;97: Wedder DO, Schmeisser D, Barish M, Kamath SK. Ipatiet ad post-discharge course of the malourished patiet. J Am Diet Assoc 1991;91: Gallagher-Allred CR, Coble Voss A, Fi SC, McCamish MA. Malutritio ad cliical outcomes: the case for medical utritio therapy. J Am Diet Assoc 1996;96: Christese KS, Gstudter KM. Hospital-wide screeig improves basis for utritio itervetio. J Am Diet Assoc 1985;85: Ferguso M, Capra S, Bauer J, Baks M. Codig for malutritio ehaces reimbursemet uder casemix-based fudig. Aust J Nutr Diet 1997;54: Ferguso M, Bauer J, Baks M, Capra S. Malutritio screeig ad assessmet resource maual. Brisbae: FBBC Nutritio Research Group; Ferguso M, Capra S, Bauer J, Baks M. Developmet of a valid ad reliable malutritio screeig tool for adult acute hospital patiets. Nutritio 1999;15: Detsky AS, Baker JP, Medelso RA, Wolma SL, Wesso DE, Jeejeebhoy KN. Evaluatig the accuracy of utritioal assessmet 96 Australia Joural of Nutritio ad Dietetics (2001) 58:2

6 techiques applied to hospitalised patiets: methodology ad comparisos. JPEN 1984;8: Detsky AS, McLaughli JR, Baker JP, Johsto N, Whittaker S, Medelso RA, et al. What is subjective global assessmet of utritioal status? JPEN 1987;11: Rya TP. Statistical methods for quality improvemet. New York: Joh Wiley; p Baxter JP. Problems of utritioal assessmet i the acute settig. Proc Nutr Soc 1999;58: Brow DM. Process aalysis improves quality ad volume of utritio screeigs. J Am Diet Assoc 1996;96: Nightigdale JMD, Reeves J. Kowledge about the assessmet ad maagemet of uderutritio: a pilot questioaire i a UK teachig hospital. Cli Nutr 1999;18: Frazoi S, Frisoi GB, Boffelli S, Rozzii R, Trabucchi M. Good utritioal oral itake is associated with equal survival i demeted ad o-demeted very old patiets. J Am Geriatr Soc 1996;44: Azad N, Murphy J, Amos SS, Toppa J. Nutritio survey i a elderly populatio followig admissio to a tertiary care hospital. Ca Med Assoc J 1999;161: Australia Joural of Nutritio ad Dietetics (2001) 58:2 97