- Email: [email protected]
Poor performance of mandatory nutritional screening of in-hospital patients
Clinical Nutrition 31 (2012) 862e867
Contents lists available at SciVerse ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original article
Poor performance of mandatory nutritional screening of in-hospital patients Nina Rica Wium Geiker a, Sisse Marie Hørup Larsen a, Steen Stender b, Arne Astrup a, c, * a
Department of Clinical Nutrition, Copenhagen University Hospital Gentofte, DK-2900 Hellerup, Denmark Department of Clinical Biochemistry Copenhagen University Hospital Gentofte, DK-2900 Hellerup, Denmark c Department of Human Nutrition, Faculty of Life Sciences, University of Copenhagen, DK-1958 Frederiksberg C, Denmark b
a r t i c l e i n f o
s u m m a r y
Article history: Received 18 January 2012 Accepted 23 March 2012
Background & aims: Since 2006 it has been mandatory at Copenhagen University Hospital Gentofte to screen all patients for nutritional risk within 24 h of admittance. Audits conducted by department staff estimate that 70e80% of assessments are correctly executed, but the validity of this estimate is unknown. The aim of the present study was to discover the true proportion of hospitalized patients receiving nutritional risk screening within the stipulated time limit and to evaluate the validity of the screening by comparison with medical records. Methods: Retrospective examination of medical records of all patients (N ¼ 3278) hospitalized in September 2008 in 11 different medical specialities were analysed in 2009e2010. Results: Of 2393 medical records 24% of the patients were screened, of these only 65% were screened within the stipulated time limit. Half of the conducted screenings were inaccurate, the most common error being underestimation of nutritional status. Forty-six percent of patients required a secondary nutritional risk screening and 30% were found to be nutritionally at risk. Conclusion: Only 8% of patients received the mandatory nutritional risk screening without procedural errors. We conclude that pre-scheduled, self-conducted audits are not viable as the basis of an assessment of the use of nutritional risk screening. Ó 2012 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Nutrition Malnutrition Patients Audit Screening
1. Introduction Several international studies have revealed a prevalence of 20e60% of moderate to severe malnutrition in hospitalized medical and surgical patients.3,5,9,11,12,14,16,19,20 A large proportion of these patients were malnourished on admittance and malnourishment increased in most of these patients during hospitalization.10 Malnourished patients are found to have prolonged convalescence and admittance, to require more medication, to be susceptible to additional infections and more severe disease, and to have a higher mortality and cost.2,3,5,11,21 There is solid scientific documentation that unintentional weight loss in obese, normal and underweight patients increases all cause mortality.1,5,6,22 In order to alleviate this problem national and international organisations
Abbreviations: A-score, Degree of malnutrition; B-score, Grade of disease; MNA, Mini-nutritional Assessment; NRS 2002, Nutrition Risk Screening; MUST, Malnutrition Universal Screening Tool; GS, Groent System (Electronic patient handling system); OPUS, Electronic patient handling system. * Corresponding author. Department of Human Nutrition, Faculty of Life Sciences, University of Copenhagen, Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark. Tel.: þ45 3533 2476. E-mail address: [email protected] (A. Astrup).
recommend routine nutritional risk screening of hospitalized patients to identify those likely to benefit from nutritional intervention (ASPEN (1995), British Dietetic Association (1999), Department of Health (2001), BAPEN (2000), Council of Europe (2002) and ESPEN.10,8 Nutritional risk screening of hospitalized patients has thus become mandatory in some countries, e.g. United Kingdom and Denmark,4 NHS Quality Improvement Scotland 2003, National Institute for Health and Clinical Excellence 2006). Several tools have been developed to assess and register nutritional status uniformly, and to identify malnutrition in clinical practice, including Mini-Nutritional Assessment (MNA), Nutrition Risk Screening (NRS 2002) and Malnutrition Universal Screening Tool (MUST).10 To respond to and improve patients’ nutritional status, reliable assessment and documentation of nutrition status upon admittance is needed. In Denmark guidelines have been developed and the health authorities have declared malnutrition in hospitals an area of priority.4 The Copenhagen Health Authority has attempted to evaluate the implementation and use of NRS 2002 by conducting half-yearly, pre-scheduled audits, based on hospital staff’s self-reporting of a self-selected sample of 20 patients per department. The results of these audits have indicated that 70e80% of patients are screened for nutritional risk (Audit from Capital Region: 8500 beds in 10 hospitals, unpublished). The aim of the
0261-5614/$ e see front matter Ó 2012 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. doi:10.1016/j.clnu.2012.03.006
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
present study was to discover the true proportion of hospitalized patients receiving nutritional risk screening within the stipulated time limit and to evaluate the validity of the screening by comparison with medical records.
2. Materials and methods A retrospective method was used to study medical records of all patients hospitalized in the period 1ste30th September 2008 at Copenhagen University Hospital Gentofte. Since 2006 it has been mandatory in the capital region of Denmark to screen all hospitalized patients for nutritional risk within 24 h of admittance, followed by repeated screening performed once each week thereafter if patients still are admitted. All nutritional risk screenings are to be performed by hospital health staff. All medical records were retrieved manually and examined by the first author (NG) from February to October 2009. Hospitalized patients were defined as being admitted to a ward and occupying a prescribed bed, and as such they were registered in two different electronic patient handling systems (GS and OPUS). Patient information such as social registration number, diagnosis, day of admittance and discharge, patient type (acute/elective), as well as medical ward and section at the Copenhagen University Hospital Gentofte, were collected from GS. The medical records were thoroughly examined for the presence of the NRS 2002, day of primary nutritional risk screening, and the results of the nutritional risk screening. Data registered in NRS 2002 was controlled for coherence with information on nutritional status otherwise registered in the medical record. The length of stay was recorded and calculated. The primary nutritional risk screening consisted of each patient’s weight, height, calculated BMI, reported dietary intake in the past week, unintentional weight loss over the past three months, and the severity of the illness that gave rise to admittance to the hospital (B-score, Table 1). A secondary nutritional screening is required if the patient’s BMI < 20.5 kg/m2, if the patient’s dietary intake has decreased in the past week, if the patient has lost more than five percent body weight in the last three months, or if the illness is severe (B-score ¼ 3). The degree of malnutrition (A-score) is evaluated and given a score of 0e3, an additional point being added if the patient is 70 years old or more. Nutritional therapy is initiated if the sum of the A-score, B-score and the age related score is equal to three or more (Table 1). The nutritional risk screening was classified as insufficiently performed if one or more of the following criteria were present: - Discrepancy between information found in the NRS 2002 and information elsewhere in the medical records, e.g. two different body weights recorded on the same day. - Failure to register a recent unintentional weight loss.
863
- Incorrect A-score, e.g. a calculated BMI of 17 kg/m2 and appetite reduced by 50% with A-score noted as 1, correct A-score being 3. - Lack of repeat screening in accordance with time of hospitalization. - Lack of secondary nutritional risk screening. - Incomplete nutritional risk screening All statistical analyses were performed using Microsoft Excel and SPSS version 15.0 for Windows. All data was non-parametric, and differences between groups were analysed by Wilcoxon Signed Ranks, Sign test and ANOVA. Data is presented as median (range). 3. Subjects Of 3287 medical records for patients admitted in the period 1ste30th. September 2008 a total of 2393 (72%) were found in the patient record archives, the remaining records probably still being on the various clinical wards. All patients had been admitted to Copenhagen University Hospital Gentofte in the period 1ste30th. September 2008. The median length of time of hospitalization was two days (1e128), with a variation between the various medical specialities. Table 2 shows the distribution of patients in the medical specialities and the length of hospitalization. Neither the Department of Gynaecology and Obstetrics nor the Department of Paediatrics performed nutritional risk screenings during the investigated period in the present study; this was due to lack of nutritional risk screening tools suitable for these two patient groups. 4. Results Seventy-six percent (N ¼ 1819) of the 2393 patients were not screened for nutritional risk at all during their hospitalization (Fig. 1). A total of 24% (N ¼ 574) were screened and 35% of these were screened later than 24 h after admittance. Thirty percent of patients screened for nutritional risk were at risk of malnutrition, and thus required nutritional therapy. Nutritional status was correctly assessed and recorded within the stipulated time limit for a total of 8% (N ¼ 194) of patients hospitalized in September 2008. Forty-six percent of the patients administered the primary nutritional risk screening required a secondary nutritional risk screening (Fig. 2). However, a secondary nutritional risk screening was conducted in only 14% of these cases. Forty-eight percent (28e63%) of all the conducted nutritional risk screenings were inadequate, or had A- or B-score in disagreement with data registered in other places within the medical record. In example there was noted a 10% weight loss the past month in the medical record but patients was only given a score of one in B-score where a score of three would have been correct
Table 1 Score for secondary nutritional screening. A Under nutrition
B Severity of illness in regards to nutritional needs (i.e.)
Score ¼ 0 None Score ¼ 1 Light
Normal nutritional status >5% weight loss in 3 months or intake 50e75% of dietary needs past week. BMI not an issue her
Score ¼ 2 Moderate
>5% weight loss in 2 months or intake 25e50% of dietary needs past week or BMI 18.5e20.5 þ affected general condition >5% weight loss in 1 month or intake 0e25% of dietary needs past week or BMI less than 18.5 þ affected general condition
Normal needs - Collum femoris fracture - Chronic patients with exacerbations: liver cirrhosis, COPD - Chronic dialysis - Diabetes - Cancer - Large abdominal surgery (colectomy, gastrectomy, hepatectomy) - Post operational ATN - Apoplexia - Severe pneumonia - Cranial trauma - More than 50% burn - Severe infections (sepsis) - Patients in intensive therapy with multiple organ failure
Score ¼ 3 Severe
864
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
Table 2 Number of retrieved medical records and length of admittance, presented for each medical specialitya. Medical speciality
Patients, N
Retrieved, N (%)
Admittance, days
Vascular surgery Geriatrics Nose and throat Internal medicine Gynaecology-obstetrics Dermatology Paediatrics Cardiology Thoracic surgery Pulmonary medicine Planned surgery TOTAL
108 287 275 565 298 42 392 702 132 137 337 3278
31 218 149 416 258 13 294 572 109 84 249 2393
3 8 3 2 2 16 1 7 2 4 2 2
(1e31) (1e56) (1e35) (1e49) (1e10) (4e29) (1e50) (1e28) (1e26) (1e128) (1e50) (1e128)
Presented as median (range).
(Table 1). This discrepancy was found in both primary and secondary nutritional risk screenings (Fig. 3). An incorrectly low Ascore (nutritional status) or lacks of repeat screening were the most frequent errors. Thirty five percent of the screenings were conducted after 24 h following admittance (Fig. 4). Eight percent of nutritional screenings were not performed until the fifth day of admittance or later. Sixty-seven percent (N ¼ 1615) of the 2393 admitted patients were acute, of those were 26% (N ¼ 417) screened for nutritional risk, were as 19% (N ¼ 148) of 778 elective patients were screened for nutritional risk during admittance (P ¼ 0.000). There was no difference in accuracy of nutritional screening among acute and elective patients. Respectively, 59% and 83% (P ¼ 0.000) of the screenings for nutritional risk of acute and elective patients were performed on the first day of admittance. Acute patients were admitted for a longer period than elective (4.6 vs. 3.8 days, P ¼ 0.00). Among female and male patients; 52% and 43% respectively were incorrectly screened for nutritional risk (P ¼ 0.031) and 72%
5. Discussion Nutritional risk screening of all hospitalized patients has been mandatory at Copenhagen University Hospital Gentofte since 2006. The present review of medical records finds that only 8% of 2393 patients hospitalized in September 2008 were appropriately screened using the recommended NRS 2002 within 24 h of admittance. In the present study almost 40% of patients admitted longer than two days were screened for nutritional risk, whereas this was the fact in only 5% (N ¼ 28) of patients with one day of hospitalization. There was a tendency that patients with a stay longer than two days were screened for nutritional risk on their second day of admission or later. A longer hospitalization could results in more opportunities for the staff to perform the registration. In the present study acute patients were admitted for a longer period and more were screened for nutritional risk than elective patients; this correlates with longer hospitalization resulted in higher percentage of screened patients. Though only 19% of elective patients were screened during their hospitalization in the present study 80% of those were screened within the stipulated time limit. Reasons for lack of nutritional risk screening have been investigated in previous studies.9,17,18 The reasons most often given for failure to perform nutritional risk screenings are lack of instruction on how to use the nutritional risk screening form, too much paperwork/lack of time, lack of knowledge of nutritional risk
% 100 90 80 70 60 50 40 30 20 10
Va sc ul ar s
L
ur ge ry G er ia No tr i se cs an d In th te ro G rn at yn al ae m co ed lo ic gy in e -o bs te tri cs De rm at ol og y Pa ed Th ia tr i or cs ac ic su rg er y Ca Pu rd lm io on lo ar gy y m e Pl di cin an ne e d su rg er y
0 TO TA
a
(29) (76) (54) (74) (87) (31) (75) (81) (83) (61) (74) (73)
and 62% respectively were acutely admitted. There was no difference between gender regarding percentage of nutritionally screened or day of performed screening (P ¼ 0.319 and P ¼ 0.249) neither was there difference between fraction of acute female and male patients being screened for nutritional risk. Five percent (N ¼ 28) of patients hospitalized for one day (N ¼ 607) and 38% of patients hospitalized more than two days (N ¼ 1101) were screened for nutritional risk, however 50% being inaccurate.
Fig. 1. Percentage of patients who were not screened for nutritional risk during their hospitalization. Data is presented for each medical speciality.
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
865
%
100 90 80 70 60 50 40 30 20 10 L TO TA
G er ia tr i cs He ad an d ne In ck te rn al m ed ic in e De rm at ol og y Ca rd io lo Th gy or ac ic su Pu rg lm er on y ar y m ed ic Pl in e an ne d su rg er y
Va sc ul ar s
ur ge ry
0
Fig. 2. Percentage of patients identified by primary nutritional risk screening results, as being in need of secondary nutritional screening, but did not receive this.
screening, or simply forgetting it.18,9 Of 857 hospital staff taking part in one of these studies 40% found it difficult to identify patients at nutritional risk despite the fact that more than 70% had used nutritional therapy as treatment. Seventy six percent of staff considered it necessary to screen patients for nutritional risk, but only 23% routinely performed nutritional risk screening when admitting a patient.18 A questionnaire similar to the one used by Rasmussen et al.18 was repeated in 2004, finding that 40% of staff now performed nutritional risk screening as a routine assessment.13 The studies by Rasmussen et al.18 and Lindorff-Larsen et al.13 do not necessarily report responses from the same staff members as the population was random, which might influence the results and make them difficult to compare. Furthermore, both studies are based on self-reported data and thus may not reflect the actual number of patients being screened. Even so, the improvement seen from 1997 to 2004 may reflect a true increase in
nutritional risk screening, though the results of both the current audit and of Rasmussen et al.19 make this questionable. Nearly 50% of the registered nutritional risk screenings in the present review of medical records were inadequate or were inconsistent with medical records. Ten percent more female than male patients had inaccurate screenings of nutritional risk performed; and 10% more female patients were acutely admitted which could result in more stressful admittance with too little time to register nutritional status accurate. But because of no difference in accuracy between acute and elective patients being screened, or between fraction of female and male patients or acute female and male patients; this seems unlikely to be the explanation. There does not seem to be a rational explanation to why mandatory nutritional risk screenings were registered in less female than male patients during admission. The most frequent mistakes were a misleadingly low A-score (evaluation of nutritional status, Table 1) and lack of
% 70 60 50 40 30 20 10
Fig. 3. Percentage of patients with inadequate or incorrect nutritional risk screening.
TO TA L
G er ia tr i cs H ea d an d ne In ck te rn al m ed ic in e De rm at ol og y Ca rd io lo Th gy or ac ic su Pu rg lm er on y ar y m ed ic Pl in e an ne d su rg er y
Va sc ul ar s
ur ge ry
0
866
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
% 100 90 80 70 60 50 40 30 20 10
1st day
2nd day
3rd day
4th day
L TO TA
G er ia tr i cs He ad an d ne In ck te rn al m ed ic in e De rm at ol og y Ca rd io lo Th gy or ac ic su Pu rg lm er y on ar y m ed ic in Pl e an ne d su rg er y
Va sc ul ar s
ur ge ry
0
5th day or later
Fig. 4. Day after admittance on which primary nutritional risk screening was conducted.
repeat screening when admitted more than one week. The low ratings given in the primary evaluations of nutritional risk may have resulted in failure to initiate nutritional therapy in cases where this was required. A previous study with both examination of patients and review of case notes revealed that malnourishment of the patient was not registered in the medical records of 52% (N ¼ 104).14 Other previously conducted audits did not assess coherence between the ratings obtained through nutritional risk screening and information regarding nutritional status obtained from other sources registered in the medical record.7,11,17,19 A self-conducted pre-scheduled audit of nutritional risk screening at Copenhagen University Hospital Gentofte performed in November 2008, found that 55% of patients were screened within 24 h of admittance. In May 2010 the results of a subsequent audit were 75% (Audit from Capital Region: 8500 beds in 10 hospitals, unpublished). In the present study only 15% of the patients were screened within 24 h of admittance, inclusive of incorrect screenings. These contradictory results indicate the fallibility of studies based on self-reported data. The self-reported audits give not only a false positive result of performance, but they also fail to take into account the contradiction between the screening results and data recorded in other sources within the medical record. Previous studies have shown a large variety in the number of patients being screened for nutritional risk while hospitalized and earlier studies with assessment of in-patients’ nutritional status have found poor documentation in medical case records7,11,14,17 A Danish study performed by Rasmussen et al.19 found that 23% of hospital staff routinely performed nutritional risk screening on admitted patients.18 When repeated three years later, the number had increased to 40% of hospital staff assessing nutritional status on a routine basis.13 These findings correlate with a study by Persenius et al., who found that 32e44% of the hospital staff performed all nutritional risk screenings administered to admitted patients.16 In the Scandinavian countries one study found that only 26% of hospital staff used nutritional risk screening of all patients as a routine method, the results from Norway, Sweden and Denmark were 16%, 21% and 40% respectively.15 However, these three
Scandinavian studies were all based on self-reported data, the first two using questionnaires and the last two using telephone interviews. The error resulting from self-reporting discovered at Copenhagen University Hospital Gentofte should be interpreted in light of the results from Rasmussen et al.,18 Lindorff-Larsen et al.,13 Persenius et al.16 and Mowe et al.15 Previous studies based on investigator performed audits, similar to the current study, have given varying results7,11,14,17,19 In a one day audit Porter et al. (N ¼ 46) found that nutritional risk screening had been performed on admittance of 17% and 61% of patients in the Department of Gastroenterology and Liver Disease and the Department of General Medicine, respectively.17 Another one day audit found that 69% of 328 patients had been screened for nutritional risk.11 Furthermore, women who were nutritionally at risk were more likely to be screened than men, and a large proportion of at risk patients were overweight and obese.11 Of 250 case records dietary factors and BMI were registered in respectively 5% and 18% of the patients’ medical records.7 McWhirter and Pennington found that 19% of 500 patients were screened.14 Audits performed in the Danish hospital system, including the present study, generally found a low percentage of patients being screened for nutritional risk. In both the current study and an audit performed by Rasmussen et al. (N ¼ 590) nutritional status were registered accurately in the medical records of only 8% of patients.19 In conclusion, only 8% of patients admitted to Copenhagen University Hospital Gentofte in September 2008 were screened for nutritional risk upon admittance, even though this procedure has been mandatory since 2006. The large proportion of inaccurate and not performed nutritional risk screenings may be explained by lack of instructions, poor training of staff and a low priority of the department management. The failure to screen for nutritional risk may result in nutritionally compromised patients remaining unidentified and not receiving the requisite nutritional therapy. In order to increase the performance of nutritional risk screening of all hospitalized patients a combined effort by hospital staff, hospital executives and The National Health Authorities is needed.
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
Statement of authorship The following authors participated in the study design: NG, AA and SS. NG performed all the practical work and prepared the manuscript and AA, SL and SS approved it. All authors participated in the interpretation of the results and in a critical review of the manuscript. Conflict of interest None declared. Acknowledgements Tina Cuthbertson, secretary at Institute of Human Nutrition, University of Copenhagen proofread the article. References 1. Allison DB, Zannolli R, Faith MS, Heo M, Pietrobelli A, VanItallie TB, et al. Weight loss increases and fat loss decreases all-cause mortality rate: results from two independent cohort studies. Int J Obes Relat Metab Disord 1999;23(6):603e11. 2. Allison SP. Hospital food as treatment. Clin Nutr 2003;22(2):113e4. 3. Correia MITD, Waitzberg DL. The impact of malnutrition on morbidity, mortality, length of hospital stay and costs evaluated through a multivariate model analysis. Clin Nutr 2003;22(3):235e9. 4. Danish Board of Health. Guidelines for doctors, nurses and clinical dieticians: screening and treatment of patients at nutritional risk. Danish Board of Health 2003. 5. Edington J, Boorman J, Durrant ER, Perkins A, Giffin CV, James R, et al. Prevalence of malnutrition on admission to four hospitals in England. Clin Nutr 2000;19(3):191e5. 6. Harrington M, Gibson S, Cottrell RC. A review and meta-analysis of the effect of weight loss on all-cause mortality risk. Nutr Res Rev 2009;22:93e108. 7. Heynes CL, Cook GA. Audit of health promotion practice within a UK hospital: results of a pilot study. J Eval Clin Pract 2006;14:103e9.
867
8. Howard P, Jonkers-Schuitema C, Furniss L, Kyle U, Meuhlebach S, ÖdlundOlin A, et al. Managing the patients journey through enteral nutritional care. Clin Nutr 2006;25:187e95. 9. Kondrup J, Johansen N, Plum ML, Bak L, Højlund Larsen I, Martinsen A, et al. Incidence of nutritional risk and causes of inadequate nutritional care in hospitals. Clin Nutr 2002;21(6):461e8. 10. Kondrup J, Allison SP, Elia M, Vellas B, Plauth M. ESPEN Guidelines for nutrition screening 2002. Clin Nutr 2003;22(4):415e21. 11. Lamb CA, Parr J, Lamb EIM, Warren MD. Adult malnutrition screening, prevalence and management in a United Kingdom hospital: cross-sectional study. Br J Nutr 2009;102:571e5. 12. Lean M, Wiseman M. Malnutrition in hospitals. Br Med J 2008;9(336):290. 13. Lindorff-Larsen K, Rasmussen HH, Kondrup J, Staun M, Ladefoged K, The Scandinavian Nutrition Group. Management and perception of hospital undernutrition e a positive change among Danish doctors and nurses. Clin Nutr 2007;26:371e8. 14. McWhirter JP, Pennington CR. Incidence and recognition of malnutrition in hospital. Br Med J 1994;203:945e8. 15. Mowe M, Bosaeus I, Rasmussen HH, Kondrup J, Unosson M, Irtun Ø. Nutritional routines and attitudes among doctors and nurses in Scandinavia: a questionnaire based survey. Clin Nutr 2006;25:524e32. 16. Persenius MW, Hall-Lord ML, Bååth C, Larsson BW. Assessment and documentation of patients’ nutritional status: perceptions of registered nurses and their chief nurses. J Clin Nurs 2008;17:2125e36. 17. Porter J, Raja R, Cant R, Aroni R. Exploring issues influencing the use of the malnutrition universal screening tool by nurses in two Australian hospitals. J Hum Nutr Diet 2009;22:203e9. 18. Rasmussen HH, Kondrup J, Ladefoged K, Staun M. Clinical nutrition in Danish hospitals: a questionnaire-based investigation among doctors and nurses. Clin Nutr 1999;18(3):153e8. 19. Rasmussen HH, Kondrup J, Staun M, Ladefoged K, Kristensen H, Wengler A. Prevalence of patients at nutritional risk in Danish hospitals. Clin Nutr 2004;23:1009e15. 20. Stratton RJ, Hackston A, Longmore D, Dixon R, Price S, Stroud M, et al. Malnutrition in hospital outpatients and inpatients: prevalence, concurrent validity and ease of use of the ‘malnutrition universal screening tool’ (‘MUST’) for adults. Br J Nutr 2004;92:799e808. 21. Stratton RJ, King CL, Stroud MA, Jackson AA, Elia M. ‘Malnutrition Universal Screening Tool’ predicts mortality and length of hospital stay in acutely ill elderly. Br J Nutr 2006;95:325e30. 22. Wannamethee SG, Shaper G, Lennon L. Reasons for intentional weight loss, unintentional weight loss, and mortality in older men. Arch Intern Med 2005;165:1035e40.
Contents lists available at SciVerse ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original article
Poor performance of mandatory nutritional screening of in-hospital patients Nina Rica Wium Geiker a, Sisse Marie Hørup Larsen a, Steen Stender b, Arne Astrup a, c, * a
Department of Clinical Nutrition, Copenhagen University Hospital Gentofte, DK-2900 Hellerup, Denmark Department of Clinical Biochemistry Copenhagen University Hospital Gentofte, DK-2900 Hellerup, Denmark c Department of Human Nutrition, Faculty of Life Sciences, University of Copenhagen, DK-1958 Frederiksberg C, Denmark b
a r t i c l e i n f o
s u m m a r y
Article history: Received 18 January 2012 Accepted 23 March 2012
Background & aims: Since 2006 it has been mandatory at Copenhagen University Hospital Gentofte to screen all patients for nutritional risk within 24 h of admittance. Audits conducted by department staff estimate that 70e80% of assessments are correctly executed, but the validity of this estimate is unknown. The aim of the present study was to discover the true proportion of hospitalized patients receiving nutritional risk screening within the stipulated time limit and to evaluate the validity of the screening by comparison with medical records. Methods: Retrospective examination of medical records of all patients (N ¼ 3278) hospitalized in September 2008 in 11 different medical specialities were analysed in 2009e2010. Results: Of 2393 medical records 24% of the patients were screened, of these only 65% were screened within the stipulated time limit. Half of the conducted screenings were inaccurate, the most common error being underestimation of nutritional status. Forty-six percent of patients required a secondary nutritional risk screening and 30% were found to be nutritionally at risk. Conclusion: Only 8% of patients received the mandatory nutritional risk screening without procedural errors. We conclude that pre-scheduled, self-conducted audits are not viable as the basis of an assessment of the use of nutritional risk screening. Ó 2012 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Nutrition Malnutrition Patients Audit Screening
1. Introduction Several international studies have revealed a prevalence of 20e60% of moderate to severe malnutrition in hospitalized medical and surgical patients.3,5,9,11,12,14,16,19,20 A large proportion of these patients were malnourished on admittance and malnourishment increased in most of these patients during hospitalization.10 Malnourished patients are found to have prolonged convalescence and admittance, to require more medication, to be susceptible to additional infections and more severe disease, and to have a higher mortality and cost.2,3,5,11,21 There is solid scientific documentation that unintentional weight loss in obese, normal and underweight patients increases all cause mortality.1,5,6,22 In order to alleviate this problem national and international organisations
Abbreviations: A-score, Degree of malnutrition; B-score, Grade of disease; MNA, Mini-nutritional Assessment; NRS 2002, Nutrition Risk Screening; MUST, Malnutrition Universal Screening Tool; GS, Groent System (Electronic patient handling system); OPUS, Electronic patient handling system. * Corresponding author. Department of Human Nutrition, Faculty of Life Sciences, University of Copenhagen, Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark. Tel.: þ45 3533 2476. E-mail address: [email protected] (A. Astrup).
recommend routine nutritional risk screening of hospitalized patients to identify those likely to benefit from nutritional intervention (ASPEN (1995), British Dietetic Association (1999), Department of Health (2001), BAPEN (2000), Council of Europe (2002) and ESPEN.10,8 Nutritional risk screening of hospitalized patients has thus become mandatory in some countries, e.g. United Kingdom and Denmark,4 NHS Quality Improvement Scotland 2003, National Institute for Health and Clinical Excellence 2006). Several tools have been developed to assess and register nutritional status uniformly, and to identify malnutrition in clinical practice, including Mini-Nutritional Assessment (MNA), Nutrition Risk Screening (NRS 2002) and Malnutrition Universal Screening Tool (MUST).10 To respond to and improve patients’ nutritional status, reliable assessment and documentation of nutrition status upon admittance is needed. In Denmark guidelines have been developed and the health authorities have declared malnutrition in hospitals an area of priority.4 The Copenhagen Health Authority has attempted to evaluate the implementation and use of NRS 2002 by conducting half-yearly, pre-scheduled audits, based on hospital staff’s self-reporting of a self-selected sample of 20 patients per department. The results of these audits have indicated that 70e80% of patients are screened for nutritional risk (Audit from Capital Region: 8500 beds in 10 hospitals, unpublished). The aim of the
0261-5614/$ e see front matter Ó 2012 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. doi:10.1016/j.clnu.2012.03.006
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
present study was to discover the true proportion of hospitalized patients receiving nutritional risk screening within the stipulated time limit and to evaluate the validity of the screening by comparison with medical records.
2. Materials and methods A retrospective method was used to study medical records of all patients hospitalized in the period 1ste30th September 2008 at Copenhagen University Hospital Gentofte. Since 2006 it has been mandatory in the capital region of Denmark to screen all hospitalized patients for nutritional risk within 24 h of admittance, followed by repeated screening performed once each week thereafter if patients still are admitted. All nutritional risk screenings are to be performed by hospital health staff. All medical records were retrieved manually and examined by the first author (NG) from February to October 2009. Hospitalized patients were defined as being admitted to a ward and occupying a prescribed bed, and as such they were registered in two different electronic patient handling systems (GS and OPUS). Patient information such as social registration number, diagnosis, day of admittance and discharge, patient type (acute/elective), as well as medical ward and section at the Copenhagen University Hospital Gentofte, were collected from GS. The medical records were thoroughly examined for the presence of the NRS 2002, day of primary nutritional risk screening, and the results of the nutritional risk screening. Data registered in NRS 2002 was controlled for coherence with information on nutritional status otherwise registered in the medical record. The length of stay was recorded and calculated. The primary nutritional risk screening consisted of each patient’s weight, height, calculated BMI, reported dietary intake in the past week, unintentional weight loss over the past three months, and the severity of the illness that gave rise to admittance to the hospital (B-score, Table 1). A secondary nutritional screening is required if the patient’s BMI < 20.5 kg/m2, if the patient’s dietary intake has decreased in the past week, if the patient has lost more than five percent body weight in the last three months, or if the illness is severe (B-score ¼ 3). The degree of malnutrition (A-score) is evaluated and given a score of 0e3, an additional point being added if the patient is 70 years old or more. Nutritional therapy is initiated if the sum of the A-score, B-score and the age related score is equal to three or more (Table 1). The nutritional risk screening was classified as insufficiently performed if one or more of the following criteria were present: - Discrepancy between information found in the NRS 2002 and information elsewhere in the medical records, e.g. two different body weights recorded on the same day. - Failure to register a recent unintentional weight loss.
863
- Incorrect A-score, e.g. a calculated BMI of 17 kg/m2 and appetite reduced by 50% with A-score noted as 1, correct A-score being 3. - Lack of repeat screening in accordance with time of hospitalization. - Lack of secondary nutritional risk screening. - Incomplete nutritional risk screening All statistical analyses were performed using Microsoft Excel and SPSS version 15.0 for Windows. All data was non-parametric, and differences between groups were analysed by Wilcoxon Signed Ranks, Sign test and ANOVA. Data is presented as median (range). 3. Subjects Of 3287 medical records for patients admitted in the period 1ste30th. September 2008 a total of 2393 (72%) were found in the patient record archives, the remaining records probably still being on the various clinical wards. All patients had been admitted to Copenhagen University Hospital Gentofte in the period 1ste30th. September 2008. The median length of time of hospitalization was two days (1e128), with a variation between the various medical specialities. Table 2 shows the distribution of patients in the medical specialities and the length of hospitalization. Neither the Department of Gynaecology and Obstetrics nor the Department of Paediatrics performed nutritional risk screenings during the investigated period in the present study; this was due to lack of nutritional risk screening tools suitable for these two patient groups. 4. Results Seventy-six percent (N ¼ 1819) of the 2393 patients were not screened for nutritional risk at all during their hospitalization (Fig. 1). A total of 24% (N ¼ 574) were screened and 35% of these were screened later than 24 h after admittance. Thirty percent of patients screened for nutritional risk were at risk of malnutrition, and thus required nutritional therapy. Nutritional status was correctly assessed and recorded within the stipulated time limit for a total of 8% (N ¼ 194) of patients hospitalized in September 2008. Forty-six percent of the patients administered the primary nutritional risk screening required a secondary nutritional risk screening (Fig. 2). However, a secondary nutritional risk screening was conducted in only 14% of these cases. Forty-eight percent (28e63%) of all the conducted nutritional risk screenings were inadequate, or had A- or B-score in disagreement with data registered in other places within the medical record. In example there was noted a 10% weight loss the past month in the medical record but patients was only given a score of one in B-score where a score of three would have been correct
Table 1 Score for secondary nutritional screening. A Under nutrition
B Severity of illness in regards to nutritional needs (i.e.)
Score ¼ 0 None Score ¼ 1 Light
Normal nutritional status >5% weight loss in 3 months or intake 50e75% of dietary needs past week. BMI not an issue her
Score ¼ 2 Moderate
>5% weight loss in 2 months or intake 25e50% of dietary needs past week or BMI 18.5e20.5 þ affected general condition >5% weight loss in 1 month or intake 0e25% of dietary needs past week or BMI less than 18.5 þ affected general condition
Normal needs - Collum femoris fracture - Chronic patients with exacerbations: liver cirrhosis, COPD - Chronic dialysis - Diabetes - Cancer - Large abdominal surgery (colectomy, gastrectomy, hepatectomy) - Post operational ATN - Apoplexia - Severe pneumonia - Cranial trauma - More than 50% burn - Severe infections (sepsis) - Patients in intensive therapy with multiple organ failure
Score ¼ 3 Severe
864
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
Table 2 Number of retrieved medical records and length of admittance, presented for each medical specialitya. Medical speciality
Patients, N
Retrieved, N (%)
Admittance, days
Vascular surgery Geriatrics Nose and throat Internal medicine Gynaecology-obstetrics Dermatology Paediatrics Cardiology Thoracic surgery Pulmonary medicine Planned surgery TOTAL
108 287 275 565 298 42 392 702 132 137 337 3278
31 218 149 416 258 13 294 572 109 84 249 2393
3 8 3 2 2 16 1 7 2 4 2 2
(1e31) (1e56) (1e35) (1e49) (1e10) (4e29) (1e50) (1e28) (1e26) (1e128) (1e50) (1e128)
Presented as median (range).
(Table 1). This discrepancy was found in both primary and secondary nutritional risk screenings (Fig. 3). An incorrectly low Ascore (nutritional status) or lacks of repeat screening were the most frequent errors. Thirty five percent of the screenings were conducted after 24 h following admittance (Fig. 4). Eight percent of nutritional screenings were not performed until the fifth day of admittance or later. Sixty-seven percent (N ¼ 1615) of the 2393 admitted patients were acute, of those were 26% (N ¼ 417) screened for nutritional risk, were as 19% (N ¼ 148) of 778 elective patients were screened for nutritional risk during admittance (P ¼ 0.000). There was no difference in accuracy of nutritional screening among acute and elective patients. Respectively, 59% and 83% (P ¼ 0.000) of the screenings for nutritional risk of acute and elective patients were performed on the first day of admittance. Acute patients were admitted for a longer period than elective (4.6 vs. 3.8 days, P ¼ 0.00). Among female and male patients; 52% and 43% respectively were incorrectly screened for nutritional risk (P ¼ 0.031) and 72%
5. Discussion Nutritional risk screening of all hospitalized patients has been mandatory at Copenhagen University Hospital Gentofte since 2006. The present review of medical records finds that only 8% of 2393 patients hospitalized in September 2008 were appropriately screened using the recommended NRS 2002 within 24 h of admittance. In the present study almost 40% of patients admitted longer than two days were screened for nutritional risk, whereas this was the fact in only 5% (N ¼ 28) of patients with one day of hospitalization. There was a tendency that patients with a stay longer than two days were screened for nutritional risk on their second day of admission or later. A longer hospitalization could results in more opportunities for the staff to perform the registration. In the present study acute patients were admitted for a longer period and more were screened for nutritional risk than elective patients; this correlates with longer hospitalization resulted in higher percentage of screened patients. Though only 19% of elective patients were screened during their hospitalization in the present study 80% of those were screened within the stipulated time limit. Reasons for lack of nutritional risk screening have been investigated in previous studies.9,17,18 The reasons most often given for failure to perform nutritional risk screenings are lack of instruction on how to use the nutritional risk screening form, too much paperwork/lack of time, lack of knowledge of nutritional risk
% 100 90 80 70 60 50 40 30 20 10
Va sc ul ar s
L
ur ge ry G er ia No tr i se cs an d In th te ro G rn at yn al ae m co ed lo ic gy in e -o bs te tri cs De rm at ol og y Pa ed Th ia tr i or cs ac ic su rg er y Ca Pu rd lm io on lo ar gy y m e Pl di cin an ne e d su rg er y
0 TO TA
a
(29) (76) (54) (74) (87) (31) (75) (81) (83) (61) (74) (73)
and 62% respectively were acutely admitted. There was no difference between gender regarding percentage of nutritionally screened or day of performed screening (P ¼ 0.319 and P ¼ 0.249) neither was there difference between fraction of acute female and male patients being screened for nutritional risk. Five percent (N ¼ 28) of patients hospitalized for one day (N ¼ 607) and 38% of patients hospitalized more than two days (N ¼ 1101) were screened for nutritional risk, however 50% being inaccurate.
Fig. 1. Percentage of patients who were not screened for nutritional risk during their hospitalization. Data is presented for each medical speciality.
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
865
%
100 90 80 70 60 50 40 30 20 10 L TO TA
G er ia tr i cs He ad an d ne In ck te rn al m ed ic in e De rm at ol og y Ca rd io lo Th gy or ac ic su Pu rg lm er on y ar y m ed ic Pl in e an ne d su rg er y
Va sc ul ar s
ur ge ry
0
Fig. 2. Percentage of patients identified by primary nutritional risk screening results, as being in need of secondary nutritional screening, but did not receive this.
screening, or simply forgetting it.18,9 Of 857 hospital staff taking part in one of these studies 40% found it difficult to identify patients at nutritional risk despite the fact that more than 70% had used nutritional therapy as treatment. Seventy six percent of staff considered it necessary to screen patients for nutritional risk, but only 23% routinely performed nutritional risk screening when admitting a patient.18 A questionnaire similar to the one used by Rasmussen et al.18 was repeated in 2004, finding that 40% of staff now performed nutritional risk screening as a routine assessment.13 The studies by Rasmussen et al.18 and Lindorff-Larsen et al.13 do not necessarily report responses from the same staff members as the population was random, which might influence the results and make them difficult to compare. Furthermore, both studies are based on self-reported data and thus may not reflect the actual number of patients being screened. Even so, the improvement seen from 1997 to 2004 may reflect a true increase in
nutritional risk screening, though the results of both the current audit and of Rasmussen et al.19 make this questionable. Nearly 50% of the registered nutritional risk screenings in the present review of medical records were inadequate or were inconsistent with medical records. Ten percent more female than male patients had inaccurate screenings of nutritional risk performed; and 10% more female patients were acutely admitted which could result in more stressful admittance with too little time to register nutritional status accurate. But because of no difference in accuracy between acute and elective patients being screened, or between fraction of female and male patients or acute female and male patients; this seems unlikely to be the explanation. There does not seem to be a rational explanation to why mandatory nutritional risk screenings were registered in less female than male patients during admission. The most frequent mistakes were a misleadingly low A-score (evaluation of nutritional status, Table 1) and lack of
% 70 60 50 40 30 20 10
Fig. 3. Percentage of patients with inadequate or incorrect nutritional risk screening.
TO TA L
G er ia tr i cs H ea d an d ne In ck te rn al m ed ic in e De rm at ol og y Ca rd io lo Th gy or ac ic su Pu rg lm er on y ar y m ed ic Pl in e an ne d su rg er y
Va sc ul ar s
ur ge ry
0
866
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
% 100 90 80 70 60 50 40 30 20 10
1st day
2nd day
3rd day
4th day
L TO TA
G er ia tr i cs He ad an d ne In ck te rn al m ed ic in e De rm at ol og y Ca rd io lo Th gy or ac ic su Pu rg lm er y on ar y m ed ic in Pl e an ne d su rg er y
Va sc ul ar s
ur ge ry
0
5th day or later
Fig. 4. Day after admittance on which primary nutritional risk screening was conducted.
repeat screening when admitted more than one week. The low ratings given in the primary evaluations of nutritional risk may have resulted in failure to initiate nutritional therapy in cases where this was required. A previous study with both examination of patients and review of case notes revealed that malnourishment of the patient was not registered in the medical records of 52% (N ¼ 104).14 Other previously conducted audits did not assess coherence between the ratings obtained through nutritional risk screening and information regarding nutritional status obtained from other sources registered in the medical record.7,11,17,19 A self-conducted pre-scheduled audit of nutritional risk screening at Copenhagen University Hospital Gentofte performed in November 2008, found that 55% of patients were screened within 24 h of admittance. In May 2010 the results of a subsequent audit were 75% (Audit from Capital Region: 8500 beds in 10 hospitals, unpublished). In the present study only 15% of the patients were screened within 24 h of admittance, inclusive of incorrect screenings. These contradictory results indicate the fallibility of studies based on self-reported data. The self-reported audits give not only a false positive result of performance, but they also fail to take into account the contradiction between the screening results and data recorded in other sources within the medical record. Previous studies have shown a large variety in the number of patients being screened for nutritional risk while hospitalized and earlier studies with assessment of in-patients’ nutritional status have found poor documentation in medical case records7,11,14,17 A Danish study performed by Rasmussen et al.19 found that 23% of hospital staff routinely performed nutritional risk screening on admitted patients.18 When repeated three years later, the number had increased to 40% of hospital staff assessing nutritional status on a routine basis.13 These findings correlate with a study by Persenius et al., who found that 32e44% of the hospital staff performed all nutritional risk screenings administered to admitted patients.16 In the Scandinavian countries one study found that only 26% of hospital staff used nutritional risk screening of all patients as a routine method, the results from Norway, Sweden and Denmark were 16%, 21% and 40% respectively.15 However, these three
Scandinavian studies were all based on self-reported data, the first two using questionnaires and the last two using telephone interviews. The error resulting from self-reporting discovered at Copenhagen University Hospital Gentofte should be interpreted in light of the results from Rasmussen et al.,18 Lindorff-Larsen et al.,13 Persenius et al.16 and Mowe et al.15 Previous studies based on investigator performed audits, similar to the current study, have given varying results7,11,14,17,19 In a one day audit Porter et al. (N ¼ 46) found that nutritional risk screening had been performed on admittance of 17% and 61% of patients in the Department of Gastroenterology and Liver Disease and the Department of General Medicine, respectively.17 Another one day audit found that 69% of 328 patients had been screened for nutritional risk.11 Furthermore, women who were nutritionally at risk were more likely to be screened than men, and a large proportion of at risk patients were overweight and obese.11 Of 250 case records dietary factors and BMI were registered in respectively 5% and 18% of the patients’ medical records.7 McWhirter and Pennington found that 19% of 500 patients were screened.14 Audits performed in the Danish hospital system, including the present study, generally found a low percentage of patients being screened for nutritional risk. In both the current study and an audit performed by Rasmussen et al. (N ¼ 590) nutritional status were registered accurately in the medical records of only 8% of patients.19 In conclusion, only 8% of patients admitted to Copenhagen University Hospital Gentofte in September 2008 were screened for nutritional risk upon admittance, even though this procedure has been mandatory since 2006. The large proportion of inaccurate and not performed nutritional risk screenings may be explained by lack of instructions, poor training of staff and a low priority of the department management. The failure to screen for nutritional risk may result in nutritionally compromised patients remaining unidentified and not receiving the requisite nutritional therapy. In order to increase the performance of nutritional risk screening of all hospitalized patients a combined effort by hospital staff, hospital executives and The National Health Authorities is needed.
N.R.W. Geiker et al. / Clinical Nutrition 31 (2012) 862e867
Statement of authorship The following authors participated in the study design: NG, AA and SS. NG performed all the practical work and prepared the manuscript and AA, SL and SS approved it. All authors participated in the interpretation of the results and in a critical review of the manuscript. Conflict of interest None declared. Acknowledgements Tina Cuthbertson, secretary at Institute of Human Nutrition, University of Copenhagen proofread the article. References 1. Allison DB, Zannolli R, Faith MS, Heo M, Pietrobelli A, VanItallie TB, et al. Weight loss increases and fat loss decreases all-cause mortality rate: results from two independent cohort studies. Int J Obes Relat Metab Disord 1999;23(6):603e11. 2. Allison SP. Hospital food as treatment. Clin Nutr 2003;22(2):113e4. 3. Correia MITD, Waitzberg DL. The impact of malnutrition on morbidity, mortality, length of hospital stay and costs evaluated through a multivariate model analysis. Clin Nutr 2003;22(3):235e9. 4. Danish Board of Health. Guidelines for doctors, nurses and clinical dieticians: screening and treatment of patients at nutritional risk. Danish Board of Health 2003. 5. Edington J, Boorman J, Durrant ER, Perkins A, Giffin CV, James R, et al. Prevalence of malnutrition on admission to four hospitals in England. Clin Nutr 2000;19(3):191e5. 6. Harrington M, Gibson S, Cottrell RC. A review and meta-analysis of the effect of weight loss on all-cause mortality risk. Nutr Res Rev 2009;22:93e108. 7. Heynes CL, Cook GA. Audit of health promotion practice within a UK hospital: results of a pilot study. J Eval Clin Pract 2006;14:103e9.
867
8. Howard P, Jonkers-Schuitema C, Furniss L, Kyle U, Meuhlebach S, ÖdlundOlin A, et al. Managing the patients journey through enteral nutritional care. Clin Nutr 2006;25:187e95. 9. Kondrup J, Johansen N, Plum ML, Bak L, Højlund Larsen I, Martinsen A, et al. Incidence of nutritional risk and causes of inadequate nutritional care in hospitals. Clin Nutr 2002;21(6):461e8. 10. Kondrup J, Allison SP, Elia M, Vellas B, Plauth M. ESPEN Guidelines for nutrition screening 2002. Clin Nutr 2003;22(4):415e21. 11. Lamb CA, Parr J, Lamb EIM, Warren MD. Adult malnutrition screening, prevalence and management in a United Kingdom hospital: cross-sectional study. Br J Nutr 2009;102:571e5. 12. Lean M, Wiseman M. Malnutrition in hospitals. Br Med J 2008;9(336):290. 13. Lindorff-Larsen K, Rasmussen HH, Kondrup J, Staun M, Ladefoged K, The Scandinavian Nutrition Group. Management and perception of hospital undernutrition e a positive change among Danish doctors and nurses. Clin Nutr 2007;26:371e8. 14. McWhirter JP, Pennington CR. Incidence and recognition of malnutrition in hospital. Br Med J 1994;203:945e8. 15. Mowe M, Bosaeus I, Rasmussen HH, Kondrup J, Unosson M, Irtun Ø. Nutritional routines and attitudes among doctors and nurses in Scandinavia: a questionnaire based survey. Clin Nutr 2006;25:524e32. 16. Persenius MW, Hall-Lord ML, Bååth C, Larsson BW. Assessment and documentation of patients’ nutritional status: perceptions of registered nurses and their chief nurses. J Clin Nurs 2008;17:2125e36. 17. Porter J, Raja R, Cant R, Aroni R. Exploring issues influencing the use of the malnutrition universal screening tool by nurses in two Australian hospitals. J Hum Nutr Diet 2009;22:203e9. 18. Rasmussen HH, Kondrup J, Ladefoged K, Staun M. Clinical nutrition in Danish hospitals: a questionnaire-based investigation among doctors and nurses. Clin Nutr 1999;18(3):153e8. 19. Rasmussen HH, Kondrup J, Staun M, Ladefoged K, Kristensen H, Wengler A. Prevalence of patients at nutritional risk in Danish hospitals. Clin Nutr 2004;23:1009e15. 20. Stratton RJ, Hackston A, Longmore D, Dixon R, Price S, Stroud M, et al. Malnutrition in hospital outpatients and inpatients: prevalence, concurrent validity and ease of use of the ‘malnutrition universal screening tool’ (‘MUST’) for adults. Br J Nutr 2004;92:799e808. 21. Stratton RJ, King CL, Stroud MA, Jackson AA, Elia M. ‘Malnutrition Universal Screening Tool’ predicts mortality and length of hospital stay in acutely ill elderly. Br J Nutr 2006;95:325e30. 22. Wannamethee SG, Shaper G, Lennon L. Reasons for intentional weight loss, unintentional weight loss, and mortality in older men. Arch Intern Med 2005;165:1035e40.