Prevalence of malnutrition in adults in Queensland public hospitals and residential aged care facilities.

Abstract

Aim: To determine the prevalence of malnutrition in Queensland public acute and residential aged care facilities, and explore effects of variables associated with malnutrition in these populations.

Methods: A multicentre, cross-sectional audit of a convenience sample of subjects was carried out as part of a larger audit of pressure ulcers in Queensland public acute and residential aged care facilities in 2002 and again in 2003. Dietitians in 20 hospitals and six aged care facilities conducted single-day nutritional status audits of 2208 acute and 839 aged care subjects using the Subjective Global Assessment, in either or both audits. Subjects excluded were obstetric, same-day, paediatric and mental health patients. Weighted average proportions of nutritional status categories for acute and residential aged care facilities across the two audits were determined and compared. The effects of gender, age, facility location and medical specialty on malnutrition were determined via logistic regression.

Results: A mean of 34.7 [+ or -] 4.0% and 31.4 [+ or -] 9.5% of acute subjects and a median of 50.0% and 49.2% of residents of aged care facilities were found to be malnourished in Audits 1 and 2, respectively. Variables found to be significantly associated with an increased odds risk of malnutrition included: older age groups, male (in residential aged care facilities), metropolitan location of facility and medical specialty, in particular, oncology and critical care.

Conclusion: Malnutrition is significant in public acute and residential aged care facilities in Queensland. Action must be taken to increase the recognition, prevention and treatment of malnutrition especially in high-risk groups.

Key words: aged, hospitalisation, malnutrition, Subjective Global Assessment.

INTRODUCTION

In Australia, disorders related to 'overnutrition' are now a national health priority. (1) There is, however, limited awareness of the existence and extent of malnutrition. Numerous studies investigating the prevalence of malnutrition and its consequences have been published. (2-25) The majority of studies have been undertaken in hospital settings, with relatively few in the community or residential care facilities. Stratton et al. (26) provided an international review of the prevalence of malnutrition in patients with different disease groups, mixed diagnoses, across different age groups and settings. They concluded that malnutrition was common in hospitals (10-60%), in residential aged care facilities ([greater than or equal to]50%), and in free-living individuals with severe or multiple diseases (>10%). The reported prevalence of malnutrition in Australian studies is 12-42% in acute settings, (15-22) 6-49% in rehabilitation settings, (20,23,24) 5% in community domiciliary care setting, (25) and no studies conducted in residential aged care settings. The wide variation in the reported prevalence of malnutrition is due to variation in the methodology and criteria used to assess nutritional status, diagnoses of patients and setting (e.g. hospital or community). This makes it difficult to compare studies and to ascertain the actual prevalence of malnutrition in hospitals, residential care facilities or the community, (27) or to apply these findings to the Australian setting in general.

The purpose of the present study was to determine the prevalence of malnutrition in Queensland public acute and residential aged care facilities, and identify variables which may be associated with malnutrition in these populations.

METHODS

The study involved the collection of nutritional status data of subjects participating in a larger study investigating the prevalence of pressure ulcers. The multicentre audits were conducted initially in 2002 (Audit 1) and repeated 12 months later (Audit 2), after the implementation of pressure ulcer guidelines in 20 hospitals and six residential aged care facilities. Facility involvement in the nutritional status audits was determined by whether a facility employed dietitians and whether the dietitians nominated to participate. Four of the 20 hospitals and four of the six residential aged care facilities participated in both nutritional status audits, with other facilities being involved in either Audit 1 or Audit 2 only. A larger number of acute facilities were able to participate in Audit 2.

Nutritional status sample

Audits were conducted on a single day for each facility involved, with all available subjects potentially eligible for inclusion. Exclusions included: obstetric, paediatric, mental health and same-day patients. The project was approved by Queensland Health as a quality improvement project, and as such, no formal ethics approval was required, although subjects or their next of kin provided informed written consent to be included in the audits. Participation in the study was dependent on whether dietitians could undertake the audits; thus the sample for acute facilities was biased towards larger facilities in which patient acuity is greater.

Variables

A data set was extracted from the larger study database for the purposes of the present study program. Variables were collected by trained audit staff, usually nurses, and were limited to those collected for the larger study and included: audit number/year, facility, age, gender and medical specialty (acute only). Nutritional status data were independently collected by dietitians.

Nutritional status of subjects was assessed using the Subjective Global Assessment (SGA), (28) which determines nutritional status based upon a medical assessment and physical examination. It has a high degree of inter-rater reliability (with assessor agreement of 80-90% and kappa statistics of 0.75-0.78), (4,6,7,21,22,28) and good predictive and convergent validity, correlating well with measures of morbidity, and traditional objective nutritional parameters. (29,30) The SGA is a valid nutrition assessment tool in a variety of patient population settings, including: surgery, (28) geriatric, (12,30-32) oncology (33) and renal, (34) allowing for comparisons to be made across different population settings. Many recent Australian and international studies investigating the prevalence of malnutrition have also used the SGA. (4,6-12,19-22)

Standardised training in performing SGA was conducted for dietitians involved in the nutritional status audits by the authors (MB, SA). Inter-rater reliability between a convenience sample of 15 dietitians assessing nine patients showed good agreement with the use of the SGA (kappa 0.9, P < 0.001).

Six medical specialties were categorised from classifications provided by facilities: medical, surgical, oncology, critical care, rehabilitation and aged care. Facility type (acute or residential) was determined by the classification of the facility by Queensland Health. Facility location was based on the Rural, Remote and Metropolitan Areas Classifications, 1991 Census edition (ABS, Canberra). Seven categories were collapsed into three: metropolitan, regional and rural/remote. Age was categorised to: [less than or equal to]40, 41-60, 61-80 and [greater than or equal to]81 years.

Determining a representative sample

Demographic variables of the sample were compared with the relevant Queensland public health facility population data where available, to determine whether the sample was representative. As population data for residential aged care were not available, comparison of demographic variables was made with the larger pressure ulcer audit study population. Queensland Health (public) facility population data, including the average daily occupied beds for all facilities, average age and gender for the populations being studied, were provided for 2002/2003 by Health Information Services (Queensland Health). Comparison of descriptive variables was made using t-tests for continuous variables and chi-squared tests for categorical variables. The proportion of subjects who declined consent or were unavailable, and their demographic data, were not available. The number of repeated cases between the two audits was quantified, with only 0.03% of acute subjects and 28.2% of aged care residents found to be in both audits. There were no significant differences between demographic variables in each audit for residential aged care when the duplicate cases were removed. Hence, comparison between Audit 1 and Audit 2 as independent data sets was deemed reasonable.

Data analyses

The percentages of well-nourished, moderately and severely malnourished subjects were determined for each facility. An average percentage across facilities was then determined, for acute and residential aged care facilities, for each audit. The average percentage was weighted by the number of cases in each facility. Analyses were carried out using SPSS for Windows (Version 12.0, 2003, SPSS Inc., Chicago, IL, USA).

To determine the effects of available variables on nutritional status in acute and residential aged care facilities, logistic regression was conducted at the bivariate level to determine crude odds ratios, and then in a multivariable model to ascertain their independent influences. The potential clustering effect of facility was accounted for in the model using an analysis of correlated data approach with SUDAAN statistical package (Version 7.5.2A, 1998, Research Triangle Institute, Research Triangle Park, NC, USA). Significant design effect was established for the variables of facility location and medical specialty in acute facilities, and facility location in residential aged care facilities, confirming the use of this approach.

Statistical significance was predetermined as the conventional level of P < 0.05. In all analyses, acute care and residential aged care facilities were kept separate, as it was decided that these types of facilities were significantly different from one another to require separate analysis and interpretation. For logistic regression, it was decided that data from Audit 1 and Audit 2 for acute facilities would be pooled, if no significant difference was found in the proportions of nutritional status categories between the two years, but that only Audit 1 data would be used for residential aged care facilities because it would be inappropriate to pool Audit 1 and Audit 2 data, due to the high percentage of dependent cases between audits. For comparison within variables, the most frequent category of a variable was chosen as the referent. The association between nutritional status and pressure ulcers was not investigated here.

RESULTS

There were 774 and 1434 acute patients from 8 and 16 hospitals, and 381 and 458 residents from 5 residential aged care facilities in Audit 1 and Audit 2, respectively This represents approximately 40% and 80% of the average daily occupied beds for the acute and residential aged care facilities, respectively, that were involved in the audits. Table 1 shows the demographic data for the sample. While there were no gender differences, the average age of the sample was significantly higher than the equivalent Queensland Health acute population by approximately eight years. There were also significantly fewer acute subjects represented from regional and rural and remote areas than metropolitan areas as expected. For residential aged care subjects, there were significantly fewer represented from regional areas, but adequate representation from rural and remote areas. The residential aged care sample was otherwise found to be similar to the larger study sample, which represented more than 80% of the Queensland public residential aged care population.

Table 2 shows the average percentages of nutritional status classifications across acute and residential aged care facilities. A mean of 34.7 [+ or -] 4.0% and 31.4 [+ or -] 9.5% of acute subjects, and a median of 50% and 49.2% of aged care subjects, were found to be malnourished in Audits 1 and 2, respectively Approximately 20% of malnourished subjects were severely malnourished. The difference in the prevalence of malnutrition between Audit 1 and Audit 2 for both acute and residential aged care facilities was not clinically or statistically significant.

The multivariable relationships of various parameters on nutritional status for the acute and residential aged care facilities are presented in Table 3. The multivariable (mutually adjusted) models generally strengthened the relationships established at the bivariate level. As there was no significant difference in the prevalence of malnutrition between Audits 1 and 2 for acute facilities, data were combined for bivariate and multivariable analyses. Gender did not have an effect on nutritional status in acute facilities, but did in residential aged care facilities. Being male in a residential aged care facility was shown to have an adjusted odds risk of 1.2 (95% CI 1.1-1.3, P = 0.003) compared with being female. Age group had a significant effect on nutritional status. Compared with the age group of 61-80 years, younger age groups have a lower odds risk of being malnourished (acute: <40 years OR = 0.6 (95% CI 0.4-0.8, P < 0.001); 41-60 years OR = 0.6 (95% CI 0.5-0.7, P < 0.001)), although this did not reach statistical significance in the residential aged care facilities. Being aged 81 years or above, compared with 61-80 years, had an adjusted odds risk of being malnourished of 1.7 (95% CI 1.5-2.0, P < 0.001) in acute facilities and 1.4 (95% CI 1.2-1.6, P < 0.001) in residential aged care facilities Facility location also had an effect on nutritional status. In the acute setting, subjects from rural and remote locations had a significantly lower odds risk of being malnourished (OR = 0.1, 95% CI 0.02-0.5, P = 0.007) than those from metropolitan facilities. This pattern was also demonstrated for regional facilities, but did not reach statistical significance. In the residential aged care setting, there was a lower odds risk of being malnourished if from a regional facility (OR = 0.1, 95% CI 0.01-0.8, P = 0.03). Again, this pattern was demonstrated for rural and remote facilities, but did not reach statistical significance. Medical specialty in acute facilities was found to have an effect on nutritional status. Compared with medical patients, oncology patients had an adjusted higher odds risk of being malnourished (OR = 2.3, 95% CI 1.5-3.8, P = 0.001) as did critical care patients (OR = 1.6, 95% CI 1.1-2.3, P = 0.02). All other medical specialties were not considered significantly different from medical subjects.

DISCUSSION

In this observational multicentre study, malnutrition was found to be present in approximately 30% of more than 2200 acute patients, and 50% of more than 800 aged care residents, in a number of public facilities in Queensland. These results are consistent with other recent studies conducted both overseas and in Australia. (2-25) The present study, however, has been conducted on a large number of cases across 20 acute and 6 residential aged care facilities, and therefore provides the first significant evidence of the extent of malnutrition in public acute and residential aged care facilities in Australia.

The prevalence of malnutrition found in the present study is similar to other across-admission studies conducted in Australian hospitals, that is, about 30-50%. (16,21,22) Studies of nutritional status conducted at admission tend to indicate a prevalence of malnutrition in the order of 12-20%. (18-20) Studies show that the prevalence of malnutrition increases as the length of stay of patients studied increases. (9) Patients who stay longer in hospital are more likely to be malnourished due to severity of illness, and that individual patients' nutritional status declines during admission, as demonstrated in several studies. (2,5,35)

There was no difference found in the prevalence of malnutrition between Audit 1 and Audit 2 for both acute and residential aged care facilities. After Audit 1, pressure ulcer guidelines, which included referral for nutrition assessment and intervention for subjects with, or at high risk of, pressure ulcers, were introduced into facilities. An unexpected finding was the unchanged prevalence of malnutrition between the two audits, when it could be expected that this would reduce with the introduction of these guidelines. Poor referral for nutrition assessment and intervention was reported by the facility dietitians, which could explain this finding. Other studies have shown that the implementation of a comprehensive nutrition screening and intervention program decreases the prevalence of malnutrition and improves patient outcomes. (36,37)

In the present study, younger age groups had a lower odds risk of being malnourished, and those over the age of 80 years had a higher odds risk of being malnourished compared with those aged 61-80 years. This has also been demonstrated in other studies. (10,11) As disease prevalence generally increases with age, rates of associated malnutrition are also likely to increase in older people. (26) As the average age of the sample was older than the Queensland Health acute population, it is expected that the level of malnutrition reported here for acute facilities would be higher than for the Queensland Health acute population in general. However, the present study demonstrates that the nutritional needs of older people in hospitals and residential aged care facilities require greater attention. Male individuals in residential aged care facilities had a higher odds risk of being malnourished. This, however, contrasts with the findings of another study in Finland which found that female aged care residents had a higher risk of malnutrition. (14) The association of gender with nutritional status needs further exploration. The odds risk of being malnourished was lower in regional and rural and remote facilities. Other studies have noted that the prevalence of malnutrition is greater in tertiary hospitals than in general hospitals. (3) This is most likely due to the greater acuity of disease, which is the primary cause of malnutrition in developed countries. (26,27) Due to the bias of the present study towards being undertaken in metropolitan (and therefore tertiary) facilities, it is expected the level of malnutrition reported here for acute facilities would be higher than for the Queensland public acute population in general. The lower odds risk of being malnourished from a regional or rural and remote residential aged care facility, compared with a metropolitan facility, was not an expected finding and requires more investigation as insufficient data are available here to make conclusive findings. The prevalence of malnutrition reported here for residential aged care facilities may be an underestimate of the true prevalence as facilities involved in the study had regular dietetic services, which should mean better nutrition practices than facilities without regular access to nutritional expertise.

Malnutrition has been identified as a significant problem in patients with respiratory disease, gastrointestinal and liver disease, HIV and AIDS, malignancy, neurological diseases, renal disease and critical illness, and in orthopaedic and surgical patients. (26) The present study found that, compared with medical patients, oncology and critical care patients had a significantly higher odds risk of being malnourished. This is most likely because patients under these medical specialties overall have greater metabolic stress. Patients in rehabilitation care have been reported to have a higher prevalence of malnutrition; (13,20,24) however, this was not demonstrated in the present study.

Malnutrition is independently associated with adverse clinical outcomes and costs. (38-40) And there is evidence that nutrition intervention results in statistically significantly and clinically relevant improvements on mortality, complications and length of stay. (26) However, many studies have demonstrated that malnutrition continues to frequently go unrecognised and untreated. (2,6,8,21,22,27) Recent Australian studies (21,22) found that a majority of patients assessed as malnourished had not been previously identified or were not documented as such, and were not receiving any specialised nutrition care. Reasons for the lack of awareness and recognition of malnutrition include: limited training and knowledge of clinical staff; misbelief that malnutrition is an inevitable part of the disease process and resistant to therapy; failure to regard nutrition as an important part of care; scarcity of specialist clinical nutrition appointments; lack of good practice guidelines and nationally agreed standards; and lack of organisation of nutritional services linking relevant disciplines. (40,41)

Limitations of the present study include: the use of a convenience sample which, for the acute sample, was not found to be representative of the Queensland acute public population, and difficulty determining whether the residential aged care sample was representative, and therefore it is difficult to apply these results to Australian public hospitals and residential aged care facilities in general; the unavailability of data on subjects who declined or were unavailable for the audits, although this was minimised as a limitation by determining the representativeness of the sample; the number of clinicians determining nutritional status was also a potential limitation of the present study, but was minimised through standardised training, which has been shown to result in good inter-rater reliability; and misclassification of subjects' medical specialty was another potential limitation of the present study, and the collapsing of medical specialties into fewer categories to ensure adequate numbers in categories may have resulted in altering or masking the effects of some medical specialties. The strengths of the present study include the large numbers across multiple facilities with the inclusion of residential aged care facilities, and the standardised nutrition assessment methodology.

CONCLUSION

This large-scale multicentre study provides evidence that malnutrition is significant in public acute and residential aged care facilities In Queensland. Being malnourished was found to be significantly associated with: older age groups, male gender (in residential aged care facilities), metropolitan location (compared with regional and rural and remote locations) and medical specialty, in particular, oncology and critical care patients. Action must be taken to increase the recognition, prevention and appropriate treatment of malnutrition especially in higher-risk groups.

ACKNOWLEDGEMENTS

The nutritional status audits were in part funded by the Queensland Health Pressure Ulcer Prevention Project. We thank project team members, and in particular Nancy Magazinovich, Project Manager, for including the nutritional status component of the audits. We acknowledge the many dietitians in Queensland Health who undertook to participate in the nutritional status audits, especially Denise Cruickshank for assisting with the organisation. Acknowledgement is also made to the late Associate Professor Carla Patterson for her initial assistance with the present study.

REFERENCES

1 NHMRC. Acting on Australia's Weight--A Strategic Plan for the Prevention of Overweight and Obesity. Canberra: National Health and Medical Research Council, 1997.

2 McWhirter JP, Pennington CR. Incidence and recognition of malnutrition in hospital. BMJ 1994; 308: 945-8.

3 Edington J, Boorman J, Durrant ER et al. Prevalence of malnutrition on admission to four hospitals in England. Clin Nutr 2000; 19: 191-5.

4 Waitzberg DL, Caiaffa WT, Correia MITD. Hospital malnutrition. The Brazilian National Survey (IBRANUTRI). A study of 4000 patients. Nutrition 2001; 17: 573-80.

5 Kondrup J, Johansen N, Plum LM et al. Incidence of nutritional risk and causes of inadequate nutritional care in hospitals. Clin Nutr 2002; 21: 461-8.

6 Correia MITD, Campos ACL. Prevalence of hospital malnutrition in Latin America: the Multicenter ELAN Study. Nutrition 2003; 19: 823-5.

7 Wyszynski DF, Perman M, Crivelli A. Prevalence of hospital malnutrition in Argentina: preliminary results of a population-based study. Nutrition 2003; 19: 115-19.

8 Planas M, Audivert S, Perez-Portabella C et al. Nutritional status among adult patients admitted to a university-affiliated hospital in Spain at the time of genoma. Clin Nutr 2004; 23: 1016-24.

9 Penie JB. State of malnutrition in Cuban hospitals. Nutrition 2005; 21: 487-97.

10 Olmos M, Vazquez M, Lopez M, del Campo Perez V. Nutritional status study of inpatients in hospitals of Galicia. Eur J Clin Nutr 2005; 59: 938-46.

11 Kyle UG, Unger P, Mensi N, Genton L, Pichard C. Nutrition status in patients younger and older than 60 years at hospital admission: a controlled population study in 995 subjects. Nutrition 2002; 18: 463-9.

12 Persson MD, Brismar KE, Katzarski KS, Nordenstrom J, Cederholm T. Nutritional status using Mini Nutritional Assessment and Subjective Global Assessment predict mortality in geriatric patients. J Am Geriatr Soc 2002; 50: 1996-2002.

13 Compan B, Di Castri A, Plaze JM, Arnaud-Battandier F. Epidemiological study of malnutrition in elderly patients in acute, sub-acute and long-term care using the MNA. J Nutr Health Aging 1999; 3: 146-51.

14 Suominen M, Muurinen S, Routasalo P et al. Malnutrition and associated factors among aged residents in all nursing homes in Helsinki. Eur J Clin Nutr 2005; 59: 578-83.

15 Marshman R, Fisher MM, Coupland GAE. Nutritional status and postoperative complications in an Australian hospital. Aust NZ J Surg 1980; 50: 516-19.

16 Askew AR, Sui S, Green MK et al. Evaluation of the nutritional state of surgical patients in Brisbane. Aust NZ J Surg 1982; 52: 117-20.

17 Wood B, Kelleher D, Lo L et al. Nutritional status in hospital inpatients: implications for nutritional support services. Aust NZ J Med 1985; 15: 435-41.

18 Zador DA, Truswell AS. Nutritional status on admission to a general surgical ward in a Sydney Hospital. Aust NZ J Med 1987; 17: 234-40.

19 Ferguson M, Capra S, Bauer J, Banks M. Coding for malnutrition enhances reimbursement under casemix-based funding. Aust J Nutr Diet 1997; 54: 102-8.

20 Beck E, Patch C, Milosavljevic M et al. Implementation of malnutrition screening and assessment by dietitians: malnutrition exists in acute and rehabilitation settings. Aust J Nutr Diet 2001; 58: 92-7.

21 Middleton MH, Nazarenko G, Nivison-Smith I, Smerdely P. Prevalence of malnutrition and 12-month incidence of mortality in two Sydney teaching hospitals. Intern Med J 2001; 31: 455-61.

22 Lazarus C, Hamlyn J. Prevalence and documentation of malnutrition in hospitals: a case study in a large private hospital setting. Nutr Diet 2005; 62: 41-7.

23 Visvanathan R, Penhall R, Chapman I. Nutritional screening of older people in sub-acute care facility in Australia and its relation to discharge outcomes. Age Aging 2004; 33: 260-65.

24 Neumann S, Miller M, Daniels L, Crotty M. Nutritional status and clinical outcomes of older patients in rehabilitation. J Hum Nutr Dietet 2005; 18: 129-36.

25 Visvanathan R, Macintosh C, Callary M et al. The nutritional status of 250 older Australian recipients of domiciliary care services and its association with outcomes at 12 months. J Am Geriatr Soc 2003; 51: 1007-11.

26 Stratton R, Green CJ, Elia M. Disease-Related Malnutrition: An Evidence-Based Approach to Treatment. Wallingford: CABI Publishing, 2003.

27 Corish CA, Kennedy NP. Protein-energy malnutrition in hospital in-patients. Br J Nutr 2000; 83: 575-91.

28 Detsky AS, McLaughlin JR, Baker JP et al. What is Subjective Global Assessment of nutritional status? JPEN J Parenter Enteral Nutr 1987; 11: 8-13.

29 Detsky AS, Baker JP, O'Rourke K et al. Predicting nutrition-associated complications for patients undergoing gastrointestinal surgery. JPEN J Parenter Enteral Nutr 1987; 11: 440-46.

30 Duerksen DR, Yeo TA, Siemens JL, O'Connor MP. The validity and reproducibility of clinical assessment of nutritional status in the elderly. Nutrition 2000; 16: 740-44.

31 Christensson L, Unosson M, Ek A. Evaluation of nutritional assessment techniques in elderly people newly admitted to municipal care. Eur J Clin Nutr 2002; 56: 810-18.

32 Sacks GS, Dearman K, Replogle WH et al. Use of Subjective Global Assessment to identify nutrition-associated complications and death in geriatric long-term care facility residents. J Am Coll Nutr 2000; 19: 570-77.

33 Thoresen L, Fjeldstad I, Krogstad K et al. Nutritional status of patients with advanced cancer: the value of using the Subjective Global Assessment of nutritional status as a screening tool. Pall Med 2002; 16: 33-42.

34 Cooper BA, Bartlett LH, Aslani A et al Validity of Subjective Global Assessment as a nutritional marker in end-stage renal disease. Am J Kidney Dis 2001; 40: 126-32.

35 Sullivan DH, Sun S, Walls RC. Protein-energy undernutrition among elderly hospitalized patients: a prospective study. JAMA 1999; 281: 2013-19.

36 O'Flynn J, Peake H, Hickson M, Foster D, Frost G. The prevalence of malnutrition in hospitals can be reduced: results from three consecutive cross-sectional studies. Clin Nutr 2005; 24: 1078-88.

37 Brugler L, DiPrinzio MJ, Bernstein LH. The five-year evolution of a malnutrition treatment program in a community hospital. Jt Comm J Qual Improv 1999; 25: 191-206.

38 Correia MITD, Waitzberg DL. The impact of malnutrition on morbidity, mortality, length of hospital stay and costs evaluated through a multivariate model analysis. Clin Nur 2003; 22: 235-9.

39 Green CJ. Existence, causes and consequences of disease-related malnutrition in the hospital and the community, and clinical and financial benefits of nutritional intervention. Clin Nutr 1999; 18 (Suppl.): 3-28.

40 Covinsky KE, Martin GE, Beyth RJ, Justice AC, Sehgal AR, Langefeld CS. The relationship between clinical assessments of nutritional status and adverse outcomes in older hospitalized medical patients. J Am Geriatr Soc 1999; 47: 532-8.

41 Pennington CR. Disease-associated malnutrition in the year 2000. Postgrad Med J 1998; 74: 65-71.

Merrilyn BANKS, (1,2) Susan ASH, (1,2) Judy BAUER (3,4) and Deanne GASKILL (3)

(1) Department of Nutrition and Dietetics, Royal Brisbane and Women's Hospital, Herston, (2) Institute of Health and Biomedical Innovation and (3) School of Public Health, Queensland University of Technology, Kelvin Grove, and (4) Wesley Hospital and Wesley Research Institute, Toowong, Queensland, Australia

M. Banks, M Hlth Sc, APD, Director

S. Ash, PhD, FDAA, Associate Professor

J. Bauer, PhD, AdvAPD, Manager, Nutrition Services and Smart Start Research Fellow

D. Gaskill, M Nursing, Lecturer

Correspondence: M. Banks, Department of Nutrition and Dietetics, Royal Brisbane and Women's Hospital, Herston, Qld 4029, Australia. Email: [e-mail]

MB initiated the study, participated in data collection, obtained the data set and carried out the statistical analysis, interpreted and wrote the manuscript; SA participated in data collection and assisted in interpretation and writing of the manuscript; JB and DG assisted in interpretation and writing of the manuscript.

Table 1 Demographic variables for subjects in the nutrition assessment
subsample

                                       Acute
Variable                Audit 1 (2002)       Audit 2 (2003)

No. of facilities         8                    16
No. of subjects         774                  1434
% of QH population (a)   18.6                  35.4
Location, n (%)
  Metro                 527 (68.1)           1276 (89.0)
  Regional              172 (22.2)             52 (3.6)
  Rural/remote           75 (9.7)             106 (7.4)
Age in years,            66.5 [+ or -] 17.8    65.0 [+ or -] 18.8
  mean [+ or -] SD
Gender, n (%)
  Female                375 (48.4)            662 (46.2)
  Male                  399 (51.6)            771 (53.8)

                                    Residential
Variable                Audit 1 (2002)       Audit 2 (2003)

No. of facilities         5                    5
No. of subjects         381                  458
% of QH population (a)   25.2                 30.3
Location, n (%)
  Metro                 243 (63.8)           294 (64.2)
  Regional               45 (11.8)             0 (0.0)
  Rural/remote           93 (24.4)           164 (35.8)
Age in years,            78.9 [+ or -] 12.5   78.7 [+ or -] 12.4
  mean [+ or -] SD
Gender, n (%)
  Female                233 (61.2)           300 (65.5)
  Male                  148 (38.8)           158 (34.5)

(a) Average daily occupied beds 2002/2003.
QH = Queensland Health.

Table 2 Weighted average percentages of nutritional status categories
according to SGA across facilities at Audits 1 and 2

                   No. of
                   facilities
Facility type and  (No. of     Well                Moderately
audit number       subjects)   nourished           malnourished

                               Mean [+ or -] SD    Mean [+ or -] SD
                               (Min.-Max.)         (Min.-Max.)

Acute -- Audit 1    8 (774)     65.3 [+ or -] 4.0%  27.8 [+ or -] 4.3%
                               (61.9-77.4)         (12.9-32.1)
Acute -- Audit 2   16 (1434)    68.5 [+ or -] 9.5%  26.1 [+ or -] 8.3%
                               (25.0-100.0)         (0-62.5)

                               Median              Median
                               (Min.-Max.)         (Min.-Max.)

Residential aged    5 (381)     50.0%               41.6%
  care -- Audit 1              (43.4-88.9)          (8.9-54.7)
Residential aged    5 (458)     50.8%               35.0%
  care -- Audit 2              (46.3-85.7)         (13.0-38.9)

Facility type and  Severely           Total
audit number       malnourished       malnourished

                   Mean [+ or -] SD   Mean [+ or -] SD
                   (Min.-Max.)        (Min.-Max.)

Acute -- Audit 1    7.0 [+ or -] 2.3%  34.7 [+ or -] 4.0%
                   (0-10.0)           (22.6-38.1)
Acute -- Audit 2    5.3 [+ or -] 3.6%  31.4 [+ or -] 9.5%
                   (0.0-13.3)          (0.0-75.0)

                   Median             Median
                   (Min.-Max.)        (Min.-Max.)

Residential aged    8.4%               50.0%
  care -- Audit 1  (1.9-25.8)         (11.1-56.6)
Residential aged   14.2%               49.2%
  care -- Audit 2  (1.3-22.2)         (13.9-53.7)

SGA = Subjective Global Assessment.

Table 3 Multivariable relationships of variables on malnutrition for
public facilities in Queensland

                             Acute facilities (Audits 1 and 2 combined)
                             No. of    Malnourished  Adjusted OR
Variables                    subjects  n (%)         (95% CI)

Gender                       2207 (b)
  Female                     1037      355 (34.2)    1.0 (c)
  Male                       1170      365 (31.2)    1.1 (0.9-1.2)
Age groups (years)           2199 (b)
  [less than or equal to]40   273       59 (21.6)    0.6 (0.4-0.8)
  41-60                       476      106 (22.3)    0.6 (0.5-0.7)
  61-80                       942      308 (32.7)    1.0 (c)
  >80                         508      244 (48.0)    1.7 (1.5-2.0)
Facility location            2208
  Metropolitan               1803      610 (33.8)    1.0 (c)
  Regional                    224       74 (33.0)    0.4 (0.2-1.2)
  Rural/remote                181       36 (19.9)    0.1 (0.02-0.5)
Specialty                    2208
  Medical                     941      285 (30.3)    1.0 (c)
  Surgical                    701      206 (29.4)    1.2 (0.9-1.8)
  Oncology                    107       60 (56.1)    2.3 (1.4-3.8)
  Critical care                62       20 (32.3)    1.6 (1.1-2.3)
  Rehabilitation              227       74 (32.6)    1.0 (0.6-1.7)
  Aged care                   170       75 (44.1)    1.8 (0.4-7.6)

                             Acute facilities (Audits 1 and 2 combined)
                                      [chi square] (df) (a)
Variables                    P-value  P-value

Gender                                 [0.8.sub.(1)]
  Female                              P = 0.38
  Male                        0.38
Age groups (years)                    [64.6.sub.(3)]
  [less than or equal to]40  <0.001   P < 0.001
  41-60                      <0.001
  61-80
  >80                        <0.001
Facility location                      [8.1.sub.(2)]
  Metropolitan                        P = 0.02
  Regional                    0.10
  Rural/remote                0.007
Specialty                             [19.7.sub.(5)]
  Medical                             P = 0.001
  Surgical                    0.26
  Oncology                    0.001
  Critical care               0.02
  Rehabilitation              0.98
  Aged care                   0.45

                             Residential facilities (Audit 1 data only)
                             No. of    Malnourished  Adjusted OR
Variables                    subjects  n (%)         (95% CI)

Gender                       381 (b)
  Female                     233        98 (42.1)    1.0 (c)
  Male                       148        66 (44.6)    1.2 (1.1-1.3)
Age groups (years)           381
  [less than or equal to]40    7         2 (38.6)    0.4 (0.04-3.7)
  41-60                       29         8 (27.6)    0.6 (0.3-1.3)
  61-80                      146        51 (34.9)    1.0 (c)
  >80                        199       103 (51.8)    1.4 (1.2-1.6)
Facility location            381
  Metropolitan               243       125 (51.4)    1.0 (c)
  Regional                    45         5 (11.1)    0.1 (0.01-0.8)
  Rural/remote                94        34 (36.6)    0.4 (0.1-2.3)
Specialty
  Medical
  Surgical
  Oncology
  Critical care
  Rehabilitation
  Aged care

                             Residential facilities (Audit 1 data only)
                                      [chi square] (df) (a)
Variables                    P-value  P-value

Gender                                 [9.5.sub.(1)]
  Female                      0.003   P = 0.002
  Male
Age groups (years)                    [23.6.sub.(3)]
  [less than or equal to]40   0.4     P < 0.001
  41-60                       0.19
  61-80
  >80                        <0.001
Facility location                      [5.1.sub.(2)]
  Metropolitan                        P = 0.08
  Regional                    0.03
  Rural/remote                0.31
Specialty
  Medical
  Surgical
  Oncology
  Critical care
  Rehabilitation
  Aged care

(a) Wald chi-square for overall model.
(b) Data missing.
(c) Referent.
Statistical method: logistic regression.
OR = odds ratio.