Background Elderly people are particularly at risk of malnutrition. There is no consensus regarding the optimal malnutrition screening test for elderly people and little is known about the prevalence of malnutrition in elderly people living in sheltered housing.Method An observational study comparing sensitivity, specificity and positive and negative predictive values of the following screening measures in elderly people living in sheltered accommodation: body mass index, mid-arm circumference, albumin, haemoglobin, lymphocyte count, cholesterol and the Malnutrition Universal Screening Tool (MUST) and Mini Nutritional Assessment (MNA). A dietitian assessment was used as the gold standard to establish whether there was a risk of malnutrition. ResultsOf 100 people recruited (31 male and 69 female with average age 79.3 years) ten were categorized at risk by the dietitian assessment. The MUST score was the most sensitive and specific screening measure (100% and 98% respectively) with a negative predictive value of 1. The sensitivity and specificity of the other measures were: MNA 80% and 90%, mid-arm circumference 70% and 99%, BMI 60% and 90%, albumin 30% and 77%, haemoglobin 50% and 61%, lymphocyte count 20% and 86%, low cholesterol 30% and 90%.Conclusions Ten per cent of elderly people in sheltered housing are at risk of malnutrition. The MUST screening tool is a sensitive and specific method of identifying those requiring further nutritional assessment. Conflict of interests, source of funding and authorshipThe authors declare that they have no conflict of interests. This study was conducted with assistance of a financial research grant from the British Geriatric Society. We are grateful for the help of Mr Joe Onwunde (Consultant Gynaecologist and Medical Statistician) for his assistance with the data statistical analysis and interpretation. DGH and NYH developed the initial research question. All authors were responsible for development of the research protocol. DGH and CD carried out the research visits and data collection. DGH, CD and NYH were responsible for analysis and interpretation of the data. All authors were involved with the drafting of the research paper, have critically reviewed its content and have approved the final version submitted for publication.PPV, positive predictive value; NPV, negative predictive value; BMI, body mass index; MUST, Malnutrition Universal Screening Tool; MNA, Mini Nutritional Assessment.

No abstract