[Validity and reliability of a nutrition screening tool in patients admitted to hospital]

ABSTRACT

Malnutrition is an important health problem in hospitals and its association with increased morbidity and mortality has been well-documented. Since the condition can be readily treated, often at low cost, it is logical that efforts should be made to identify malnutrition for timely nutritional interventions. A nutrition screening tool [a questionnaire] was designed on the basis of recommendations of the British Association for Parenteral and Enteral Nutrition and used to identify individuals who are malnourished, at-risk of malnutrition, or may need nutritional support. The aim of this study was to evaluate the validity and reliability of this tool in a hospital in Tehran, Iran. A total of 446 patients, aged over 18 years, admitted consecutively to different wards of Taleghani Hospital in Tehran, Iran during six months were studied in the first 24 hours of admission. For testing validity of the questionnaire used, first a nutritionist made a full nutritional assessment of all the patients. The assessment included the following items a. measurements of weight, height, body mass index [BMI], triceps skinfold [TSF], and mid-arm circumference [MAC]; b. weight history and percentage of involuntary weight loss over the previous 3-6 months; a 24-hour dietary recall; c. gastro-intestinal symptoms. The results of the assessment were, then, compared with the results obtained from the screening questionnaire. The inter-rater reliability of the screening tool was assessed by two nurses who completed the screening questionnaire separately during 72 hours following admission of the patients. The agreement between results was calculated. There was 66% agreement between the nutritionist's assessment results and the nurses's assessment results using the nutritional screening tool. The sensitivity and specificity of the screeing tool were 83.7% [95%CI 80.0-87.4%] and 84.7% [95%CI 81.1-88.1%], respectively. Further analysis of the data showed the positive predictive value to be 92.3% [95%CI 89.6-95./0%] and the negative predictive value to be 70.2% [95%CI 65.8-74.8%]. The mean inter-rater reliability [K=0.51] was moderate. The screening tool, completed separately for each patient by the two nurses, is reliable and, in comparison with the more objective nutritional assessment, has an acceptable validity. The questionnaire is useful for identifying malnourished patients in a hospital setting: