Nutritional screening and assessment tools for use by nurses: literature review

Abstract

This paper reports a literature review describing the range of published tools available for use by nurses to screen or assess nutritional status of patients/clients, and to examine whether the validity, reliability, sensitivity, specificity and acceptability of the tool have been investigated. Tools developed specifically for use with older adults are not considered in this review. A screening/assessment tool is described as a tool that uses a questionnaire-type format, contains more than one risk factor for malnutrition and gives an assessment of risk. The incidence of malnutrition in people cared for by nurses is high and screening or assessment tools are often used to identify those with, or at risk of, malnutrition. A comprehensive literature review methodology was employed. A range of electronic databases was searched from 1982 to 2002. Search terms incorporating "nutrition", "screening", assessment, feeding, instrument, tool, validity, reliability, sensitivity and specificity were combined. Manual searches were also conducted. Seventy-one nutritional screening/assessment tools were identified of which 35 were reviewed. Tools not reviewed included those which incorporated significant use of biochemical measures (8), included complex anthropometric measures (3), were concerned specifically with ingestion (4) or were designated for use with an older population (21). The tools reviewed use a wide variety of risk factors for malnutrition, varying from anthropometric measurements to socio-cultural aspects of eating behaviour. Some identify an action plan based on the score obtained. Many have not been subjected to validity and/or reliability testing and yet appear to be in use in clinical practice. The sensitivity, specificity and acceptability of the tools are often not investigated. There are many published nutritional screening/assessment tools available for use by nurses to screen or assess the nutritional status of patients/clients. Many have not been subject to rigorous testing. Future work should consider a more standardized approach to the use of these tools.