Considerations for screening tool selection and role of predictive and concurrent validity.

PURPOSE OF REVIEW: Nutrition screening tool selection can be difficult. This review critically examines the relevance of validity, specifically concurrent (agreement between tools) and predictive validity (prediction of outcomes), which have been the focus of several recent studies. An operational framework for screening tool selection is provided to contextualise the findings. RECENT
FINDINGS: Studies of predictive and concurrent validity involving screening tools comprising a variable number of nutritional and non-nutritional items (some nonmodifiable) have yielded inconsistent results. The use of one tool as a gold standard to judge the relative merits of other tools can be misleading because there is no agreed gold standard and different tools were designed for diagnostic, prognostic or other purposes. The use of observed outcomes (without nutritional intervention) as the gold standard may not adequately reflect the value of tools designed to assess nutritional status and need for nutritional intervention.
SUMMARY: Over-reliance on concurrent and predictive validity can be confusing and even counter-productive if used inappropriately. A proposed framework for screening tool selection indicates many factors should be considered so that there is purpose and harmony between the screening tool and the screening programme.