The need for consistent criteria for identifying malnutrition.

The lack of consistent criteria for diagnosing malnutrition and protein-energy malnutrition (PEM) creates problems in educating medical students and physicians, setting the parameters for observational and controlled clinical trials, and formulating clinical guidelines. There is no validated formal definition of malnutrition (or PEM), and the tools that have been developed to screen for it, or diagnose it, vary in their agreement. I make the following suggestions. First, avoid unqualified use of the term 'malnutrition', as it is ambiguous. Second, carefully distinguish between screening and diagnosis, which have different aims and implications. Third, consider the notion that in medicine the diagnosis of PEM is reached by 'narrative-interpretive' reasoning, which regards the disease as a pathophysiological entity in a specific clinical context. I recommend that the concept of PEM as a disease (not a score) be imbedded in teaching and the practice of medicine, and in the design of clinical trials and the setting of guidelines. Fourth, disagreements in screening-derived risk scores and uncertainty in diagnosis are difficult to avoid, but only in the grey zone. It would be prudent, at least until the greater medical world considers the nutritional paradigm plausible enough to invest in it, to enroll only patients who have unambiguously diagnosed PEM in prospective trials with hard clinical endpoints.