A protein energy malnutrition scale (PEMS).

Most assessments of protein energy malnutrition have relied on clinical impressions, laboratory tests, or anthropometric measurements. There has been a split in the field between those who favor clinical versus laboratory approaches, and each alone misclassifies about 20% of the patients. A 23-item scale is described that provides four subscores derived from anthropometric measurements, clinical history, physical examination, and laboratory tests, as well as a total score. Items rated on logarithmic scales relate to degree of deviation from standards. Reliability was tested by having two physicians rate the same 25 patients and items with better than r = 0.66 retained. Validating information came from the finding that the scale discriminated before surgery (p less than 0.001) between patients who did and did not develop postoperative complications. Use of the scale appears to be a good approach for assessing degree of malnutrition as well as for monitoring changes that occur over time.