Estimation of 24-hour urinary creatinine excretion from patient variables: A novel approach to identify patients with low muscle mass and malnutrition and relationship to outcomes.

Background: Low muscle mass has been correlated with adverse outcomes in patients who are critically ill. Methods to identify low muscularity such as computed tomography scans or bioelectrical impedance analyses are impractical for admission screening. Urinary creatinine excretion (UCE) and creatinine height index (CHI) are associated with muscularity and outcomes but require a 24-h urine collection. The estimation of UCE from patient variables avoids the need for a 24-h urine collection and may be clinically useful.

Methods: Variables of age, height, weight, sex, plasma creatinine, blood urea nitrogen (BUN), glucose, sodium, potassium, chloride, and carbon dioxide from a deidentified data set of 967 patients who had UCE measured were used to develop models to predict UCE. The model identified with the best predictive ability was validated and then retrospectively applied to a separate sample of 120 veterans who were critically ill to examine if UCE and CHI predicted malnutrition or were associated with outcomes.

Results: A model was identified that included variables of plasma creatinine, BUN, age, and weight and was found to be highly correlated, moderately predictive of UCE, and statistically significant. Patients with model-estimated CHI   60% had significantly lower body weight, body mass index, plasma creatinine, and sera albumin and prealbumin levels; were 8.0 times more likely to be diagnosed with malnutrition; and were 2.6 times more likely to be readmitted in 6 months.

Conclusion: A model that predicts UCE offers a novel method to identify patients with low muscularity and malnutrition on admission without the use of invasive tests.