Background
Undernutrition is associated with increased hospital length of stay (LOS). Anthropometrics is required for undernutrition screening and diagnosis. However, the measurements which are more strongly associated with LOS are to be specified. This study aims to measure the independent association of weight, triceps skinfold thickness (TST), mid‐upper arm circumference (MUAC), adductor pollicis muscle thickness (APMT), mid‐upper arm muscle area (AMA), mid‐upper arm muscle circumference (AMC), and body mass index (BMI) with LOS and investigating whether these indicators had similar validity in predicting LOS.
Methods
Six hundred ninety‐five patients participated in a prospective observational study. Participants were dichotomized indicators studied according to the 5th and 25th percentiles described in the literature (for APMT only the 5th percentile was used). Cox regression analysis was used to estimate adjusted hazard ratios (HRs) and corresponding 95% CIs.
Results
Patients that presented values <5th percentile for TST (HR, 0.759; 95% CI, 0.579–0.995), MUAC (HR, 0.822; 95% CI, 0.687–0.983), APMT (HR, 0.791; 95% CI, 0.671–0.933), AMA (HR, 0.797, 95% CI, 0.660–0.962), and AMC (HR, 0.746; 95% CI, 0.611–0.911) showed a lower probability of being discharged to usual residence. Patients whose TST values were <25th percentile also presented lower probability of being discharged to usual residence (HR, 0.798; 95% CI, 0.673–0.946). No associations were found between weight and BMI with LOS.
Conclusion
Depletion of TST, MUAC, APMT, AMA, and AMC was associated with lower probability of being discharged to usual residence. A thicker TST was independently associated with this outcome and thus TST should be considered for undernutrition screening and diagnosis.