Andy H Lee1, Wing K Fung, Bo Fu. 1. School of Public Health, Curtin University of Technology, Perth, WA, Australia. Andy.Lee@curtin.edu.au
Abstract
BACKGROUND: Length of stay (LOS) is an important measure of hospital activity and health care utilization, but its empirical distribution is often positively skewed. OBJECTIVE: This study reviews the mean and median regression approaches for analyzing LOS, which have implications for service planning, resource allocation, and bed utilization. METHODS: The two approaches are applied to analyze hospital discharge data on cesarean delivery. Both models adjust for patient and health-related characteristics, and for the dependency of LOS outcomes nested within hospitals. The estimation methods are also compared in a simulation study. RESULTS: For the empirical application, the mean regression results are somewhat sensitive to the magnitude of trimming chosen. The identified factors from median regression, namely number of diagnoses, number of procedures, and payment classification, are robust to high-LOS outliers. The simulation experiment shows that median regression can outperform mean regression even when the response variable is moderately positively skewed. CONCLUSION: Median regression appears to be a suitable alternative to analyze the clustered and positively skewed LOS, without transforming and trimming the data arbitrarily.
BACKGROUND: Length of stay (LOS) is an important measure of hospital activity and health care utilization, but its empirical distribution is often positively skewed. OBJECTIVE: This study reviews the mean and median regression approaches for analyzing LOS, which have implications for service planning, resource allocation, and bed utilization. METHODS: The two approaches are applied to analyze hospital discharge data on cesarean delivery. Both models adjust for patient and health-related characteristics, and for the dependency of LOS outcomes nested within hospitals. The estimation methods are also compared in a simulation study. RESULTS: For the empirical application, the mean regression results are somewhat sensitive to the magnitude of trimming chosen. The identified factors from median regression, namely number of diagnoses, number of procedures, and payment classification, are robust to high-LOS outliers. The simulation experiment shows that median regression can outperform mean regression even when the response variable is moderately positively skewed. CONCLUSION: Median regression appears to be a suitable alternative to analyze the clustered and positively skewed LOS, without transforming and trimming the data arbitrarily.
Authors: D Steffling; M Ritzka; W Jakob; A Steinbrecher; S Schwab-Malek; B Kaiser; P Hau; S Boy; K Fuchs; U Bogdahn; F Schlachetzki Journal: Nervenarzt Date: 2012-06 Impact factor: 1.214
Authors: Mitesh S Patel; Neha Patel; Dylan S Small; Roy Rosin; Jeffrey I Rohrbach; Nathaniel Stromberg; C William Hanson; David A Asch Journal: J Gen Intern Med Date: 2016-03-25 Impact factor: 5.128
Authors: Nasir Wabe; Ling Li; Robert Lindeman; Ruth Yimsung; Maria R Dahm; Susan McLennan; Kate Clezy; Johanna I Westbrook; Andrew Georgiou Journal: J Clin Microbiol Date: 2019-03-28 Impact factor: 5.948
Authors: Saam Morshed; Theodore Miclau; Oliver Bembom; Mitchell Cohen; M Margaret Knudson; John M Colford Journal: J Bone Joint Surg Am Date: 2009-01 Impact factor: 5.284
Authors: Katherine R Courtright; Erich M Dress; Jaspal Singh; Brian A Bayes; Marzana Chowdhury; Dylan S Small; Timothy Hetherington; Lindsay Plickert; Michael E Detsky; Jason N Doctor; Michael O Harhay; Henry L Burke; Michael B Green; Toan Huynh; D Matthew Sullivan; Scott D Halpern Journal: Ann Am Thorac Soc Date: 2021-02