A supervised pattern analysis of the length of stay for hip replacement admissions

Dimitrios Zikos, Ashara Shrestha, Taylor Colotti, Leonidas Fegaras

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

Hip replacement is the most common surgical procedure among Medicare patients in the US and worldwide. The hospital length of stay (LOS) for hip replacement admissions is therefore important to be controlled, contributing to savings for hospitals. This study combined medical claims and hospital structure and service data to examine LOS fluctuations and trends, and admission distribution patterns, during weekdays, for hip replacement cases. The study furthermore examined associations of these patterns with the LOS performance. Most hospitals were found to admit hip replacement cases at the start of the week (Monday through Wednesday). There is an upward LOS trend as we approach late weekday admissions. Multiple linear regression analysis showed that LOS weekday inconsistencies, a large proportion of hip replacement admissions on Thursday and Friday, the government ownership status, the bed size, and the critical access status are associated with an increased LOS. On the other hand, the rate of hip replacement admissions over total ones, and the hospital being accredited, are associated with a lower LOS. Findings stress out the need for hospitals to maintain an effective and balanced distribution of hip replacement admissions, evenly during the week, and the need for standardized case management, to avoid practice variability and, therefore, LOS fluctuations for their hip replacement cases.

Original languageEnglish
Article number58
JournalHealthcare (Switzerland)
Volume7
Issue number2
DOIs
StatePublished - Jun 2019

Keywords

  • Hip replacement
  • Hospital
  • Length of stay
  • Medicare
  • Pattern analysis

Fingerprint

Dive into the research topics of 'A supervised pattern analysis of the length of stay for hip replacement admissions'. Together they form a unique fingerprint.

Cite this