PREOPERATIVE PREDICTORS OF INPATIENT STAY FOLLOWING TOTAL HIP REPLACEMENT: A PROSPECTIVE MULTIVARIATE ANALYSIS OF 2302 PATIENTS

Abstract

Whilst clinical pathways have reduced the length of inpatient stay (LOS) and costs associated with total hip replacement (THR), it would be useful to be able to predict which factors affect the LOS following THR. In this way units could improve patient flow, and thus increase efficiency, by optimising case loads and staffing levels in all related departments. We hypothesised that there were pre-operative factors that will predict the length of inpatient stay following a unilateral primary THR.

Prospective data from 2302 patients who underwent primary unilateral THR for osteoarthritis during a nine-year period from January 01 1998 were included. The relationship between each prospectively recorded factor and LOS was analyzed separately using t-tests or Pearson correlation. Multiple linear regression was used to analyze the effect of each significant factor adjusted for others.

Length of stay varied from 3 to 58 days, with a mean of 8.1 and a median of 7.0. The pre-operative factors that were not significantly associated with the LOS were BMI, Scottish Index of Multiple Deprivation, disabling knee, back or contra-lateral hip pain. A number of pre-operative factors were found to be highly significant predictors of LOS when subjected to univariate analysis, but not significant after multivariate analysis. These were smoking, heart disease, diabetes, pre-operative aspirin, pre-operative haemoglobin and overall Short Form (SF-36) score. Pre-operative factors that were significantly associated with LOS after adjusting for the effect of others were age, sex, pre-operative NSAIDs, consultant surgeon, combined function/activity dimension to Harris hip score, general health perception (GHP) dimension of SF-36, the day of the week and the year of surgery.

Younger age, male sex, higher combined Harris hip function and activity score, higher GHP dimension of SF-36 score and NSAID use are all significantly associated with reduced length of inpatient stay following primary THR for osteoarthritis.