Survival following total hip replacement

Abstract

Background: Several studies have shown that patients who undergo total hip replacement live longer than control subjects, but the reason for this apparent protective effect is not clear. The purpose of our investigation was to assess whether the association had the characteristics of a causal relationship or whether it might appear to be due to comorbidity or other factors.

Methods: We compared survival over a six-year period for 28,469 Medicare enrollees who had elective primary total hip replacement in 1996 and a control group from the general Medicare population, matched 5:1 on year of birth, sex, race, and whether the Medicaid program paid the Medicare premium (a proxy for low income). For both the patients treated with total hip replacement and the control subjects, we used Medicare hospital claims to ascertain comorbidity, that is, whether the patient had had any of sixteen serious discharge diagnoses in the year prior to the total hip replacement (or an equivalent date for the controls). The survival patterns for the total hip arthroplasty and control cohorts were compared with use of proportional hazards regression for three follow-up periods: the first ninety days after surgery (or an equivalent date for the controls), ninety days to five years after surgery, and more than five years after surgery.

Results: The total hip replacement cohort had less comorbidity than the controls, with an approximately 30% lower prevalence for most serious diseases. Sex, age, Medicaid eligibility, and earlier serious comorbid diagnoses were important predictors of survival, but, even in analyses that adjusted for these variables, the total hip replacement cohort had better long-term survival than the controls. The patients who had a total hip replacement had a higher rate of mortality than the controls immediately after surgery, but, by three months postoperatively, the mortality rate for the patients was well below that for the controls. The overall rate of survival during the first ninety days was comparable for the two groups. From three months to five years after surgery, the mortality rate for the patients who had a total hip replacement was only two-thirds of that for the controls. After five years, the mortality rates for the two groups appeared to be converging.

Conclusions: Total hip replacement recipients survive longer than do matched controls in the Medicare population. The very rapid emergence of the lower mortality rate suggests that it is due to the selection of low-risk patients for elective surgery. However, our admittedly crude adjustment for comorbidity did not diminish the protective effect, so some effect of the procedure itself cannot be ruled out.