Abstract
We determined the association between frailty and 30-day mortality following total hip arthroplasty (THA) and the impact of THA on 30-day mortality compared to a control population.
We used primary care data (Clinical Practice Research Datalink), linked secondary care data (Hospital Episode Statistics) and Office for National Statistics mortality data. Frailty was assessed using a validated frailty index based on coded data in the primary care record and categorised as fit, mild, moderate, and severe frailty. The association between frailty and 30-day mortality following THA due to osteoarthritis was assessed using Cox regression, adjusted for year of birth, sex, quintile of index of multiple deprivation and year of surgery. Mortality following THA was also compared to a control population who had osteoarthritis but no previous THA, matched on year of birth, sex, and quintile of index of multiple deprivation.
103,563 cases who had a THA and their matched controls contributed data. Among those who had THA, compared to fit participants, 30-day mortality increased with increasing frailty; adjusted hazard ratio (HR) (95% CI) among mild frailty, 0.87 (0.66, 1.15); moderate frailty 1.73 (1.26, 2.38); and severe frailty, 2.85 (1.84, 4.39).
Compared to fit controls who did not have surgery, 30-day mortality was higher among fit people who had THA, adjusted HR 1.60 (1.15, 2.21). There was, however, no statistically significant difference in 30-day mortality among cases with mild, moderate and severe frailty compared to controls in the same frailty category.
Among people who had THA, 30-day mortality increased with increasing frailty. While 30-day mortality was increased among fit individuals who had THA compared to fit controls who did not have surgery, there did not appear to be increased mortality among individuals with mild, moderate or severe frailty compared to controls in the same frailty category. A healthy surgery (selection) effect may have impacted on the comparison of mortality among cases and controls.
