Predicting Outcome after Total Hip Arthroplasty: The Role of Preoperative Patient-Reported Measures

Choosing the appropriate patient for surgery is crucial for good outcome in total hip arthroplasty (THA). Therefore, parameters predicting outcome preoperatively are of major interest. In the current study, we compared the predictive power of different presurgical measures in minimally invasive THA. In the course of a prospective clinical trial preoperative HOOS, EQ-5D and SF-36 were obtained in 140 patients undergoing THA. Responder rate was defined by the modified OMERACT-OARSI criteria at six-month-, one-year, two-year, and three-year follow-up. Logistic regression was performed to compare the different questionnaires regarding their power of predicting positive responders. ROC-curve analysis was used to define benchmarks in preoperative measures associated with good outcome. Preoperative HOOS (p<0.001), EQ-5D (p=0.007), and PCS of SF-36 (p<0.001) were higher in responders than in nonresponders whereas no differences between responders and nonresponders were found for preoperative MCS (p=0.96) of SF-36. However, preoperative HOOS revealed best predictive power (OR=0.84 95%CI=0.78-0.90, p<0.001, Pseudo R-Squared according to Nagelkerke=0.48, effect size according to Cohen=0.96) compared to all other preoperative measures. Multivariable analysis confirmed preoperative HOOS as an independent parameter correlating with postoperative responder status (OR=0.76, 95% CI=0.66-0.88, p<0.001). In ROC-curve analysis nonresponders were identified with a sensitivity of 91.7% and specificity of 68.9% using a cutoff in preoperative HOOS of 40.3. Presurgical HOOS can predict outcome in THA better than other preoperative outcome measures. Patients with a preoperative HOOS value less than 40.3 have the highest probability of a positive response in terms of pain and function after THA.