3040 — Race and Joint Replacement Recommendation: The Role of Preference and Doctor-Patient Communication

Hausmann L (VA Pittsburgh Healthcare System Center for Health Equity Research and Promotion (VAPHS CHERP)), Hanusa BH (VAPHS CHERP), Kresevic D (Louis Stokes DVA Medical Center, University Hospitals Case Medical Center), Mor M (VAPHS CHERP, University of Pittsburgh), Kwoh CK (VAPHS CHERP, University of Pittsburgh), Gordon HS (Jesse Brown VA Medical Center, University of Illinois at Chicago), Geng M (VAPHS CHERP), Ibrahim SA (VAPHS CHERP, University of Pittsburgh)

Objectives:
There is a marked racial disparity in utilization of joint replacement (JR) in the management of knee/hip osteoarthritis (OA). Whether patient preferences and doctor-patient communication in an orthopedic setting contribute to this disparity remains unclear. We examined the impact of patient race, patient preference, and doctor-patient communication on the likelihood of receiving a recommendation for knee/hip JR during an orthopedic consultation.

Methods:
The sample included African American (AA) and white veterans who were potential candidates for JR and were referred to Pittsburgh and Cleveland VA orthopedic clinics. Prior to a clinic visit, we assessed patient demographics, severity of OA, and preferences regarding JR (i.e., willingness to undergo JR). We then audio-recorded the visit and coded the recordings using the Roter Interaction Analysis System, which assesses qualities of communication such as patient-centeredness, physician verbal dominance, and effect. After first examining the relationship between patient race and JR recommendation using logistic regression adjusting for study site, changes in JR recommendation rates over time, and clustering by provider, we used backwards stepwise selection to identify additional factors significantly associated with the likelihood of receiving a recommendation for JR.

Results:
Analyses included visits of 278 white and 103 AA patients with 69 orthopedic doctors. After adjusting for site and changes over time, AA patients were significantly less likely than whites to receive a recommendation for JR (OR = .40, 95% CI = .22-.75). In the final adjusted model, however, this difference was not significant (OR = .60, 95% CI = .30-1.19). Patient preference was the strongest predictor of recommendation for JR, such that patients who were unsure, probably not willing, or definitely not willing to undergo JR were less likely to receive a JR recommendation (OR = .07, 95% CI = .01-.57). Other significant predictors of JR recommendations included positive doctor affect (OR = 1.14, 95% CI = 1.03-1.28), age greater than 70 (OR = 2.30, 95% CI = 1.06-4.99), and severity of OA (OR = 5.70, 95% CI = 1.77-18.38).

Implications:
In this VA sample, the relationship between race and being recommended for joint replacement is primarily a function of patient preference.

Impacts:
Given previous evidence showing racial variation in patient preferences for elective treatments, understanding factors that shape patient preferences and how preferences get communicated may help reduce healthcare disparities.