Patient Decision Aids Improve Decision Quality and Patient Experience and Reduce Surgical Rates in Routine Orthopaedic Care: A Prospective Cohort Study.

BACKGROUND: Patient decision aids are effective in randomized controlled trials, yet little is known about their impact in routine care. The purpose of this study was to examine whether decision aids increase shared decision-making when used in routine care.
METHODS: A prospective study was designed to evaluate the impact of a quality improvement project to increase the use of decision aids for patients with hip or knee osteoarthritis, lumbar disc herniation, or lumbar spinal stenosis. A usual care cohort was enrolled before the quality improvement project and an intervention cohort was enrolled after the project. Participants were surveyed 1 week after a specialist visit, and surgical status was collected at 6 months. Regression analyses adjusted for clustering of patients within clinicians and examined the impact on knowledge, patient reports of shared decision-making in the visit, and surgical rates. With 550 surveys, the study had 80% to 90% power to detect a difference in these key outcomes.
RESULTS: The response rates to the 1-week survey were 70.6% (324 of 459) for the usual care cohort and 70.2% (328 of 467) for the intervention cohort. There was no significant difference (p > 0.05) in any patient characteristic between the 2 cohorts. More patients received decision aids in the intervention cohort at 63.6% compared with the usual care cohort at 27.3% (p = 0.007). Decision aid use was associated with higher knowledge scores, with a mean difference of 18.7 points (95% confidence interval [CI], 11.4 to 26.1 points; p < 0.001) for the usual care cohort and 15.3 points (95% CI, 7.5 to 23.0 points; p = 0.002) for the intervention cohort. Patients reported more shared decision-making (p = 0.009) in the visit with their surgeon in the intervention cohort, with a mean Shared Decision-Making Process score (and standard deviation) of 66.9 ± 27.5 points, compared with the usual care cohort at 62.5 ± 28.6 points. The majority of patients received their preferred treatment, and this did not differ by cohort or decision aid use. Surgical rates were lower in the intervention cohort for those who received the decision aids at 42.3% compared with 58.8% for those who did not receive decision aids (p = 0.023) and in the usual care cohort at 44.3% for those who received decision aids compared with 55.7% for those who did not receive them (p = 0.45).
CONCLUSIONS: The quality improvement project successfully integrated patient decision aids into a busy orthopaedic clinic. When used in routine care, decision aids are associated with increased knowledge, more shared decision-making, and lower surgical rates. CLINICAL RELEVANCE: There is increasing pressure to design systems of care that inform and involve patients in decisions about elective surgery. In this study, the authors found that patient decision aids, when used as part of routine orthopaedic care, were associated with increased knowledge, more shared decision-making, higher patient experience ratings, and lower surgical rates.