PREFERENCE ELICITATION TO AID DECISION-MAKING FOR TREATMENT OF A FIRST TIME ANTERIOR SHOULDER DISLOCATION

Purpose: To elicit individual-level preferences for factors relevant to choosing surgical versus non-surgical management of first-time anterior shoulder dislocation (FTASD) and to evaluate whether preferences were associated with treatment choice.      

Method: We designed and pilot-tested a survey instrument using Sawtooth Software that included an adaptive conjoint analysis (ACA), inclusive of the following attributes: (1) chance of recurrent dislocation (5%, 20%, and 80%); (2) out-of-pocket cost ($0, $1,000, $2,000); (3) short-term limits on shoulder motion (none, cannot lift arm above shoulder, arm in sling); (4) limits on participation in high-risk activities (1 month, 3 months, 1 year); and (5) duration of physical therapy (4, 8 and 12 weeks). Following the ACA exercise, participants were given the option of surgical or non-surgical treatment for a hypothetical shoulder dislocation wherein the risk of recurrent dislocation was personalized according to the participant’s gender and age. We administered the survey to an online panel in February 2014. 

Result:  Overall, 374 (75%) participants met inclusion criteria and satisfied validation checks. Most participants were male (61%), less than 30 years old (68%) and reported to participate in high-risk activities at least weekly (76%). Nearly 1 in 3 (32%) reported prior shoulder dislocation. On a scale from 0 to 100, mean (SD) importance weights were 27.9 (8.9) for chance of recurrent dislocation, 24.6 (7.8) for out-of-pocket cost, 18.7 (5.2) for limits on high-risk activities, 15.2 (4.4) on limits on shoulder motion, and 13.5 (4.2) on duration of physical therapy.  Among the importance weights, only chance of recurrent dislocation was significantly predictive of respondents’ choice of surgery at $1,000 (versus no surgery) when adjusting for age and gender. Nearly 9 out of 10 (87%) agreed or strongly agreed that their individual-level importance weights matched their expectations, and 89% indicated that they would share the information with their physician. 

Conclusion: Given the levels specified for the five attributes, an individual’s level of concern about risk of recurrence was found to be the most important and was predictive of an individual opting for surgical management. Participants’ acceptance of and their desire to share results with their providers warrant further study of a preference-based tool in patients with FTASD to evaluate its utility in clinical care.