EP545 - A Pilot Cluster Randomized Trial of a Diverticulitis Decision Aid

Image or Table (Do Not add text here):

Purpose/Background: Shared decision-making is crucial in choosing the ideal treatment for recurrent uncomplicated diverticulitis as it involves balancing medical management and surgical interventions in the context of patient preferences. Shared decision-making tools in the form of decision aids are valuable in the face of preference-sensitive decisions as they help patients and providers navigate the complexities of care options available. Our group previously developed an online diverticulitis decision aid to facilitate effective surgeon-patient communication. We performed a pilot trial to describe the feasibility of this decision aid and to generate effect sizes to inform a larger randomized trial.

Methods/Interventions: In a pilot cluster randomized study, 31 patients were randomized at the provider level to receive the decision aid or proceed with standard care. Feasibility outcomes such as recruitment timing, enrollment cadence, and participant retention were collected. Pre-consultation questionnaires captured demographic, disease severity, and baseline decisional conflict levels. After review of the decision aid and subsequent clinician consultation, patients completed surveys immediately after their clinic appointment and at three months to capture quality of shared decision-making, change in decisional conflict levels, decision regret, and fidelity of decision aid exposure.

Results/Outcomes: 2137 colorectal surgery clinic patients were screened over a 6-month period, of which 101 (4.7%) were seen for a diagnosis of diverticulitis. Of these, 56 (55.4%) met eligibility criteria and were contacted for recruitment prior to their clinic appointment. 31 (55.4%) of the eligible patients consented to participate and 28 (90.3%) patients completed the 6-month study. Our enrollment cadence was 0.2 patients per surgeon per week. Over 92% of patients randomized to the intervention arm used the decision aid prior to their clinic visit with 83% of participants recommending the patient decision aid to others. The mean age of participants was 56.9 years and 51.6% were female. 64.5% of patients had 1-5 lifetime episodes of diverticulitis, 25.8% had 6-10, and 9.7% had 11+. We observed baseline mean (standard deviation) pre-consultation DCS scores of 44.8 (18.7), indicating high baseline levels of conflict. Marked reductions in conflict scales in both groups were observed, but a greater mean reduction was observed in the decision aid group [33.5 (18.5)] vs control [13.3 (38.0)]. The mean decision regret score was 16.7 (17.3) in the control group vs 11.2 (14.3) in the decision aid group.

Conclusion/Discussion: This study illustrates the feasibility of the diverticulitis decision aid and informs effect sizes to generate a power calculation for a larger trial. Preliminary results suggest a meaningful difference in DCS score reductions between intervention and control. A vast majority of participants would recommend the use of the decision aid to other patients with diverticulitis. Next steps include a large, randomized trial to determine efficacy compared with standard of care.