Promoting Clinical Guidelines for Opioid Prescribing

The proposed study evaluates a sequence and combination of implementation strategies that is tailored to the needs of stakeholders at different levels (health system, clinic, and prescriber). We will deliver an adaptive version of systems consultation that progressively and adaptively drills down to offer more and more personalized levels of implementation support. The intervention starts with academic detailing, a systems-level strategy consisting of an expert-led training session plus distance-based follow up support. This strategy continues for the 21-month intervention for all clinics, but at 3 months, half of the clinics will be randomized to receive practice facilitation. Practice facilitation is a clinic-level strategy in which a highly-skilled external change agent helps clinics improve processes related to opioid prescribing. At 6 months, half of prescribers will be randomized to receive physician peer coaching. Physician peer coaching is a clinic-level strategy in which a physician expert gives one-on-one support to prescribers in managing their patients on long-term opioid therapy. These 3 discrete strategies will be delivered in a sequential, multiple-assignment randomized trial to 38 clinics from 2 Wisconsin health systems. The study has 3 specific aims:

• Compare the effect over 21 months of (1). An adaptive systems consultation implementation strategy (intervention group) vs. (2). Academic detailing alone (control group) on average morphine milligram equivalent dose (the primary outcome).
• Develop an assessment of contextual factors that influence the effectiveness of different implementation strategies. This aim will test 4 moderators and assess other factors that affect implementation. The goal is to develop a tool that decision-makers can use to predict which implementation strategies will be most effective in different settings.
• Estimate the costs of delivering 4 different sequences and combinations of strategies, including the incremental cost effectiveness of adding facilitation and physician peer coaching. Results will help decision-makers weigh the costs and effects of using different implementation strategies.

Inclusion criteria

Clinics will be eligible for the study if they:

• are a primary care clinic (non-pediatric primary care, internal medicine, or family medicine);
• have not received the systems consultation intervention;
• do not explicitly prohibit initiating opioid therapy;
• do not exceed the performance on key measures of guideline concordance (fewer than 80% of long-term opioid patients have treatment agreements and a urine drug screen in the past 12 months)

Prescribers will be eligible if they:

• are a primary care provider at the clinic;
• are not temporary providers who do not manage stable panels or patients;

While patients are not subjects of study, de-identified prescriber panel data will be used to assess outcome measures. To be included in the de-identified prescriber panel data, patients must:

• have a primary care provider at the clinic;
• are prescribed opioid therapy for at least 3 consecutive months;
• do not have a cancer diagnosis or are receiving hospice care.

Exclusion criteria

Clinics will be excluded if they are not a primary care clinic, have received the systems consultation intervention, prohibit initiating opioid therapy, or exceed the threshold on key measures of guideline concordance.

Prescribers will be excluded if they don't have prescribing privileges or are temporary providers who do not manage stable panels or patients.

De-identified prescriber panel data will be excluded from outcome measures if they do not have a primary care provider at the clinic, are not prescribed opioid therapy for at least 3 consecutive months, or have a cancer diagnosis or are receiving hospice care.