This pragmatic trial examines the uptake and effects of primary care clinician commitments to follow 3 Choosing Wisely® recommendations. The investigators hypothesize that pre-encounter invitations to clinicians to commit to the recommendations will decrease ordering of: (1) imaging tests for low back pain, (2) antibiotics for acute sinusitis, and (3) imaging tests for headaches. The study is a mixed-methods, stepped wedge cluster randomized trial in which the intervention will be sequentially introduced to 6 clinics in southeastern Michigan in a randomly assigned order.

Clinicians often make decisions about ordering low-value services in the midst of clinical encounters when their thinking can be rushed and susceptible to patient demands. This pragmatic trial examines the uptake and effects of shifting clinicians' decisions about ordering low-value services to the pre-encounter period when their thinking is slower and more deliberative; inviting clinicians to commit to avoid ordering those services during patient encounters; and providing to clinicians who committed decision support resources such as point-of-care reminders, patient education handouts, and access to brief communications training.

The investigators will conduct a mixed-methods, stepped wedge cluster randomized trial in 6 IHA primary care clinics in southeastern Michigan. The intervention will be introduced in each clinic in sequence approximately one month apart and in a randomly assigned order (i.e., temporal randomization). In each clinic the intervention will continue for 1 to 6 months so that the intervention ends in all clinics at the same time. For example, in the first clinic the intervention will run for 6 months; in the sixth clinic the intervention will run for 1 month. The total length of the control period plus the intervention period will be 8 months at all clinics.

Prior to the start of the study, IHA made its primary care clinicians aware of publicly available recommendations from the ABIM Foundation's Choosing Wisely® campaign applicable to avoiding overuse of low-value services for 3 common conditions: (1) low back pain, (2) acute sinusitis, and (3) headaches. However, research has shown that clinician behavior seldom changes as a result of information alone. Therefore, at the start of the intervention period clinicians will be invited to make a pre-encounter commitment to follow the 3 Choosing Wisely recommendations. Clinicians who choose to commit will throughout the intervention period receive point-of-care commitment reminders, Choosing Wisely patient education handouts, and weekly emails with decision support resources. Clinicians will remain free to order services at any time without penalty.

The investigators will measure whether the clinician-focused intervention leads to declines in rates of ordering imaging tests for low back pain, antibiotics for acute sinusitis, and imaging tests for headaches. The study team will analyze ordering patterns and costs using clinical operations datasets (drawn from electronic health records and from the billing and practice management system). The following hypotheses will be tested:

Hypothesis 1: Inviting primary care clinicians to commit to avoid ordering imaging for low back pain, antibiotics for acute sinusitis, and imaging for headaches will decrease ordering of these services.

Hypothesis 2: Inviting primary care clinicians to commit to avoid ordering imaging for low back pain, antibiotics for acute sinusitis, and imaging for headaches will reduce health care costs in patient visits for low back pain, acute sinusitis, and headaches.

Hypothesis 3: Inviting primary care clinicians to commit to avoid ordering low-value services will reduce per-clinician per-visit health care costs across clinical conditions.

For each outcome investigators will also compare differences between clinicians who did and did not commit to follow the 3 Choosing Wisely recommendations. Additionally, investigators will compare differences in outcomes between the 3-month follow-up period (immediately after the intervention has ended) and the control and intervention periods.

After the intervention period, the study team will conduct surveys and semi-structured interviews with study clinicians. The survey results will be used to identify correlates of primary care clinician uptake of invitations to commit to avoid ordering low-value services. The semi-structured interviews will elicit qualitative data on clinician attitudes towards overuse of low-value services, clinicians' responses to the intervention, and facilitators of and barriers to avoidance of delivering low-value care.

Inclusion Criteria:

• Primary care clinicians in 6 primary care (family medicine and internal medicine) clinics within the IHA organization of Ann Arbor, Michigan

Exclusion Criteria:

• None