Nudges and Incentives to Enhance the Opioid Treatment Workforce

This project seeks to examine a critical barrier to optimizing the health care workforce for the treatment of opioid use disorders. Without a dramatic increase in the number of primary care providers trained and comfortable with the many nuances of prescribing medication-assisted treatment (MAT), the staggering increases in opioid overdose deaths will continue to skyrocket. However, Drug Addiction Treatment Act (DATA) 2000 waiver training alone is not enough to facilitate prescribing for patients who desperately need services; an estimated 40% of physicians with waivers do not initiate MAT prescriptions. To address this problem, North Carolina developed a learning collaborative framework to promote MAT training. Learning collaboratives have been shown to be an efficacious approach to increase utilization of MAT, but engagement among providers in North Carolina has been low. To date, the need to encourage provider collaborative participation at scale has not been addressed. This is the critical problem focused on in this proposal. The death rate from accidental opioid overdoses continues to climb at an alarming rate, with overdose deaths in 2016 almost five times the number from 1999. The daily death rate from opioid overdoses in the U.S. alone is now estimated at 115, so every day that evidence-based treatment is not available leads to more preventable deaths. North Carolina is one of the states with both an opioid overdose death rate greater than the national average (11.9 vs 10.4 deaths per 100,000, age-adjusted) and a rate of increase in opioid overdose deaths greater than the national average (19% vs. 16%). North Carolina is also one of four states with an Agency for Healthcare Research and Quality (AHRQ) funded Extension for Community Healthcare Outcomes (ECHO) MAT learning collaborative available to primary care practices, but engagement among providers is low. While the main barriers to engagement are incompletely understood, recent evidence from provider interviews conducted by the study team in December 2017 and January 2018 suggest that one substantial barrier is the time required for weekly ECHO clinics.

This study is designed to assess how to effectively recruit and enroll primary care providers into participation into an ECHO MAT learning collaborative. For study participants that enroll in the learning collaborative, the study team will track participation over 6 months. The study is designed to be implemented in two phases. Phase 1 is the recruitment phase. Providers will be randomized to receive recruitment letters and emails from 1 of 4 conditions. Randomization for this phase will occur at the practice level, as determined by provider address. Participants that contact the study team and enroll in the study will begin Phase 2 of the study. In Phase 2 of the study participants will be randomly assigned to rewards for participation in ECHO MAT learning collaborative. Their participation in the learning collaborative will be tracked. Phase 2 of the study is a pilot. Recruitment: Primary care providers will receive recruitment letters, and may receive emails and phone calls based on the recruitment condition they are assigned to. Informed Consent: Since it is not possible to consent before recruitment, participants will complete the informed consent process at the beginning of Phase 2 of the study. The consent form will be shared with eligible participants when they contact the study team to enroll. Before enrolling in the ECHO MAT learning collaborative, participants will be asked to provide consent in order to enroll in the study. Randomization: Phase 1: Primary care providers will be randomized by practice to 1 of 4 study arms. Computerized randomization will assign practices to study arms, practices and primary care providers will have an equal chance of being randomized to each study arm . Phase 2: Eligible participants will be randomized to the Phase 2 study arm upon enrollment in the study. Randomization will be assigned using a predetermined randomly ordered list of study condition assignments. Upon enrolling in the study, the participants will be assigned to the condition that is next on the list. Participants will have an equal chance of being randomized to each study arm. Assessment: Assessment for Phase 1 will occur when participants contact study staff with interest in the study or to enroll. Any primary care providers that are sent letters but do not contact study staff will be included in study analyses as 'No' for outcomes for Phase 1. Assessment for Phase 2 will occur throughout the 6 months after participants enroll in the study. Outcomes for this study phase are related to participation in ECHO MAT learning collaboratives, which is tracked for each training by study staff. Detailed Description of the Intervention: Phase 1 recruitment letters are designed in a 2 (prosocial mention of need for primary care providers trained in opioid use disorder treatment vs. none) x 2 (mention of additional practice supports available for participation in training vs. none). This yields 4 study conditions for recruitment letters: 1) Letter including prosocial messages and additional mention of practice supports available for participation, 2) Letter including prosocial messages but no additional mention of practice support available for participation, 3) Letter including an additional mention of practice support available for participation but no prosocial messages, and 4) Recruitment as usual letter that does not include prosocial messages or an additional mention of practice support available for participation. Phase 2 examines participation in the learning community by group. Participants will be randomly assigned to 1 of 3 groups. At this time the investigators are not disclosing the Phase 2 conditions, and will disclose these at the end of the study, as approved by UNC's institutional review board (IRB).

Providers assigned to this condition will be sent recruitment materials that are 'recruitment as usual' (Phase 1 Condition 1). These materials will not include prosocial messaging and will not include an additional mention of financial support available for participation. If they enroll, in Phase 2 of the study these participants will be in condition 1.

Providers assigned to this condition will be sent recruitment materials that are 'recruitment as usual' (Phase 1 Condition 1). These materials will not include prosocial messaging and will not include an additional mention of financial support available for participation. If they enroll, in Phase 2 of the study these participants will be in condition 2.

Providers assigned to this condition will be sent recruitment materials that are 'recruitment as usual' (Phase 1 Condition 1). These materials will not include prosocial messaging and will not include an additional mention of financial support available for participation. If they enroll, in Phase 2 of the study these participants will be in condition 3.

Providers assigned to this condition will be sent recruitment materials that include additional mention of financial support available for participation, but will not include prosocial messaging (Phase 1 Condition 2). If they enroll, in Phase 2 of the study these participants will be in condition 1.

Providers assigned to this condition will be sent recruitment materials that include additional mention of financial support available for participation, but will not include prosocial messaging (Phase 1 Condition 2). If they enroll, in Phase 2 of the study these participants will be in condition 2.

Providers assigned to this condition will be sent recruitment materials that include additional mention of financial support available for participation, but will not include prosocial messaging (Phase 1 Condition 2). If they enroll, in Phase 2 of the study these participants will be in condition 3.

Providers assigned to this condition will be sent recruitment materials that include prosocial messaging, but do not include an additional mention of financial support available for participation (Phase 1 Condition 3). If they enroll, in Phase 2 of the study these participants will be in condition 1.

Providers assigned to this condition will be sent recruitment materials that include prosocial messaging, but do not include an additional mention of financial support available for participation (Phase 1 Condition 3). If they enroll, in Phase 2 of the study these participants will be in condition 2.

Providers assigned to this condition will be sent recruitment materials that include prosocial messaging, but do not include an additional mention of financial support available for participation (Phase 1 Condition 3). If they enroll, in Phase 2 of the study these participants will be in condition 3.

Providers assigned to this condition will be sent recruitment materials that include prosocial messaging and an additional mention of financial support available for participation (Phase 1 Condition 4). If they enroll, in Phase 2 of the study these participants will be in condition 1.

Providers assigned to this condition will be sent recruitment materials that include prosocial messaging and an additional mention of financial support available for participation (Phase 1 Condition 4). If they enroll, in Phase 2 of the study these participants will be in condition 2.

Providers assigned to this condition will be sent recruitment materials that include prosocial messaging and an additional mention of financial support available for participation (Phase 1 Condition 4). If they enroll, in Phase 2 of the study these participants will be in condition 3.

Providers assigned to this condition will be sent recruitment materials that are 'recruitment as usual' (Phase 1 Condition 1). These materials will not include prosocial messaging and will not include an additional mention of financial support available for participation.

Providers assigned to this condition will be sent recruitment materials that include an additional mention of financial support available for participation, but will not include prosocial messaging (Phase 1 Condition 2).

Providers assigned to this condition will be sent recruitment materials that include prosocial messaging, but do not include an additional mention of financial support available for participation (Phase 1 Condition 3).

Providers assigned to this condition will be sent recruitment materials that include prosocial messaging and an additional mention of financial support available for participation (Phase 1 Condition 4).

Providers who enroll in the study and are assigned to this condition will be in this group. Other than the opportunity to participate in a learning collaborative, participants receive no compensation.

Providers who enroll in the study and are assigned to this condition will be in this group. In addition to the opportunity to participate in the learning collaborative, participants receive a reimbursement at 30% of the NC Medicaid rate for a patient visit.

Providers who enroll in the study and are assigned to this condition will be in this group. In addition to the opportunity to participate in the learning collaborative, participants receive a reimbursement at 100 percent of the NC Medicaid rate for a patient visit for their time participating in the learning collaborative.

Phase 1 recruitment interventions will be compared based on the number of recruited providers that reach out to study staff via email, phone, or any other method.

Phase 2 study interventions will be compared based on the number of UNC ECHO clinic sessions that a participant attends. Subsequent to COVID-19, the number of ECHO clinic sessions was decreased and the assessment period was extended by 3 additional months for those who had not completed the Follow up period as of 3/9/2020.

Phase 1 recruitment interventions will be compared based on the number of recruited providers that enroll in the Phase 2 study.

Phase 2 study interventions will be compared based on the total amount of time that a participant attends a UNC ECHO clinic session (P2). Subsequent to COVID-19, the number of ECHO clinic sessions was decreased and the assessment period was extended by 3 additional months for those who had not completed the Follow up period as of 3/9/2020.

Phase 2 study interventions will be compared based on the number of participants that receive a DATA 2000 waiver. MAT waiver training is an 8 hour course for physicians and a 24 hour course for advanced practitioners that is required for providers to prescribe and dispense MAT. Subsequent to COVID-19, the number of ECHO clinic sessions was decreased and the assessment period was extended by 3 additional months for those who had not completed the Follow up period as of 3/9/2020.

Phase 2 study interventions will be compared based on the number of participants that begin prescribing MAT paid by Medicaid within one year of enrolling in Phase 2.

Inclusion Criteria: Licensed Primary Care Provider (Doctor of Medicine (MD), Doctor of Osteopathic Medicine (DO), Physician Assistant (PA), Nurse Practitioner (NP)) Currently practice in North Carolina For Nurse practitioners, the following specialties were included: Addiction Psychiatry Addiction/Chemical Dependency Adolescent Medicine Emergency Medicine Family Practice Family Practice, Geriatric General Practice General Preventive Medicine Geriatrics Gerontology Gynecology Infectious Disease Internal Medicine Internal Medicine/Immunology Diagnostic Internal Medicine/Pulmonary Dis. & Crit Medicine/Pediatrics Medicine/Psychiatry Obstetrics Obstetrics and Gynecology/Critical Care Obstetrics/Gynecology Other Specialty Pain Management (phys med) Pain Medicine Pediatrics Psychiatry Psychiatry, Child Psychiatry/Geriatric Public Health Unspecified Hospitalist For MD/DOs practitioners that were considered primary by the medical board care were included except those specializing in sports, addiction, alcohol, and surgery For PAs practitioners were included that are considered primary care by the medical board Exclusion Criteria: Another member of the same practice is already enrolled in the study (one provider per practice in Phase 2) Active participation in University of North Carolina at Chapel Hill (UNC) ECHO Collaborative at beginning of study

Deidentified individual data that supports the results will be shared beginning 9 to 36 months following publication. Some of the outcomes proposed by the study are covered by data use agreements, we will not share measures restricted by data use agreements. We will share data, as allowed, provided the investigator who proposes to use the data has approval from an Institutional Review Board (IRB), Independent Ethics Committee (IEC), or Research Ethics Board (REB), as applicable, and executes a data use/sharing agreement with UNC.

We will share data, as allowed, provided the investigator who proposes to use the data has approval from an Institutional Review Board (IRB), Independent Ethics Committee (IEC), or Research Ethics Board (REB), as applicable, and executes a data use/sharing agreement with UNC.