A New Quality Improvement Toolkit to Improve Opioid Prescribing in Primary Care

Constance van Eeghen, DrPH, MHSA, MBA; Amanda G. Kennedy, PharmD, BCPS; Mark E. Pasanen, MD; Charles D. MacLean, MD


J Am Board Fam Med. 2020;33(1):17-26. 

In This Article


Trial 1

The 10 ambulatory clinics that participated in 2012 included a combination of family medicine, primary care internal medicine, and surgical specialties. Half were owned by health systems and the remaining were independent clinics or members of FQHCs. The primary driver that motivated clinic participation was an engaged champion (60%); other drivers included a corporate decision to improve opioid prescribing or an investigation by the state's medical practice board regarding high prescribing (Table 2). All clinics completed the structured QI project through team meetings with providers, clinical staff, and front desk staff in an average of 3 months (range, 1 to 7). The teams met an average of 6 times (range, 3 to 10), spending a mean of 9.1 hours of team time (range, 5 to 12) to study workflow and design changes that incorporated a selection of best practice strategies. Teams implemented a range of 3 to 13 of 15 best practice strategies presented to them and half implemented 8 or 9 strategies. The most frequently chosen strategies (36%) were those responding to an emerging policy statement about best opioid-prescribing practices from the state Medical Practice Board, finalized in 2014,[20] with the remaining strategies focused on streamlining workflow (28%), providing peer support among members within the clinic (22%), and responding to patients at risk (10%) (Table 3).

Respondents to the pre- and postproject surveys included providers, clinical staff, front desk staff, and other clinic members. We matched respondents from both time periods and maintained a net response rate of 81%. In the matched group, 36 were providers and 83 were nonproviders. Both providers and staff reported that opioid prescribing was improved across a variety of domains. For example, provider satisfaction with opioid-prescribing management increased from 42% to 96% after the intervention, providers' confidence in prescribing opioids improved from 69% to 88%, and staff satisfaction increased from 54% to 81%. Provider confidence in many other components of CDC guideline-based prescribing improved as well (Table 4). After 3 months of implementation, clinic staff evaluated strategy completion, with 47% reporting selected strategies as "All done," 21% as "Partly done," and 25% as "Started." Furthermore, 76% of respondents felt that their QI project was effective in improving opioid management.

Trial 2

The first edition of the toolkit incorporated all 15 strategies presented to and used by the clinics in trial 1. Of the 7 clinics participating from 2014 to 2016, almost half were family medicine followed by primary care internal medicine and specialty care. As before, they were predominately owned by health systems with the remaining an independent clinic and a member of an FQHC. The primary driver of clinic change was again an engaged champion (71%) as well as a corporate decision to improve (Table 2). All clinics completed the structured QI project through team meetings with providers, clinical staff, and front desk staff in an average of 7.7 months (range, 4 to 15), more than twice the average of trial 1 clinics working with an expert facilitator (3.0 months). The teams met an average of 7.3 times (range, 2 to 11), spending a mean of 8.7 hours of team time (range, 2 to 13) to select and redesign workflows, similar to trial 1 results of 6.0 meetings and 9.1 hours, respectively.

Teams implemented an average of 8.6 best practice strategies (range, 1 to 13), similar to the trial 1 average of 8.9, with 5 of the 7 clinics implementing 9 or more strategies. As before, the most commonly chosen category of strategies (35%) responded to the now published state Medical Practice Board guidelines, with the remaining strategies focused on streamlining workflow (32%), providing peer support (20%), and responding to patients at risk (13%) (Table 3). Pre- and postproject surveys were modified and used at the discretion of the clinics; their QI outcomes are not available, although 1 clinic did publish its work.[21] Our field notes indicated that these projects were also highly rated by clinic members and satisfaction with the process and outcomes were high. As trial 2 progressed, we updated the toolkit to reflect updates to the CDC guidelines, state law, and feedback from clinics and facilitators (second ed).

Trial 3

Toolkit 2019 (third ed) expanded to 28 strategies (from 16) but reduced the length of the instructions, offering a workbook format more usable by office personnel. These strategies, organized into workflow sections from "Before Prescribing" to "Manage the Population," highlight CDC guidelines as well as commonly used strategies to manage pain and provide opioid therapy when chosen as part of treatment. This toolkit is undergoing testing with new clinics; a complete list of all strategies is found in Table 1. Access to this and previous toolkit editions are available at http://www.med.uvm.edu/ahec/workforceresearchdevelopment/toolkits-and-workbooks.

In 2019, 10 primary care clinics came forward to participate in this project and have maintained their engagement. Over half are FQHC clinics, 3 are owned by a health system, and 1 is an independent group clinic. The primary driver motivating change is, once again, a champion in the clinic (60%), and other drivers include a corporate decision to improve and a medical practice board investigation. The owners of these clinics separately selected the same 3 strategies: a team approach, the use of consistent strategies across all providers in the clinic, and the initiation of a "pain management council" to formalize peer-to-peer support in managing patients at risk (Table 3). One clinic also added 7 strategies for process improvement and regulatory compliance. When we explored the shared emphasis on peer support as an initial set of strategies, we received comments such as "It is my job to manage substance use disorder" and "We will find a way to care for every chronic pain patient." These comments come at a time when the number of available specialists in chronic pain or opioid use disorder has declined across the state and the need for supportive strategies has increased.

Combined Results

We aggregated the strategies chosen across all 27 clinics and identified the top 12, stratified by trial (Figure 2). We found 3 peer-to-peer strategies (team approach, consistent strategies, and pain management council) among the top 12, with greater representation in the third trial. Other frequently selected strategies include screening urine (on a scheduled basis or randomly), patient education followed by an agreement and consent, checking the PDMP regularly, ongoing risk assessments, dedicated chronic pain visits, and rosters/registries to support chronic pain care. These and other commonly selected strategies directly affect clinic workflow and are supported by the toolkit.

Figure 2.

Top 12 strategies selected by 27 ambulatory care clinics by trial period. Abbreviations: PDMP, prescription drug monitoring programs; Pt, Patient.