Table VI.

Effect of Implementation Challenges on Study Design

Original Element of Study Design	Challenge	Impact on Study Design
Recruit 63 patients at each of the 30 participating practices	Smaller practices did not have enough eligible patients to reach recruitment target of 63 patients per practice.	Adjusted our recruitment target to an average of 63 patients per practice, over recruiting at other practices in the same intervention arm within the same health system when possible or, if not possible, over recruiting at practices in other health systems that have a similar demographic makeup to the small practice.
Recruit patients of White, Black and Latinx background in 1/3–1/3–1/3 proportions to examine racial and ethnic disparities among three racial groups	Participating practice sites had fewer eligible Latinx patients than would allow us to reach this goal.	Changed the analysis plan to examine racial and ethnic disparities among two racial groups (African-American and white), and changing the Latinx vs. white comparison to an exploratory one.
Embed one full-time CM and one full-time CHW at each of the 15 practices assigned to the Collaborative Care/Stepped Care study arm	There were limited health system and study resources to fully fund full-time CMs and CHWs at every practice site.	Health systems either identified an acceptable percentage of effort for CMs and CHWs to dedicate to the study or combined per practice resources to employ full-time CMs and CHWs responsible for more than one practice site.
CMs, CHWs, and PCPs would participate in at-least once weekly, in-person, “rounding” meetings as part of the collaborative care team	Several health system leaders expressed preferences for regular, within EMR, CM-CHW-PCP communication and ad hoc CM-CHW-PCP in-person or telephonic communication rather than regularly scheduled meetings.	Allowed for more flexibility in the communication among the collaborative care team to reflect health system leaders’ preferences.
CMs would “step-up” patients not achieving BP control or other health goals after 3 months to either a CHW or to the specialist core and initiate an additional step-up 3 months after the first step-up, if BP control/other health goals are not achieved	CM and CHW supervisors identified the phrase “step-up” was problematic as it implied a hierarchy between the CM and CHW. Health system leaders preferred the term “consultation” for specialists to avoid the perception that patients would actually need to see the specialist team. CMs and CHWs provided early feedback that patients were in need of additional resources, particularly resources related to social determinants of health, earlier than 3 months into working with the CM.	Discontinued use of the phrase “step-up” and adopted “CHW referral and “specialist consultation.” Adapted the protocol to allow for earlier CHW referrals and specialist consultations. Identified specific criteria for CMs to follow to determine if a patient qualified for immediate, 1 month, or 3 month referral to a CHW. Identified specific criteria for CMs to follow to determine if a patient qualified for specialist consultation at 1 month of 3 months.
All participating practices would receive a hypertension dashboard, but only practices in the Collaborative Care/Stepped Care study arm would receive a race and ethnicity-stratified hypertension dashboard	Due to limitations of the population health management software at the time of the dashboard development and concerns about an inconsistent approach to quality improvement across the organization, the largest participating health system determined that they would prefer to release the race and ethnicity-stratified hypertension dashboard to all practices, regardless of study arm or study participation.	All participating practices receive a race and ethnicity-stratified hypertension dashboard. Leaders from the Collaborative Care/Stepped Care practices receive additional coaching on dashboard promotion and use through quarterly system level leadership coaching calls.
Each health system would identify a clinical and an administrative champion at the system-level and a clinical champion, preferably a provider, at the practice-level, for each participating practice	While all health systems identified clinical and administrative champions at the health system level, not all health systems identified practice-level clinical champions.	Allowed for flexibility in the practice-level champion role, including allowing medical assistants identified by health system leaders to serve as project champions.