Leveraging human-centered design and causal pathway diagramming toward enhanced specification and development of innovative implementation strategies: a case example of an outreach tool to address racial inequities in breast cancer screening.

BACKGROUND: Implementation strategies are strategies to improve uptake of evidence-based practices or interventions and are essential to implementation science. Developing or tailoring implementation strategies may benefit from integrating approaches from other disciplines; yet current guidance on how to effectively incorporate methods from other disciplines to develop and refine innovative implementation strategies is limited. We describe an approach that combines community-engaged methods, human-centered design (HCD) methods, and causal pathway diagramming (CPD)-an implementation science tool to map an implementation strategy as it is intended to work-to develop innovative implementation strategies. METHODS: We use a case example of developing a conversational agent or chatbot to address racial inequities in breast cancer screening via mammography. With an interdisciplinary team including community members and operational leaders, we conducted a rapid evidence review and elicited qualitative data through interviews and focus groups using HCD methods to identify and prioritize key determinants (facilitators and barriers) of the evidence-based intervention (breast cancer screening) and the implementation strategy (chatbot). We developed a CPD using key determinants and proposed strategy mechanisms and proximal outcomes based in conceptual frameworks. RESULTS: We identified key determinants for breast cancer screening and for the chatbot implementation strategy. Mistrust was a key barrier to both completing breast cancer screening and using the chatbot. We focused design for the initial chatbot interaction to engender trust and developed a CPD to guide chatbot development. We used the persuasive health message framework and conceptual frameworks about trust from marketing and artificial intelligence disciplines. We developed a CPD for the initial interaction with the chatbot with engagement as a mechanism to use and trust as a proximal outcome leading to further engagement with the chatbot. CONCLUSIONS: The use of interdisciplinary methods is core to implementation science. HCD is a particularly synergistic discipline with multiple existing applications of HCD to implementation research. We present an extension of this work and an example of the potential value in an integrated community-engaged approach of HCD and implementation science researchers and methods to combine strengths of both disciplines and develop human-centered implementation strategies rooted in causal perspective and healthcare equity.

Automated Opt-Out vs Opt-In Patient Outreach Strategies for Breast Cancer Screening: A Randomized Clinical Trial.

Importance: Optimal strategies for population-based outreach for breast cancer screening remain unknown. Objective: To evaluate the effect on breast cancer screening of an opt-out automatic mammography referral strategy compared with an opt-in automated telephone message strategy. Design, Setting, and Participants: This pragmatic randomized clinical trial was conducted from April 2022 to January 2023 at a single Veterans Affairs (VA) medical center. Participants were female veterans aged 45 to 75 years who were eligible for breast cancer screening and enrolled in VA primary care. Intervention: Veterans were randomized 1:1 to receive either an automatic mammography referral (opt-out arm) or an automated telephone call with an option for mammography referral (opt-in arm). Main Outcomes and Measures: The primary outcome was completed mammography 100 days after outreach. Secondary outcomes were scheduled or completed mammography by 100 days after outreach and referrals canceled if mammography was not scheduled within 90 days. Both intention-to-treat analyses and a restricted analysis were conducted. The restricted analysis excluded veterans who were unable to be reached by telephone (eg, a nonworking number) or who were found to be ineligible after randomization (eg, medical record documentation of recent mammography). Results: Of 883 veterans due for mammography (mean [SD] age, 59.13 [8.24] years; 656 [74.3%] had received prior mammography), 442 were randomized to the opt-in group and 441 to the opt-out group. In the intention-to-treat analysis, there was no significant difference in the primary outcome of completed mammography at 100 days between the opt-out and opt-in groups (67 [15.2%] vs 66 [14.9%]; P = .90) or the secondary outcome of completed or scheduled mammography (84 [19%] vs 106 [24.0%]; P = .07). A higher number of referrals were canceled in the opt-out group compared with the opt-in group (104 [23.6%] vs 24 [5.4%]; P < .001). The restricted analysis demonstrated similar results except more veterans completed or scheduled mammography within 100 days in the opt-out group compared with the opt-in group (102 of 388 [26.3%] vs 80 of 415 [19.3%]; P = .02). Conclusions and Relevance: In this randomized clinical trial, an opt-out population-based breast cancer screening outreach approach compared with an opt-in approach did not result in a significant difference in mammography completion but did lead to substantially more canceled mammography referrals, increasing staff burden. Trial Registration: ClinicalTrials.gov Identifier: NCT05313737.

Factors associated with lung cancer risk factor documentation.

Objectives: To identify factors associated with the minimum necessary information to determine an individual's eligibility for lung cancer screening (ie, sufficient risk factor documentation) and to characterize clinic-level variability in documentation. Study Design: Cross-sectional observational study using electronic health record data from an academic health system in 2019. Methods: We calculated the relative risk of sufficient lung cancer risk factor documentation by patient-, provider-, and system-level variables using Poisson regression models, clustering by clinic. We compared unadjusted, risk-adjusted, and reliability-adjusted proportions of patients with sufficient smoking documentation across 31 clinics using logistic regression models and 2-level hierarchical logit models to estimate reliability-adjusted proportions across clinics. Results: Among 20,632 individuals, 60% had sufficient risk factor documentation to determine screening eligibility. Patient-level factors inversely associated with risk factor documentation included Black race (relative risk [RR], 0.70; 95% CI, 0.60-0.81), non-English preferred language (RR, 0.60; 95% CI, 0.49-0.74), Medicaid insurance (RR, 0.64; 95% CI, 0.57-0.71), and nonactivated patient portal (RR, 0.85; 95% CI, 0.80-0.90). Documentation varied across clinics. The reliability-adjusted intraclass correlation coefficient decreased from 11.0% (95% CI, 6.9%-17.1%) to 5.3% (95% CI, 3.2%-8.6%), adjusting for covariates. Conclusions: We found a low rate of sufficient lung cancer risk factor documentation and associations of risk factor documentation based on patient-level factors such as race, insurance status, language, and patient portal activation. Risk factor documentation rates varied across clinics, and only approximately half the variation was explained by factors in our analysis.

Improving smoking history documentation in the electronic health record for lung cancer risk assessment and screening in primary care: A case study.

Improving risk factor documentation in the electronic health record (EHR) is important in order to determine patient eligibility for lung cancer screening. System-level prioritization combined with a clinic-level initiative can improve risk factor documentation rates. Multi-faceted interventions that include training, process improvement, data management, and continuous performance feedback are effective and can be integrated into existing workflows.

Long-term outcomes in patients who received a corneal graft for keratoconus between 1980 and 1986.

PURPOSE: To estimate the probability of allograft rejection, graft failure, and recurrent keratoconus (KC) and to assess vision-specific quality of life 20 to 25 years after corneal transplantation for KC. DESIGN: Retrospective case series. METHODS: Two hundred nineteen eyes of 184 subjects were identified from the Michigan Corneal Transplantation Patient Registry as receiving corneal transplantation for KC from 1980 through 1986. Current ophthalmic examinations and the 25-item National Eye Institute Visual Function Questionnaire were obtained. Kaplan-Meier analyses were used to estimate the time-related probability of allograft rejection, graft failure, and KC recurrence. Cox regression was used to identify predictive factors of these outcomes. The 25-item National Eye Institute Visual Function Questionnaire scores were summarized with descriptive statistics. RESULTS: Follow-up was available up to 27 years after surgery (median, 10 years). Rejection occurred in 98 of 219 grafts. Most rejections occurred in the first 2 years (probability, 41%; standard error, 3%). KC recurrence was noted in 6 grafts 9 to 20 years after surgery, with a 20-year probability of 10% (standard error, 4%). Eighteen grafts failed, with a 20-year probability of 12% (standard error, 3%). Larger host trephine size, male donor gender, and nonwhite donor race were associated with increased rejection hazard. Worse astigmatism and nonwhite recipient race were associated with increased failure hazard. Twenty-eight subjects completed the 25-item National Eye Institute Visual Function Questionnaire at an average of 23 years after surgery (range, 18 to 26 years). Their mean composite score was 84.5 (standard deviation, 12.1). CONCLUSIONS: Allograft rejection is frequent in the 2 years after corneal graft for KC. However, the 20-year probabilities of graft failure and recurrent KC are low. Given the relative youth of KC graft recipients, these statistics should enhance the information they receive.