Sarah Feldman

Supplementary Table S1: Data collection by phase, quarter, system

Background:Guidelines for management of abnormal cervical cancer screening results have increased in complexity over the past two decades. Little is known about how patient-, clinician-, and organization-level factors influence implementation when guidelines change. Process mapping may offer insights into organizational processes and facilitate visualization for potential intervention opportunities.Methods:We conducted an iterative multimodal qualitative assessment to compare abnormal cervical cancer screening management between two health systems: a safety-net institution and an integrated health system. We interviewed clinicians and staff to generate (phase I, May 2019–March 2021) and validate (phase II, July–Oct 2022) process maps at both systems. We conducted a rapid and thematic content analysis and engaged clinical and nonclinical stakeholders during interpretation.Results:At both health systems, process maps informed by phase I participants (n = 31) identified a gap in care during patients’ transition back to primary care following resolution of abnormal tests by gynecologists. In phase II, participants (n = 21) validated and revised maps, noting guideline updates and quality improvement initiatives. Although each system deployed unique strategies to address gaps in care, strategies in common included creating electronic health record–based clinical decision support tools, enabling gynecologists to provide real-time e-consults to primary care clinicians, and engaging patients via the portal.Conclusions:The complexity of cervical cancer screening management guidelines elevates the importance of identifying system-level tools to support clinician decision-making and coordinate between primary and specialty care teams.Impact:Process maps are valuable in generating cross-system comparisons by documenting clinical workflows, identifying care gaps, and engaging participants in formulating potential interventions.

Supplementary Table S2: Standards for Reporting Qualitative Research (SRQR)

Background: Guidelines for management of abnormal cervical cancer screening results have increased in complexity over the past two decades. Little is known about how patient-, clinician-, and organization-level factors influence implementation when guidelines change. Process mapping may offer insights into organizational processes and facilitate visualization for potential intervention opportunities. Methods: We conducted an iterative multimodal qualitative assessment to compare abnormal cervical cancer screening management between two health systems: a safety-net institution and an integrated health system. We interviewed clinicians and staff to generate (phase I, May 2019–March 2021) and validate (phase II, July–Oct 2022) process maps at both systems. We conducted a rapid and thematic content analysis and engaged clinical and nonclinical stakeholders during interpretation. Results: At both health systems, process maps informed by phase I participants (n = 31) identified a gap in care during patients’ transition back to primary care following resolution of abnormal tests by gynecologists. In phase II, participants (n = 21) validated and revised maps, noting guideline updates and quality improvement initiatives. Although each system deployed unique strategies to address gaps in care, strategies in common included creating electronic health record–based clinical decision support tools, enabling gynecologists to provide real-time e-consults to primary care clinicians, and engaging patients via the portal. Conclusions: The complexity of cervical cancer screening management guidelines elevates the importance of identifying system-level tools to support clinician decision-making and coordinate between primary and specialty care teams. Impact: Process maps are valuable in generating cross-system comparisons by documenting clinical workflows, identifying care gaps, and engaging participants in formulating potential interventions.

Background Guidelines for managing abnormal cervical cancer screening results are complex and adherence is challenging for clinicians. Previous studies have identified gaps in knowledge as a possible cause; few have explored the confidence clinicians have in their management decisions. Confidence in decision-making may influence management practices, particularly when guidelines are complex and evolving. Objective Assess whether confidence in decision-making is associated with making guideline-concordant recommendations for abnormal cervical cancer screening results. Design A clinician survey used vignettes to ask clinicians to make a management recommendation for different abnormal results and rate their level of confidence in their response. Participants Physicians and advanced practice providers (APPs) at three diverse health systems in Washington, Texas, and Massachusetts. Main Measures Correct response to each vignette based on either the 2012 or 2019 American Colposcopy and Cervical Pathology (ASCCP) management guidelines. Key Results In total, 501 clinicians completed the survey between October and December 2020 (response rate 53.7%). Overall, most clinicians made guideline-recommended management decisions for two vignettes (73.2 and 73.7%), but fewer were confident in their selection (48.3% and 46.6%, respectively). Clinicians who reported high levels of confidence were more often correct than those who reported lower levels of confidence (85.8% vs. 62.2% and 87.5% vs. 60.7%, both p<0.001). After adjusting for clinician and practice characteristics, confidence remained significantly associated with selecting the correct answer. Conclusions Clinician confidence in management decisions for abnormal cervical cancer screening results was significantly associated with knowing guideline-concordant recommendations. Given the complexity of cervical cancer management guidelines, solutions to improve clinician confidence in decision-making are needed.

Supplementary Table S2 shows the parameters used to define guideline-concordant management after the index abnormality and the initial colposcopy.

Supplementary Table S4 shows associations between patient characteristics and guideline-concordant follow-up after the index abnormality and the initial colposcopy based on multivariable logistic regression among patients at PH.

Supplementary Table S2 shows the parameters used to define guideline-concordant management after the index abnormality and the initial colposcopy.

Supplementary Table S3 shows associations between patient characteristics and guideline-concordant follow-up after the index abnormality and the initial colposcopy based on multivariable logistic regression among patients at MGB.

Supplementary Table S1 shows the inclusion/exclusion criteria applied to determine the study sample and the sample size after applying each criteria to the overall and site-specific samples.