Bedford VA Research Corporation

Pre-exposure prophylaxis (PrEP) is an effective HIV prevention strategy for adolescents and young adults (AYAs, ages 14-26). PrEP uptake among AYAs remains low. Barriers to uptake include provider discomfort discussing sexuality, challenges engaging parents in sexual health conversations, and low perceived HIV risk. To address these gaps, we aimed to develop a toolkit to enhance AYA-serving primary care providers' (PCPs) comfort and confidence in discussing and prescribing PrEP. The Consolidated Framework for Implementation Research informed our approach. Four focus groups with youth-serving PCPs across three clinics in a Northeastern US city were held in April and May 2024. PCPs included residents (n = 10), attending physicians (n = 11), nurse practitioners (n = 2), and a physician assistant (n = 1). Debrief summaries were created following each focus group and analyzed. We used a rapid qualitative process, guided by an a priori codebook and reflexive thematic analysis. PCPs emphasized that the toolkit should be easy to access, interactive, and provide practical, age-appropriate prescribing information. They identified outer-setting factors such as the role of parents in AYAs' health care, sexual health education in schools, and uncertainty around laws that impact PrEP provision to minors. Within the clinical setting, providers noted that limited time often hindered sexual health discussions. Individually, PCPs reported moderate comfort with PrEP and were motivated to prescribe PrEP when AYAs expressed interest. Findings informed the development of a PrEP toolkit tailored for AYA-serving PCPs and future implementation tailored for AYA-serving PCPs. This approach may help expand access and address persistent barriers to PrEP uptake among AYAs.

Health care systems increasingly incorporate social risk data into electronic health records (EHRs) to address needs like food, housing, and transportation insecurity. This study evaluated whether EHR-integrated social clinical decision support (SCDS) tools improved control of blood pressure (BP) and hemoglobin A1c (HbA1c) and increased social risk-informed care and documentation in community-based clinics. We conducted a cluster randomized trial in a large primary care network. This pragmatic trial was designed to assess tool impact in real-world clinic conditions. Six clinics received SCDS tools embedded in the EHR; 44 clinics served as controls. The tools supported clinic-wide workflows and targeted decision support. A screening alert was triggered for adult patients lacking up-to-date social risk screening. Additional components were activated for patients with uncontrolled hypertension or diabetes, or with a diagnosis of either condition combined with a visit no-show rate of at least 50%. Primary outcomes were BP and HbA1c control. Secondary outcomes included social risk screening and documentation. Generalized linear mixed models accounted for patient clustering. We also examined use patterns of individual tool components. Blood pressure control improved over 12 months in both arms, with significantly greater improvement in intervention clinics. Control of HbA1c showed no significant differences. Intervention clinics had significantly greater odds of social risk screening and documentation. Use of individual SCDS tool components varied widely across clinics. Access to EHR-integrated SCDS tools was associated with increased documentation of social risks and greater improvements in BP control. These findings support embedding social risk data into clinical workflows to enhance chronic disease management in primary care.

Clinical champions can be effective for increasing uptake of evidence-based interventions. However, little is known about how to prepare them to be impactful, particularly within large healthcare systems. We present a conceptual model, grounded in the Awareness, Desire, Knowledge, Ability, Reinforcement (ADKAR®) change management framework, to guide training for clinical champions. In 2021, the U.S. Department of Veterans Affairs implemented clinical champions in primary care and mental health services to facilitate uptake of Whole Health, a person-centered holistic approach to healthcare. Our conceptual model was created through iterative team discussions about learnings from our evaluation of Whole Health clinical champion implementation. This evaluation included two rounds of interviews with clinical champions, and three rounds of a practice reflection survey (aligned with ADKAR) administered to champions. Drawing on these data and ADKAR, we developed a conceptual model of how clinical champions can be supported through two complementary and sequential change management processes. The first process is related to their practice change. Clinical champions must start by gaining awareness of and interest in the new practice. They can then develop foundational knowledge and skills to enact it. Finally, they will only maintain the practice if they observe benefits of its use. Once they have progressed through the ADKAR stages in relation to the practice change, the second process they must undertake is in relation to the clinical champion role. They must first understand why clinical champions are needed and have an interest in the role. They then need training and skills for the role (e.g., overcoming barriers, mentorship). Finally, to continue the role over time they must see that champions are making a difference. Only after champions have gone through both processes can they effectively support their colleagues in progressing through the ADKAR stages to implement the change in their practice. Given that clinical champions are a widely used implementation strategy, this work holds promise for improving its impact on implementation and effectiveness outcomes. By supporting tailoring training to where champions are in the change management processes, our data-driven conceptual model can improve champions' effectiveness as change agents.

This qualitative study evaluates ambient scribe technology using standardized patient encounters in a simulated clinical environment, combining a validated measure of documentation quality with qualitative insights.

ABSTRACTObjectiveEmbedded research partnerships can advance the implementation of evidence‐based policies and practices, including those aligned with person‐centered care. Care delivery model transformations, such as the VA's person‐centered Whole Health System (WHS), can benefit from an ongoing cycle of program implementation and their evaluation to inform future evolution. This paper describes how embedded researchers partnered with policy makers leading VA's WHS transformation to support its development, implementation, and scaling to illustrate lessons learned for embedded research.Data Sources and Study SettingFifty‐eight embedded research projects were conducted from FY2013 to FY2024.Study DesignWe recorded each project's scope, methodology, results, products, and impact. Through a group reflection process, we identified common cross‐project lessons that fostered this successful embedded research partnership. Finally, we mapped projects to phases of Kilbourne's Knowledge to Action Framework (Pre‐Implementation, Implementation, and Sustainment) to demonstrate how embedded researchers defined evaluation questions, evaluated WHS transformation, and assessed outcomes to inform the implementation and sustainment of VA's WHS transformation.Data Collection/Extraction MethodsProjects used multiple qualitative, survey, and large database methods.Principal FindingsAcross 58 projects, 380 discrete products were used by our operational partner to refine the WHS model, improve implementation support, scale effective practices, inform new policy, and sustain transformation. Three practices cut across these projects to contribute to our successful embedded research partnership: agility, collaboration, and continuous learning and improvement. Additionally, the purpose, questions, and methods of embedded research projects varied as operational partner activities moved across pre‐implementation, implementation, and sustainment phases, iteratively impacting WHS transformation.ConclusionsEmbedded research can transform a healthcare system through the timely translation of data into practice, enabling evidence‐based policy and practice decisions. Our embedded research‐operations partnership, characterized by trust, respect, and strong communication, ensured that evaluations generated an evidence base to inform both implementation and impact.

Clinicians often elicit and incorporate patient priorities into care plans. This process is complicated when patients and their family caregivers express incongruent priorities. We investigated patient and family caregiver priorities expressed during medical visits and clinician responses to divergent priorities. Methods: Our data consisted of four first-time medical appointments, each attended by an older patient, at least one family caregiver, and a team of geriatrics clinicians. We audio-recorded and transcribed the appointments, each lasting at least 120 min, and performed thematic analysis of the expressed priorities and clinician responses. Results: When incongruence in priorities emerged, clinicians responded with acknowledgement to each party’s priority and highlighted the incongruence itself, pointing out common ground when present, and pivoted to care planning, while exhibiting warmth and avoiding taking sides. Conclusions: Medical visits with patients and their family caregivers with divergent views present an opportunity for clinicians to promote candor and perspective-taking while finding common ground and assist throughout complex conversations. Further research should explore divergent priorities in larger patient samples with more diversity in cognitive ability and in varied healthcare settings.

Black individuals have higher incidence and mortality rates of prostate cancer than White individuals, and existing literature on treatment sequencing strategies in metastatic castration-resistant prostate cancer (mCRPC) is limited and conflicting. Understanding the interplay between race and treatment could optimize care and improve overall survival (OS). This retrospective observational cohort study assessed survival outcomes based on race and initial androgen receptor pathway inhibitor (ARPI) treatment among veterans with mCRPC. Participants included White or Black veterans diagnosed with mCRPC between January 1, 2011, and March 12, 2024, initially treated with either abiraterone or enzalutamide. The primary outcome was OS measured from the date of diagnosis and from the start of initial therapy. Kaplan-Meier estimates and Cox proportional hazards models were used to evaluate outcome differences by race and treatment. A total of 12,355 patients met eligibility criteria. The mean age of the cohort was 74.57 years, with Black patients diagnosed approximately 4 years younger than White patients (71.77 vs 75.60, respectively; P<.001). Median prostate-specific antigen and mean Charlson comorbidity index scores were higher among Black patients. Median OS was higher in Black patients compared with White patients (36.3 months [95% CI, 35-38] vs 31 months [95% CI, 30.3-31.7], respectively; inverse probability of treatment weighting hazard ratio [IPTW HR], 0.80; P<.0001) Median OS was also longer among patients prescribed enzalutamide compared with those prescribed abiraterone (28 months [95% CI, 27.0-28.9] vs 23.7 months [95% CI, 23.1-24.2], respectively; IPTW HR, 1.14; P<.0001). No significant race-by-treatment interaction was observed. Black veterans with prostate cancer treated with ARPIs lived approximately 6 months longer after developing mCRPC but had a shorter lifespan-by about 3 years-compared with White patients. Differences in tumor biology may help explain why Black individuals develop prostate cancer at a younger age and exhibit better responses to cancer therapy.

Telehealth can improve glycemic control for patients with type 2 diabetes (T2D), but programs must be designed in a manner that facilitates real-world uptake and effectiveness within learning health systems (LHS). To highlight how the Veterans Affairs (VA) LHS supports real-world practice change by examining Advanced Comprehensive Diabetes Care (ACDC), a telehealth program for persistently poor T2D control, in the context of the Learning Health System Consolidated Framework (LHS-CF) taxonomy. Evaluation of program implementation within an LHS. N = 1154 real-world ACDC participants INTERVENTIONS: ACDC is a 6-month program that leverages widely available VA telehealth infrastructure to deliver telemonitoring, self-management support, and medication management. In applying the LHS-CF, we accessed clinical hemoglobin A1c (HbA1c) data, examined ACDC encounter completion, and interviewed participants and staff. ACDC's development benefited from LHS-CF Enabling Conditions, including a workforce prepared to engage in structured learning and quality improvement; well-developed clinical data systems; strong VA investments in innovation; and a collaborative culture. Consequently, ACDC has contributed to multiple LHS-CF-aligned Bodies of Work, including building new evidence to improve T2D outcomes; translation of effective, telehealth-based T2D care into practice; authentic engagement of partners; ongoing analyses of data to support effective care; and development of new approaches to foster innovation and improvement. In practice, ACDC has reduced HbA1c from 9.7% at baseline to 8.0% at 6months (-1.7%, 95% CI -1.7, -1.6, p < 0.0001), with retention of this benefit at 4years; the program has been delivered with good fidelity and has been valued by participants and staff. The conditions created by VA's LHS facilitate the development, testing, and implementation of programs like ACDC, which promote further LHS success. As a result of its intentional development within VA's LHS, ACDC is well-positioned to improve outcomes for Veterans with poorly controlled T2D nationwide.

Abstract Background Hemolysis detection for whole blood samples is getting increased attention due to its unrecognized impact on results especially when testing is performed at the point-of-care (POC). GEM Premier 7000 with iQM3 (GEM 7000) is a new blood gas system that automatically detects hemolysis in samples using an integrated hemolysis detection module. Here, we investigate potential cross-interference from bilirubin and lipids on hemolysis measurement. Methods GEM 7000 utilizes 570 nm as the primary wavelength for hemoglobin and 610 nm as the secondary wavelength to monitor for sample turbidity. For evaluating cross-interference, native or hemolysate spiked whole blood samples were run in duplicate before and after the addition of bilirubin or lipid interferences. Two levels of lipids (150, 250 mg/dL) were added to three levels of native or hemolysate spiked whole blood (&amp;lt;20, 240 – 260, 430 – 510 mg/dL plasma free hemoglobin [pfHb]). Two levels of bilirubin (9.9 and 24.7 mg/dL) were added to two levels of native or hemolysate spiked whole blood (&amp;lt;20, 96 – 106 mg/dL pfHb). GEM 7000 provides results in hemolysis categories, therefore the agreement of the result category in the presence and absence of interferent was determined. Additionally, pfHb values were extracted to estimate the observed bias with and without interference. pfHb was also taken into consideration when evaluating bias at very low or very high concentrations. Results The presence of lipids or bilirubin showed no cross-interference on the GEM 7000 H index. Difference charts using the extracted pfHb values showed a mean bias of 3.6 and 5.6 mg/dL for all samples at the tested lipid and bilirubin levels, respectively. For hemolyzed samples (pfHb &amp;gt; 20), the presence of lipids and bilirubin resulted in a mean bias of &amp;lt;10%. Conclusion These results demonstrate that the GEM 7000 hemolysis detection is not impacted by the presence of high levels of lipids and bilirubin that tend to co-exist, especially in neonatal samples. By providing integrated hemolysis detection in whole blood measurement, the GEM 7000 system can provide quality and confidence in the interpretation of the results at POC.

Large language models (LLMs) have shown promising capabilities across diverse domains, yet their application to complex clinical prediction tasks remains limited. In this study, we present CARE-AD (Collaborative Analysis and Risk Evaluation for Alzheimer's Disease), a multi-agent LLM-based framework for forecasting Alzheimer's disease (AD) onset by analyzing longitudinal electronic health record (EHR) notes. CARE-AD assigns specialized LLM agents to extract signs and symptoms relevant to AD and conduct domain-specific evaluations-emulating a collaborative diagnostic process. In a retrospective evaluation, CARE-AD achieved higher accuracy (0.53 vs. 0.26-0.45) than baseline single-model approaches in predicting AD risk 10 years prior to the first recorded diagnosis code. These findings highlight the feasibility of using multi-agent LLM systems to support early risk assessment for AD and motivate further research on their integration into clinical decision support workflows.