Savitz, Samuel T.

Introduction/Objectives The rise in screening for social drivers of health (SDoH) in primary care settings has led to questions about effective intervention design. A large primary care practice at Mayo Clinic developed a model centered on a full-time Health Equity Coordinator (HEC) to facilitate clinic-community partnerships in 2022. The goal of this analysis was to assess the model on process outcomes, resolution of SDoH needs, and changes in healthcare utilization. Methods Using a cohort study design, we identified primary care patients with unmet SDoH needs at the intervention clinic and two comparator clinics from March 2022 to December 2023. We assessed process outcomes descriptively and resolution of SDoH needs and healthcare utilization using multivariable regression. Results The HEC contacted 348 patients and conducted 748 meetings with community-based organization staff. While there were decreases for SDoH needs and missed or late cancelled appointments in intervention group, there were no significant difference in resolution of SDoH and healthcare utilization relative to comparator groups. Conclusions The HEC model for multi-stakeholder coordination showed promise in referring patients to services and engaging key stakeholders in a primary care setting. Longer term follow up may be needed to identify changes in SDoH needs and healthcare utilization.

ABSTRACT Introduction Community health workers (CHWs) help patients navigate community resources. CHW programs can improve health outcomes and reduce healthcare utilization, but identifying eligible patients is challenging. We developed an electronic health record (EHR)‐based algorithm trained to predict referrals to a CHW program. Our objective was to evaluate whether an algorithm trained on historical referrals identifies a subgroup of patients that differs from the subgroup identified through questionnaire‐based health‐related social needs (HRSN) screening alone. Methods This analysis used data from Mayo Clinic primary care patients in Southeast Minnesota. We developed a gradient boosted time‐to‐event model that incorporated demographics, diagnoses, HRSN questionnaire responses, and HRSNs derived from clinical notes using natural language processing. We compared the characteristics of algorithm‐identified patients with those identified through questionnaires and assessed the variable importance of model features. Results The algorithm had AUC: 0.87 in the training and 0.92 in the hold out set. Patients flagged by the algorithm tended to use interpreter services more commonly (12.6% vs. 1.3%), have a non‐English preferred language (14.1% vs. 2.3%), and have more instances of health literacy documented (5.36 vs. 1.29) compared to patients identified by the HRSN questionnaire. Only 47.0% of patients flagged by the algorithm had responded to the HRSN questionnaire within 2 years. Variable importance was highest for health literacy and financial strain from unstructured notes. Conclusions The algorithm is a chart‐review prioritization tool for referral to the CHW program for patients that may overcome challenges to identifying patients with HRSNs, especially when questionnaire data is incomplete.

Background Area‐based deprivation indices have been linked to hypertension and diabetes outcomes, but limited information exists about their relative performance. We compared 2 area‐based deprivation measures (Area Deprivation Index and Yost Index) as risk factors for hypertension and diabetes control. Methods This cohort study identified adults with hypertension or diabetes in a 19‐county region of Southeast Minnesota covered by the Rochester Epidemiology Project. The deprivation indices were operationalized as quintiles for state and national ranks. We assessed agreement using the weighted kappa statistic. We used modified Poisson regression to evaluate associations of the deprivation indices with blood pressure or diabetes control in 2022. We compared relative model fit using the area under the receiver operating characteristic curve. Results We identified 67 386 individuals with hypertension and 30 435 with diabetes. There was moderate to substantial agreement between indices, but the agreement was higher for the state rank (weighted kappa 0.68 for both cohorts) than the national rank (0.54 for both). Although the indices were not associated with hypertension control, greater deprivation was associated with worse diabetes control for the Yost Index (relative risk [RR], 1.21 [95% CI, 1.07–1.38]) and Area Deprivation Index (RR, 1.39 [95% CI, 1.12–1.73]) national ranks. The full‐model area under the receiver operating characteristic curve was similar across indices but differed across cohorts with worse fit for hypertension (0.53 for both) than diabetes (0.63 for both). Conclusions Two common area‐based deprivation indices had similar strength of association with diabetes control, suggesting that either index would be appropriate for researching and developing interventions for diabetes control.

AbstractPurposeThe Collaborative care model (CoCM) is the leading model for integrating behavioral health into primary care for patients with major depressive disorder (MDD). However, CoCM requires engagement and ongoing participation. We aimed to assess whether two area‐based measures, the area‐deprivation index (ADI) and rurality, were associated with enrollment, participation, and outcomes with CoCM.MethodsThis was an observational analysis of Mayo Clinic patients eligible for CoCM: adults aged ≥18 years, empaneled in primary care, and with a PHQ‐9 of ≥10. We operationalized ADI as quintiles with Q1 being least deprived and Q5 being most deprived and rurality using RUCA codes with two categories: urban and rural. We evaluated enrollment in CoCM, drop out defined by leaving the program early, the count and type of contacts with the care coordinator, and clinical improvement measured using the PHQ‐9.FindingsWe identified 54,030 individuals with 16,532 (30.6%) residing in rural areas and 11,122 (20.6%) residing in the most deprived ADI quintile (Q5). Living in a rural area was associated with lower enrollment in CoCM (–2.3 percentage points [95% confidence interval (CI): –2.5, 2.2]), longer length in CoCM (18.6 days [95% CI: 5.7, 31.5]), more contacts with the care coordinator (1.1 contacts [95% CI: 0.2, 2.0]), and worse response and remission. In contrast, ADI Q5 was only associated with worse response and remission.ConclusionsRurality was associated with lower enrollment, greater engagement, and worse clinical outcomes. More work may be needed to address enrollment barriers for individuals living in rural areas to improve clinical outcomes.

ABSTRACTObjectiveTo build consensus from different stakeholders on the content and digital interactivity of a shared decision making (SDM) tool to guide patient‐clinician collaborative conversations about treatment for bipolar depression.Patients and MethodsThis study was conducted between September 2022 and September 2023. In the first phase, we conducted a narrative review on pharmacological, neuromodulatory, and psychotherapeutic treatments available for bipolar depression. For each option, we assessed treatment effectiveness, side effects, need for clinical monitoring, clinical considerations, and cost. In the second phase, committees comprised of mental health clinicians and people with lived experience with bipolar disorder from a national mood disorders advocacy group identified key features and content to be included in the SDM tool. A thematic analysis of these data then contributed to populate the SDM tool.ResultsTwenty‐one evidence‐based treatment options identified from the narrative review were presented to the four convened committees (clinical, lived experience, peer council, and integrated steering committees) in five meetings. After discussions and thematic analysis, 12 outcomes of interest for each of the therapeutic interventions were delineated for inclusion in a web‐based SDM tool. A detailed description of the development process and SDM tool is provided.ConclusionThis is the first partnership between clinicians and persons with bipolar disorder lived experience charged with the development of a patient‐centered SDM tool to integrate clinical evidence with real‐world clinical practice and patient preferences for use during clinical encounters. Human‐centered design and testing in real clinical encounters will now contribute to refining the instrument and readying it for rigorous evaluation and implementation in practice.

AbstractThe Adult Medical Care Coordination program (“the program”) was implemented at Mayo Clinic to promote patient self‐management and improve 30‐day unplanned readmission for patients with high risk for readmission after hospital discharge. This study aimed to evaluate the impact of the program compared to usual care using a pragmatic, stepped wedge cluster randomized trial (“stepped wedge trial”). However, several challenges arose including large differences between the study arms. Our goal is to describe the challenges and present lessons learned on how to overcome such challenges and generate evidence to support practice decisions. We describe the challenges encountered during the trial, the approach to addressing these challenges, and lessons learned for other learning health system researchers facing similar challenges. The trial experienced several challenges in implementation including several clinics dropping from the study and care disruptions due to COVID‐19. Additionally, there were large differences in the patient population between the program and usual care arms. For example, the mean age was 76.8 for the program and 68.1 for usual care. Due to these differences, we adapted the methods using the propensity score matching approach that is traditionally applied to observational designs and adjusted for differences in observable characteristics. When conducting pragmatic research, researchers will encounter factors beyond their control that may introduce bias. The lessons learned include the need to weigh the tradeoffs of pragmatic design elements and the potential value of adaptive designs for pragmatic trials. Applying these lessons would promote the successful generation of evidence that informs practice decisions.

BACKGROUND: Following regulatory approval, medical devices may be used “off-label.” Patent foramen ovale (PFO) closure is indicated to reduce recurrent stroke but has been proposed for other indications, including migraine, transient ischemic attack, and diving decompression illness. We sought to evaluate PFO closure rates and indications relative to the timing of regulatory approval and publication of key randomized trials. METHODS: We performed a retrospective cohort study using the OptumLabs Data Warehouse of US commercial insurance enrollees from 2006 to 2019. We quantified PFO closure among individuals with ≥2 years of preprocedure coverage to establish indications, classified hierarchically as stroke/systemic embolism, migraine, transient ischemia attack, or other. RESULTS: We identified 5315 patients undergoing PFO closure (51.8% female, 29.2%≥60 years old), which increased from 4.75 per 100 000 person-years in 2006 to 6.60 per 100 000 person-years in 2019. Patients aged ≥60 years accounted for 29.2% of closures. Procedure volumes corresponded weakly with supportive clinical publications and device approval. Among patients with PFO closure, 58.6% underwent closure for stroke/systemic embolism, 10.2% for transient ischemia attack, 8.8% for migraine, and 22.4% for other indications; 17.6% of patients had atrial fibrillation at baseline; and 11.9% developed atrial fibrillation postprocedure. Those aged ≥60 years and male were less likely to undergo closure for migraine than stroke/systemic embolism. CONCLUSIONS: From 2006 to 2019, PFO closure use was consistently low and corresponded weakly with clinical trial publications and regulatory status. Nearly half of patients underwent PFO closure for indications unapproved by the Food and Drug Administration. Regulators and payers should coordinate mechanisms to promote utilization for approved indications to ensure patient safety and should facilitate clinical trials for other possible indications.

AbstractRationaleExisting literature describing differences in survival following percutaneous coronary intervention (PCI) by patient sex, race‐ethnicity and the role of socioeconomic characteristics (SEC) is limited.Aims and ObjectivesEvaluate differences in 1‐year survival after PCI by sex and race‐ethnicity, and explore the contribution of SEC to observed differences.MethodsUsing a 20% sample of Medicare claims data for beneficiaries aged 65+, we identified fee‐for‐service patients who received PCI from 2007 to 2015. We performed logistic regression to assess how sex and race‐ethnicity relate to procedural indication, inpatient versus outpatient setting, and 1‐year mortality. We evaluated whether these relationships are moderated by sequentially controlling for factors including age, comorbidities, presence of acute myocardial infarction (AMI), county SEC, medical resource availability and inpatient versus outpatient procedural status.ResultsWe identified 300,491 PCI procedures, of which 94,863 (31.6%) were outpatient. There was a significant transition to outpatient PCI during the study period, especially for men compared with women and White patients compared with Black patients. Black patients were 3.50 percentage points (p < 0.001) and women were 3.41 percentage points (p < 0.001) more likely than White and male patients to undergo PCI at the time of AMI, which typically occurs in the inpatient setting. Controlling for age and calendar year, Black patients were 2.87 percentage points more likely than non‐Hispanic White patients to die within 1 year after PCI. After controlling for Black–White differences in comorbidities, the differences in 1‐year mortality decreased to 0.95 percentage points, which then became nonsignificant when further controlling for county resources and state of residence.ConclusionWomen were more likely to experience PCI in the setting of AMI and had less transition to outpatient care during the period. Black patients experienced higher 1‐year mortality following PCI, which is explained by differences in baseline comorbidities, county medical resources, and state of residence.

The objective was to assess the value of routinely collected patient-reported health-related social needs (HRSNs) measures for predicting utilization and health outcomes. The authors identified Mayo Clinic patients with cancer, diabetes, or heart failure. The HRSN measures were collected as part of patient-reported screenings from June to December 2019 and outcomes (hospitalization, 30-day readmission, and death) were ascertained in 2020. For each outcome and disease combination, 4 models were used: gradient boosting machine (GBM), random forest (RF), generalized linear model (GLM), and elastic net (EN). Other predictors included clinical factors, demographics, and area-based HRSN measures—area deprivation index (ADI) and rurality. Predictive performance for models was evaluated with and without the routinely collected HRSN measures as change in area under the curve (AUC). Variable importance was also assessed. The differences in AUC were mixed. Significant improvements existed in 3 models of death for cancer (GBM: 0.0421, RF: 0.0496, EN: 0.0428), 3 models of hospitalization (GBM: 0.0372, RF: 0.0640, EN: 0.0441), and 1 of death (RF: 0.0754) for diabetes, and 1 model of readmissions (GBM: 0.1817), and 3 models of death (GBM: 0.0333, RF: 0.0519, GLM: 0.0489) for heart failure. Age, ADI, and the Charlson comorbidity index were the top 3 in variable importance and were consistently more important than routinely collected HRSN measures. The addition of routinely collected HRSN measures resulted in mixed improvement in the predictive performance of the models. These findings suggest that existing factors and the ADI are more important for prediction in these contexts. More work is needed to identify predictors that consistently improve model performance.