Charisse Madlock-Brown

Background: Social isolation and social support are clinically important but inconsistently represented in structured electronic health record data. Their identification from clinical narratives is complicated by nuanced, context-dependent language and by variation in how social connectedness is documented across clinical settings, patient populations, and geographic locations. Models intended for health care use must therefore distinguish social isolation from documented support, negation, ambiguous social circumstances, and non-social clinical uses of similar terminology while remaining robust to differences in documentation practices. Objective: This study aimed to develop and evaluate locally deployable, fine-tuned language models for context-aware classification of social isolation, social support, and contextually similar nonrelevant clinical text across multiple health systems. Materials and Methods: We conducted a retrospective study using electronic health record data from 326,847 adults aged 50 years or older who received care at 3 Tennessee health systems from 2020 to 2023. Candidate clinical text was identified using social-environment ICD-10-CM Z codes and an established social isolation lexicon. Human annotators classified 9,748 spans from 9,578 clinical notes contributed by 6,881 patients as social isolation, social support, or no social isolation. The annotated corpus included social isolation and support concepts including negations, infection-control and microbiology terminology, institutional living arrangements, and other contextually ambiguous examples. Uncertainty-based active learning was used to prioritize informative spans for annotation. FLAN-T5-Large, BERT, RoBERTa, and Gemma-2-2B were fine-tuned for 3-class span classification. Full fine-tuning and parameter-efficient fine-tuning using low-rank adaptation approaches were evaluated. Performance was assessed using site-based internal-external cross-validation, with macro-F1 as the primary metric. Results: The annotated corpus included 2,318 social isolation spans, 638 social support spans, and 6,792 spans classified as no social isolation. Fully fine-tuned FLAN-T5-Large achieved the highest and most consistent performance, with a mean macro-F1 of 0.92 (SD 0.04). Class-specific F1 scores were 0.91 (SD 0.03) for social isolation, 0.90 (SD 0.04) for social support, and 0.94 (SD 0.05) for no social isolation. Gemma-2-2B achieved a macro-F1 of 0.89 (SD 0.10), compared with 0.80 (SD 0.21) for RoBERTa and 0.77 (SD 0.17) for BERT. Full fine-tuning also outperformed parameter-efficient fine-tuning in the corresponding held-out evaluation. Remaining errors primarily involved patients who lived alone but received assistance, the presence of family members without meaningful support, grief-related language, and non-social clinical uses of isolation terminology. Conclusions: Fine-tuned language models can distinguish social isolation, social support, and contextually similar but unrelated language across heterogeneous clinical settings. FLAN-T5-Large provided the strongest and most stable performance, including for the less frequent social support class. The 3-class annotation framework, inclusion of difficult negative examples, and multisite evaluation offer a practical approach for converting nuanced social-connectedness documentation into structured data. Further work should evaluate end-to-end identification in unselected clinical notes and prospective integration into clinical and population health workflows.

BACKGROUND Recent advancements in large language models (LLMs) have improved the identification of social isolation from clinical narratives, which vary widely in linguistic patterns and documentation practices. However, LLMs fine-tuned on a single dataset often show reduced performance when applied to different healthcare settings or clinical note types. Rigorous evaluation of cross-site generalizability and fairness is therefore essential to ensure accurate and equitable detection of social isolation across diverse populations and clinical contexts. OBJECTIVE This study aimed to evaluate a span-level fine-tuned FLAN-T5-Large model for extracting ‘social isolation’ indicators from unstructured clinical text and to assess its generalizability and fairness across diverse populations and healthcare data sources. METHODS A total of 2,967 unique annotated spans from 9,578 clinical notes across three healthcare systems were used to fine-tune an FLAN-T5-Large model using a contextualized span classification framework. A Gemma‑2-2B model was evaluated in a sensitivity analysis to assess architecture‑related performance differences. Performance was assessed using precision, recall, and macro F1. Fairness was evaluated across demographic variables, social vulnerability strata, and note types using statistical parity difference (SPD) and equal opportunity difference (EOD). RESULTS Incorporating contextual windows around annotated spans improved macro-F1 from 0.90 to 0.94 during validation. In full note evaluation across 900 manually reviewed notes, FLAN-T5-Large achieved high recall for social isolation (0.94 – 0.98) and macro F1 values ranging from 0.69 to 0.81 across sites. Fairness analysis showed generally consistent performance across age, gender, race, and social vulnerability groups, with equitable sensitivity (EOD 0.02 – 0.04) and moderate variation in positive prediction rates (SPD). Performance variability was strongly associated with documentation type, with note type driving substantially greater variability in both performance and fairness metrics than patient demographic factors. CONCLUSIONS Fine-tuned FLAN‑T5-Large demonstrated strong capability in detecting ‘social isolation’ from clinical narratives, while maintaining sensitivity parity across subgroups. The observed heterogeneity was largely driven by documentation context rather than by patient characteristics, highlighting the importance of note type-aware evaluation in clinical NLP. These findings support the use of instruction‑tuned LLMs for equitable extraction of social context information from EHR text.

The COVID-19 pandemic impacted nearly every aspect of modern society. The medical industry was underprepared for the pandemic-related shifts in emergency d

This study assesses the relationship between BMI and healthcare burden, stratified by race and healthcare utilization, among middle-aged patients in the US. We used data from the Cerner HealthFacts database for 2016–2017 as our study period. We employed regression analysis to maximize the area under the curve (AUC) of the receiver operating characteristic (ROC) curve. This analysis assessed the relationship between BMI and healthcare burden using logistic regression and identified optimal BMI cutoffs stratified across different racial groups and levels of healthcare utilization by maximizing the AUC of the ROC curve. BMI was normalized between 0 and 1, and odds ratios were interpreted as the change in odds of a CCI score greater than zero associated with a full-range (min-to-max) increase in BMI. Results indicated that BMI was strongly associated with CCI across regular, low, and non-utilization cohorts. Specifically, in the regular healthcare utilizer cohort, a min-to-max increase in BMI was associated with a significantly higher likelihood of a CCI score greater than zero. The odds ratios for a min-to-max BMI change were notably high for the White and Asian/Pacific Islander groups (24.3 and 38.18, respectively), compared to 5.04 and 3.82 for the Black and Native American groups. The AUC analysis revealed the highest value for the Asian/Pacific Islander cohort (0.71) and the lowest for the Black cohort (0.63), with optimal BMI cutoffs identified as 34 for African Americans, 35 for American Indians/Alaska Natives, 32 for Whites, and 27 for Asians/Pacific Islanders. The findings underscore the necessity of stratifying patients by healthcare utilization, particularly as regular utilizers in the White and Asian/Pacific Islander populations had lower BMI cutoffs. This study advocates for a paradigm shift in obesity diagnosis, emphasizing the need for refined metrics and additional research on BMI’s role in healthcare burden across diverse populations.

Background Multimorbidity with diabetes mellitus (DM and ≥ 1 chronic disease) presents challenges to maintaining adequate DM control. Objective This study evaluates the risk of DM-related complications associated with various multimorbidity/DM patterns for patients seen in community-based health centers (CHCs). Design Retrospective cohort study from the ADVANCE multi-state practice-based clinical data network. Participants Study included data from 132,765 patients age ≥ 45 years with DM seen in 2493 CHCs across 26 states from 10/01/2015 to 12/31/2019. Main Measures We assessed accrual of conditions and risk of experiencing DM complications during follow-up. Primary outcome of DM complication was categorized into acute, microvascular, microvascular (end-stage), macrovascular, or other. Key exposures included mutually exclusive multimorbidity categories: (1) DM + cardiometabolic, (2) DM + other somatic, (3) DM + mental, (4) DM + mental + somatic. Key Results At baseline, 17.2% of patients had DM only, 55.0% had DM + cardiometabolic multimorbidity, 2.3% had DM + other somatic multimorbidity, 3.0% had DM + mental multimorbidity, and 22.5% had DM + mental + somatic multimorbidity. Most DM-only patients (76.5%) developed multimorbidity with DM by study end. Compared with DM-only, adjusted risk differences of DM complications ranged from 1.4% (acute) to 8.8% (microvascular). DM + mental + somatic multimorbidity was associated with 13.4% (95%CI 12.8–14.1%) higher adjusted risk of experiencing any DM complication. Conclusions CHCs care for increasingly complex populations of patients with DM. Tailoring disease management strategies to address comorbid cardiovascular and/or mental health conditions may be important to prevent acute, microvascular, and macrovascular complications in these settings.

Background: Nirmatrelvir with ritonavir (Paxlovid) is indicated for patients with Coronavirus Disease 2019 (COVID-19) who are at risk for progression to severe disease due to the presence of one or more risk factors. Millions of treatment courses have been prescribed in the United States alone. Paxlovid was highly effective at preventing hospitalization and death in clinical trials. Several studies have found a protective association in real-world data, but they variously used less recent study periods, correlational methods, and small, local cohorts. Their estimates also varied widely. The real-world effectiveness of Paxlovid remains uncertain, and it is unknown whether its effect is homogeneous across demographic strata. This study leverages electronic health record data in the National COVID Cohort Collaborative’s (N3C) repository to investigate disparities in Paxlovid treatment and to emulate a target trial assessing its effectiveness in reducing severe COVID-19 outcomes. Methods and findings: This target trial emulation used a cohort of 703,647 patients with COVID-19 seen at 34 clinical sites across the United States between April 1, 2022 and August 28, 2023. Treatment was defined as receipt of a Paxlovid prescription within 5 days of the patient’s COVID-19 index date (positive test or diagnosis). To emulate randomization, we used the clone-censor-weight technique with inverse probability of censoring weights to balance a set of covariates including sex, age, race and ethnicity, comorbidities, community well-being index (CWBI), prior healthcare utilization, month of COVID-19 index, and site of care provision. The primary outcome was hospitalization; death was a secondary outcome. We estimated that Paxlovid reduced the risk of hospitalization by 39% (95% confidence interval (CI) [36%, 41%]; p < 0.001), with an absolute risk reduction of 0.9 percentage points (95% CI [0.9, 1.0]; p < 0.001), and reduced the risk of death by 61% (95% CI [55%, 67%]; p < 0.001), with an absolute risk reduction of 0.2 percentage points (95% CI [0.1, 0.2]; p < 0.001). We also conducted stratified analyses by vaccination status and age group. Absolute risk reduction for hospitalization was similar among patients that were vaccinated and unvaccinate, but was much greater among patients aged 65+ years than among younger patients. We observed disparities in Paxlovid treatment, with lower rates among black and Hispanic or Latino patients, and within socially vulnerable communities. This study’s main limitation is that it estimates causal effects using observational data and could be biased by unmeasured confounding. Conclusions: In this study of Paxlovid’s real-world effectiveness, we observed that Paxlovid is effective at preventing hospitalization and death, including among vaccinated patients, and particularly among older patients. This remains true in the era of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Omicron subvariants. However, disparities in Paxlovid treatment rates imply that the benefit of Paxlovid’s effectiveness is not equitably distributed.

Individually, the COVID-19 and HIV pandemics have differentially impacted minoritized groups due to the role of social determinants of health (SDoH) in the U.S. Little is known how the collision of these two pandemics may have exacerbated adverse health outcomes. We evaluated county-level SDoH and associations with hospitalization after a COVID-19 diagnosis among people with (PWH) and without HIV (PWOH) by racial/ethnic groups. We used the U.S. National COVID Cohort Collaborative (January 2020-November 2023), a nationally-sampled electronic health record repository, to identify adults who were diagnosed with COVID-19 with HIV (n = 22,491) and without HIV (n = 2,220,660). We aggregated SDoH measures at the county-level and categorized racial/ethnic groups as Non-Hispanic (NH) White, NH-Black, Hispanic/Latinx, NH-Asian and Pacific Islander (AAPI), and NH-American Indian or Alaskan Native (AIAN). To estimate associations of county-level SDoH with hospitalization after a COVID-19 diagnosis, we used multilevel, multivariable logistic regressions, calculating adjusted relative risks (aRR) with 95% confidence intervals (95% CI). COVID-19 related hospitalization occurred among 11% of PWH and 7% of PWOH, with the highest proportion among NH-Black PWH (15%). In evaluating county-level SDoH among PWH, we found higher average household size was associated with lower risk of COVID-19 related hospitalization across racial/ethnic groups. Higher mean commute time (aRR: 1.76; 95% CI 1.10–2.62) and higher proportion of adults without health insurance (aRR: 1.40; 95% CI 1.04–1.84) was associated with a higher risk of COVID-19 hospitalization among NH-Black PWH, however, NH-Black PWOH did not demonstrate these associations. Differences by race and ethnicity exist in associations of adverse county-level SDoH with COVID-19 outcomes among people with and without HIV in the U.S.

Background: Leveraging the National COVID-19 Cohort Collaborative (N3C), a nationally sampled electronic health records repository, we explored associations between individual-level social determinants of health (SDoH) and COVID-19-related hospitalizations among racialized minority people with human immunodeficiency virus (HIV) (PWH), who have been historically adversely affected by SDoH. Methods: We retrospectively studied PWH and people without HIV (PWoH) using N3C data from January 2020 to November 2023. We evaluated SDoH variables across three domains in the Healthy People 2030 framework: (1) healthcare access, (2) economic stability, and (3) social cohesion with our primary outcome, COVID-19-related hospitalization. We conducted hierarchically nested additive and adjusted mixed-effects logistic regression models, stratifying by HIV status and race/ethnicity groups, accounting for age, sex, comorbidities, and data partners. Results: Our analytic sample included 280,441 individuals from 24 data partner sites, where 3,291 (1.17%) were PWH, with racialized minority PWH having higher proportions of adverse SDoH exposures than racialized minority PWoH. COVID-19-related hospitalizations occurred in 11.23% of all individuals (9.17% among PWH, 11.26% among PWoH). In our initial additive modeling, we observed that all three SDoH domains were significantly associated with hospitalizations, even with progressive adjustments (adjusted odds ratios [aOR] range 1.36–1.97). Subsequently, our HIV-stratified analyses indicated economic instability was associated with hospitalization in both PWH and PWoH (aOR range 1.35–1.48). Lastly, our fully adjusted, race/ethnicity-stratified analysis, indicated access to healthcare issues was associated with hospitalization across various racialized groups (aOR range 1.36–2.00). Conclusion: Our study underscores the importance of assessing individual-level SDoH variables to unravel the complex interplay of these factors for racialized minority groups.