Development of CARRA/PReS-endorsed consensus Core and Expanded Datasets in childhood-onset systemic lupus erythematosus for international registry-based research.

OBJECTIVES: Childhood-onset systemic lupus erythematosus (cSLE), representing 15%-20% of individuals with SLE, has been difficult to study globally due to differences between registries. This initiative, supported by Childhood Arthritis Rheumatology Research Alliance (CARRA) and Paediatric Rheumatology European Society (PReS), aims to create Core and Expanded cSLE Datasets to standardise and enhance research worldwide. METHODS: 21 international cSLE experts and 4 patients participated in a Delphi process (questionnaires, 2 topic-specific focus groups and 3 virtual consensus meetings) to create 2 standardised cSLE datasets. The Core cSLE Dataset was designed to include data essential to meaningful clinical research across many settings. The Expanded cSLE Dataset was designed for centres able to consistently collect data to address broader research questions. Final data items for the Core and Expanded datasets were determined by consensus defined as >80% agreement) using an adapted nominal group technique and voting. RESULTS: The resulting Core cSLE Dataset contains 46 items, including demographics, clinical features, laboratory results, medications and significant adverse events. The Expanded cSLE Dataset adds 26 additional items and includes patient-reported outcomes. Consensus was also achieved regarding the frequency and time points for data collection: baseline, quarterly follow-up visits, annually and flare visits. CONCLUSION: Standardised Core and Expanded cSLE Datasets for registry-based international cSLE research were defined through the consensus of global experts and patient/caregiver representatives, endorsed by CARRA and PReS. These datasets incorporate disease-specific and patient-specific features, optimised for diverse settings to facilitate international collaborative research for children and adolescents with SLE worldwide.

Factors Associated With COVID-19 Vaccine Hesitancy in Rheumatology Outpatients in New York City.

OBJECTIVE: The aim of this study was to measure COVID-19 vaccine hesitancy among rheumatology outpatients from an early COVID-19 "hotspot" during the initial period of vaccine availability. METHODS: In March 2021, a Web-based survey was sent to 7505 adults seen at a Rheumatology Division in New York City. We evaluated characteristics associated with 3 categories of COVID-19 vaccination status: declined, undecided, and willing/already received. We used multinomial logistic regression models to calculate relative risk ratios assessing predictors of vaccination status. RESULTS: Among 2384 (32%) respondents (80% female, 87% White, 59% with systemic rheumatic disease), 2240 (94.0%) were willing/already received COVID-19 vaccination, 88 (3.7%) were undecided, and 56 (2.3%) declined. Compared with those willing/already vaccinated, those declining or undecided were younger, more likely identified as Black or Hispanic/Latinx, and had lower household income and educational attainment. Immunosuppressive medication use did not differ among groups. After multivariable adjustment, every 1-year increase in age was associated with a 0.96 lower relative risk of declining or being undecided versus willing/already vaccinated. Respondents identifying as Black versus White had a higher relative risk ratio of being undecided (4.29 [95% confidence interval, 1.96-9.36]), as did those identifying as Hispanic/Latinx versus non-Hispanic/non-Latinx (2.81 [95% confidence interval, 1.29-6.09]). Those declining vaccination were least likely to believe in general vaccine importance or the safety and efficacy of the COVID-19 vaccine. CONCLUSIONS: Among rheumatology patients in New York City with and without systemic rheumatic disease, COVID-19 vaccine uptake was high after its initial availability. Sociodemographic but not medication-related factors were associated with vaccine hesitancy; these findings can inform future rheumatology vaccination programs.

The link between post-traumatic stress disorder and systemic lupus erythematosus.

Systemic lupus erythematosus (SLE) is a heterogeneous, multisystem autoimmune disorder characterized by unpredictable disease flares. Although the pathogenesis of SLE is complex, an epidemiologic link between posttraumatic stress disorder (PTSD) and the development of SLE has been identified, suggesting that stress-related disorders alter the susceptibility to SLE. Despite the strong epidemiologic evidence connecting PTSD and SLE, gaps remain in our understanding of how the two may be connected. Perturbations in the autonomic nervous system, neuroendocrine system, and at the genomic level may cause and sustain immune dysregulation that could lower the threshold for the development and propagation of SLE. We first describe shared risk factors for SLE and PTSD. We then describe potential biological pathways which may facilitate excessive inflammation in the context of PTSD. Among those genetically predisposed to SLE, systemic inflammation that accompanies chronic stress may fan the flames of smoldering SLE by priming immune pathways. Further studies on the connection between trauma and inflammation will provide important data on pathogenesis, risk factors, and novel treatments for SLE.

Evaluation and management of pericarditis in rheumatic diseases.

ABSTRACT: This review summarizes the evaluation for underlying rheumatic conditions in patients presenting with acute pericarditis, treatment considerations for specific rheumatic conditions, and the role of imaging in diagnosis and monitoring. Pericarditis may be one of the initial presentations of a rheumatic disease or identified in a patient with known rheumatic disease. There is also growing evidence for using anti-inflammatory and immunosuppressive agents for treating recurrent pericarditis, which can overlap with the treatment of rheumatic diseases.

Exploration of machine learning methods to predict systemic lupus erythematosus hospitalizations.

OBJECTIVES: Systemic lupus erythematosus (SLE) is a heterogeneous disease characterized by disease flares which can require hospitalization. Our objective was to apply machine learning methods to predict hospitalizations for SLE from electronic health record (EHR) data. METHODS: We identified patients with SLE in a longitudinal EHR-based cohort with ≥2 outpatient rheumatology visits between 2012 and 2019. We applied multiple machine learning methods to predict hospitalizations with a primary diagnosis code for SLE, including decision tree, random forest, naive Bayes, logistic regression, and an ensemble method. Candidate predictors were derived from structured EHR features, including demographics, laboratory tests, medications, ICD-9/10 codes for SLE manifestations, and healthcare utilization. We used two approaches to assess these variables over longitudinal follow-up, including the incorporation of lagged features to capture changes over time of clinical data. The performance of each model was evaluated by overall accuracy, the F statistic, and the area under the receiver operator curve (AUC). RESULTS: We identified 1996 patients with SLE. 4.6% were hospitalized for SLE in their most recent year of follow-up. Random forest models had highest performance in predicting SLE hospitalizations, with AUC 0.751 and AUC 0.772 for two approaches (averaging and progressive), respectively. The leading predictors of SLE hospitalizations included dsDNA positivity, C3 level, blood cell counts, and inflammatory markers as well as age and albumin. CONCLUSION: We have demonstrated that machine learning methods can predict SLE hospitalizations. We identified key predictors of these events including known markers of SLE disease activity; further validation in external cohorts is warranted.

Comparative risks of cardiovascular disease events among SLE patients receiving immunosuppressive medications.

OBJECTIVES: SLE patients have elevated cardiovascular disease (CVD) risk, but it is unclear whether this risk is affected by choice of immunosuppressive drug. We compared CVD risks among SLE patients starting MMF, CYC or AZA. METHODS: Using Medicaid Analytic eXtract (2000-2012), adult SLE patients starting MMF, CYC or AZA were identified and propensity scores (PS) were estimated for receipt of MMF vs CYC and MMF vs AZA. We examined rates of first CVD event (primary outcome), all-cause mortality, and a composite of first CVD event and all-cause mortality (secondary outcomes). After 1:1 PS-matching, Fine-Gray regression models estimated subdistribution hazard ratios (HRs.d.) for risk of CVD events. Cox regression models estimated HRs for all-cause mortality. The primary analysis was as-treated; 6- and 12-month intention-to-treat (ITT) analyses were secondary. RESULTS: We studied 680 PS-matched pairs of patients with SLE initiating MMF vs CYC and 1871 pairs initiating MMF vs AZA. Risk of first CVD event was non-significantly reduced for MMF vs CYC [HRs.d 0.72 (95% CI: 0.37, 1.39)] and for MMF vs AZA [HRs.d 0.88 (95% CI: 0.59, 1.32)] groups. In the 12-month ITT, first CVD event risk was lower among MMF than AZA new users [HRs.d 0.68 (95% CI: 0.47, 0.98)]. CONCLUSION: In this head-to-head PS-matched analysis, CVD event risks among SLE patients starting MMF vs CYC or AZA were not statistically reduced except in one 12-month ITT analysis of MMF vs AZA, suggesting longer-term use may convey benefit. Further studies of potential cardioprotective benefit of MMF are necessary.

Patient experiences and strategies for coping with SLE: A qualitative study.

OBJECTIVE: This study explored challenges that patients with systemic lupus erythematosus (SLE) and childhood-onset SLE (cSLE) face to identify modifiable influences and coping strategies in patient experiences. METHODS: Participants were recruited from two academic medical centers through a Lupus Registry of individuals ≥18 years old and ≥4 1997 ACR classification criteria for SLE and a centralized data repository of cSLE patients, and participated in three focus groups. Transcripts were coded thematically and adjudicated by two independent reviewers. RESULTS: Thirteen adults, 7 (54%) with cSLE, participated in focus groups. Themes were categorized into two domains: (1) challenges with SLE diagnosis and management; and (2) patient coping strategies and modifiable factors of the SLE experience. Participants identified five primary challenges: diagnostic odyssey, public versus private face of SLE, SLE-related stresses, medication adherence, and transitioning from pediatric to adult care. Coping strategies and modifiable factors included social support, open communication about SLE, and strong patient-provider relationships. Several participants highlighted positive lessons learned through their experiences with SLE, including empathy, resilience, and self-care skills. CONCLUSIONS: Patients with cSLE and SLE identified common challenges, modifying influences and coping strategies based on personal experiences. A strong patient-provider relationship and trust in the medical team emerged as key modifiable factors. Deriving optimism from experiences with SLE was unique to several patients diagnosed as children or young adults. Leveraging factors that improved the participants' experiences living with SLE may be used in future studies to address vulnerabilities in care.

Initial disease severity, cardiovascular events and all-cause mortality among patients with systemic lupus erythematosus.

OBJECTIVE: SLE is associated with high risks of cardiovascular disease (CVD) and mortality, and has a wide spectrum of presentations. We investigated whether SLE severity at diagnosis was associated with CVD or mortality risk. METHODS: Within Medicaid (2000-10), we identified patients 18-65 years of age with incident SLE. Initial SLE severity was classified-mild, moderate, or severe-during the baseline year prior to the start of follow-up (incident index date) using a published algorithm based on SLE-related medications and diagnoses. Patients were followed from the index date to the first CVD event or death, disenrollment, loss to follow-up or end of follow-up period. Cox and Fine-Gray regression models, adjusted for demographics and comorbidities accounting for the competing risk of death (for CVD), estimated CVD and mortality risks by baseline SLE severity. RESULTS: Of 15 120 incident SLE patients, 48.7% had mild initial SLE severity, 33.9% moderate and 17.4% severe. Mean (s.d.) follow-up was 3.3 (2.4) years. After multivariable adjustment, CVD subdistribution hazard ratios (HRSD) were higher for initially severe [HRSD 1.64 (95% CI 1.32, 2.04)] and moderate [HRSD 1.19 (95% CI 1.00, 1.41)] SLE vs mild SLE. Mortality HRs were also higher for initially severe [HR 3.11 (95% CI 2.49, 3.89)] and moderate [HR 1.61 (95% CI 1.29, 2.01)] SLE vs mild SLE. CONCLUSION: SLE patients with high initial severity had elevated mortality and CVD events risks compared with those who presented with milder disease. This has implications for clinical care and risk stratification of newly diagnosed SLE patients.

Associations between daily alcohol consumption and systemic lupus erythematosus-related cytokines and chemokines among US female nurses without SLE.

OBJECTIVES: Moderate alcohol consumption has been associated with decreased systemic lupus erythematosus risk, but the biologic basis for this association is unknown. We aimed to determine whether moderate alcohol consumption was associated with lower concentrations of systemic lupus erythematosus-associated chemokines/cytokines in an ongoing cohort of female nurses without systemic lupus erythematosus, and whether the association was modified by the presence of systemic lupus erythematosus-related autoantibodies. METHODS: About 25% of participants from the Nurses' Health Study (n = 121,700 women) and Nurses' Health Study 2 (n = 116,429) donated a blood sample; of these, 1177 women were without systemic lupus erythematosus at time of donation. Cumulative average and current (within 4 years) intakes of beer, wine or liquor were assessed from pre-blood draw questionnaires. Chemokine/cytokine concentrations (stem cell factor, B-lymphocyte stimulator, interferon-inducible protein-10, interferon-alpha, interleukin-10) and antibodies against dsDNA and extractable nuclear antigens were obtained using enzyme-linked immunosorbent assays. Antinuclear antibodies were detected by indirect immunofluorescence on HEp-2 cells. RESULTS: At blood draw, the women's mean age was 56 years and 22% were antinuclear antibody positive; 36% were African-American. About half (46%) reported consuming 0-5 g/day of alcohol. Stem cell factor levels were 0.5% lower (p < 0.0001) for every gram per day increase in cumulative average alcohol consumption. Women who consumed >5 g/day had mean stem cell factor levels 7% lower (p = 0.002) than non-drinkers. Other cytokines were not significantly associated with alcohol intake. Autoantibody status did not modify observed associations. CONCLUSION: In this study of female nurses, moderate alcohol consumption was associated with lower stem cell factor levels, suggesting a plausible mechanism through which alcohol may lower systemic lupus erythematosus risk might be by decreasing circulating stem cell factor.

Comparison of an administrative algorithm for SLE disease severity to clinical SLE Disease Activity Index scores.

Systemic lupus erythematosus (SLE) severity, reflecting both disease intensity and duration, is heterogeneous making it challenging to study in administrative databases where severity may confound or mediate associations with outcomes. Garris et al. developed an administrative claims-based algorithm employing claims over a 1-year period to classify SLE severity as mild, moderate or severe. We sought to compare this administrative algorithm to a measure of SLE activity, the SLE Disease Activity Index-2000 (SLEDAI-2K) score at clinical visits. We identified 100 SLE patients followed in the Brigham and Women's Hospital (BWH) Lupus Center (in 2008-2010) with SLEDAI-2K scores at each visit over a 1-year period per person. We obtained data for the Garris algorithm for the same year per subject. We compared Garris SLE severity to the highest SLEDAI-2K in that year, with SLEDAI-2K categories of mild < 3, moderate 3-6, and severe > 6. We compared classification using weighted kappa statistics, and positive and negative predictive values (PPV, NPV). We also assessed the binary comparison of mild vs. moderate/severe. We calculated sensitivity, specificity, and McNemar's test. We analyzed 377 SLEDAI-2K assessments (mean 3.8 [SD 2.6] per subject/year). For classifying moderate/severe vs. mild SLE severity, the sensitivity was 85.7%, specificity 67.6%, PPV 81.8% and NPV 73.5%. The Garris algorithm for classifying SLE severity in administrative datasets had moderate agreement for classification of mild vs. moderate/severe SLE activity assessed by SLEDAI-2K assessments in an academic lupus center. It may be a useful tool for classifying SLE severity in administrative database studies.

Association between socioeconomic status and pain, function and pain catastrophizing at presentation for total knee arthroplasty.

BACKGROUND: Patients with higher socioeconomic status (SES) are shown to have better total knee arthroplasty (TKA) outcomes compared to those with lower SES. The relationship between SES and factors that influence TKA use is understudied. We examined the association between SES and pain, function and pain catastrophizing at presentation for TKA. METHODS: In patients undergoing TKA at an academic center, we obtained preoperative pain and functional status (WOMAC Index 0-100, 100 worst), pain catastrophizing (PCS, ≥ 16 high), and mental health (MHI-5, <68 poor). We described individual-level SES using education as a proxy, and area-level SES using a validated composite index linking geocoded addresses to U.S. Census data. We measured associations between these indicators and pain, function and pain catastrophizing, adjusting for age, sex and BMI. RESULTS: Among 316 patients, mean age was 65.9 (SD 8.7), 59% were female, and 88% were Caucasian; 17% achieved less than college education and 62% were college graduates. The median area SES index score was 59 (U.S. median 51). Bivariable analyses demonstrated associations between higher individual- and area-level SES and lower pain, higher function and less pain catastrophizing (all p<0.05). Adjusted analyses demonstrated statistically significant associations between higher individual- and area-level SES and better function and less pain. CONCLUSION: In this cohort, patients with higher individual- and area-level SES had lower pain and higher function at the time of TKA than lower SES patients. Further research is needed to assess what constitutes appropriate levels of pain and function to undergo TKA in these higher SES groups.

Women's political participation and health: a health capability study in rural India.

Understanding the relationship between women's political participation and health has eluded researchers and cannot be adequately studied using traditional epidemiological or social scientific methodologies. We employed a health capability framework to understand dimensions of health agency to illuminate how local political economies affect health. Exploiting a cluster-randomized controlled trial of a community-based behavior change management intervention in northern India, we conducted a qualitative study with semistructured, in-depth focus groups in both intervention and nonintervention villages. We presented scenarios to each group regarding the limitations and motivations involved in women's political participation and health. Thematic analysis focused on four domains of health agency -- participation, autonomy, self-efficacy, and health systems -- relevant for understanding the relationship between political participation and health. Elder women demonstrated the greatest sense of self-efficacy and as a group cited the largest number of successful health advocacy efforts. Participation in an associated community-based neonatal intervention had varying effects, showing some differences in self-efficacy, but only rare improvements in participation, autonomy, or health system functioning. Better understanding of cultural norms surrounding autonomy, the local infrastructure and health system, and male and female perceptions of political participation and self-efficacy are needed to improve women's health agency. For a community-based participatory health intervention to improve health capability effectively, explicit strategies focused on health agency should be as central as health indicators.

Lymphoma risk in systemic lupus: effects of disease activity versus treatment.

OBJECTIVE: To examine disease activity versus treatment as lymphoma risk factors in systemic lupus erythematosus (SLE). METHODS: We performed case-cohort analyses within a multisite SLE cohort. Cancers were ascertained by regional registry linkages. Adjusted HRs for lymphoma were generated in regression models, for time-dependent exposures to immunomodulators (cyclophosphamide, azathioprine, methotrexate, mycophenolate, antimalarial drugs, glucocorticoids) demographics, calendar year, Sjogren's syndrome, SLE duration and disease activity. We used adjusted mean SLE Disease Activity Index scores (SLEDAI-2K) over time, and drugs were treated both categorically (ever/never) and as estimated cumulative doses. RESULTS: We studied 75 patients with lymphoma (72 non-Hodgkin, three Hodgkin) and 4961 cancer-free controls. Most lymphomas were of B-cell origin. As is seen in the general population, lymphoma risk in SLE was higher in male than female patients and increased with age. Lymphomas occurred a mean of 12.4 years (median 10.9) after SLE diagnosis. Unadjusted and adjusted analyses failed to show a clear association of disease activity with lymphoma risk. There was a suggestion of greater exposure to cyclophosphamide and to higher cumulative steroids in lymphoma cases than the cancer-free controls. CONCLUSIONS: In this large SLE sample, there was a suggestion of higher lymphoma risk with exposure to cyclophosphamide and high cumulative steroids. Disease activity itself was not clearly associated with lymphoma risk. Further work will focus on genetic profiles that might interact with medication exposure to influence lymphoma risk in SLE.

Epidemiology and sociodemographics of systemic lupus erythematosus and lupus nephritis among US adults with Medicaid coverage, 2000-2004.

OBJECTIVE: Systemic lupus erythematosus (SLE) and lupus nephritis (LN) disproportionately affect individuals who are members of racial/ethnic minority groups and individuals of lower socioeconomic status (SES). This study was undertaken to investigate the epidemiology and sociodemographics of SLE and LN in the low-income US Medicaid population. METHODS: We utilized Medicaid Analytic eXtract data, with billing claims from 47 states and Washington, DC, for 23.9 million individuals ages 18-65 years who were enrolled in Medicaid for >3 months in 2000-2004. Individuals with SLE (≥3 visits >30 days apart with an International Classification of Diseases, Ninth Revision [ICD-9] code of 710.0) and with LN (≥2 visits with an ICD-9 code for glomerulonephritis, proteinuria, or renal failure) were identified. We calculated SLE and LN prevalence and incidence, stratified by sociodemographic category, and adjusted for number of American College of Rheumatology (ACR) member rheumatologists in the state and SES using a validated composite of US Census variables. RESULTS: We identified 34,339 individuals with SLE (prevalence 143.7 per 100,000) and 7,388 (21.5%) with LN (prevalence 30.9 per 100,000). SLE prevalence was 6 times higher among women, nearly double in African American compared to white women, and highest in the US South. LN prevalence was higher among all racial/ethnic minority groups compared to whites. The areas with lowest SES had the highest prevalence; areas with the fewest ACR rheumatologists had the lowest prevalence. SLE incidence was 23.2 per 100,000 person-years and LN incidence was 6.9 per 100,000 person-years, with similar sociodemographic trends. CONCLUSION: In this nationwide Medicaid population, there was sociodemographic variation in SLE and LN prevalence and incidence. Understanding the increased burden of SLE and its complications in this low-income population has implications for resource allocation and access to subspecialty care.

Screening to understand pregnancy preferences and offer referrals and treatment (SUPPORT): Results of a pilot quality improvement initiative.

OBJECTIVE: To assess the feasibility of integrating a pregnancy intention assessment screening algorithm into the electronic medical record (EMR) at a multispecialty clinic focused on the health of women and people assigned female at birth (AFAB). STUDY DESIGN: This pilot quality improvement project implemented a series of clinician reminders, new data fields in the patient record, and templated clinical notes to prompt care providers across specialties to ask AFAB reproductive age individuals about their desire for future pregnancies. Investigators created a novel screening question based on prior literature and expert input. Prospective observational study of one year of during-intervention EMR data on screening uptake and documentation, contraceptive use, and referrals to obstetrics and gynecology (OBGYN) for preconception care, contraceptive care, and related services. RESULTS: SUPPORT launched in February 2020 and was paused for 6 months due to the COVID-19 pandemic. During the intervention period through July 2021, 18% of patients for whom the automated screening reminder was activated had a documented pregnancy intention. Patients were screened in OBGYN, internal medicine, and eight subspecialty medical clinics. Among those screened, individuals who reported they did not desire pregnancy in the next year were more likely to use contraception (aOR 1.8, 95% CI 1.1, 3.1). Individuals that did desire pregnancy in the next year were more likely to be subsequently referred to OBGYN (aOR 2.7, 95% CI 1.2, 6.0). CONCLUSIONS: Despite the competing demands of the COVID-19 pandemic, the SUPPORT intervention was utilized at higher rates than prior similar interventions and across multiple disease specialties. IMPLICATIONS: Results from the SUPPORT pilot suggest that pregnancy intention screening of reproductive age AFAB individuals with an EMR-based screening prompt is feasible at scale.

Effects of Social Vulnerability and Environmental Burden on Care Fragmentation and Social Needs Among Individuals With Rheumatic Conditions.

OBJECTIVE: Environmental hazards and heightened neighborhood social vulnerability coexist and disproportionately affect minoritized populations. We investigated associations between exposure to adverse environmental burden concentrated in areas with high social vulnerability and care fragmentation (missed appointments, emergency department visits, and hospitalizations) and social needs (eg, food and housing insecurity) among individuals with rheumatic conditions. METHODS: We identified adults receiving care in a Massachusetts multihospital system with at least two rheumatic disease codes and complete street addresses. Geocoded addresses were linked to the Centers for Disease Control and Prevention/Agency for Toxic Substances and Disease Registry Social-Environmental Ranking (SER), which combines census-tract social vulnerability variables (eg, socioeconomic status) with environmental hazards (eg, air and water pollution). Social needs were obtained from self-reported surveys. Multilevel, multinomial regression models estimated associations between SER quartiles and care fragmentation and social need burden, accounting for demographics and comorbidities. RESULTS: Among 16,856 individuals with rheumatic conditions, 70% were female, 6% were Black, 82% were White, and 7% resided in the highest combined social vulnerability and environmental burden (SER quartile 4) areas. Among 7,083 with social needs data, 19% experienced more than one challenge. Individuals in SER quartile 4 areas (vs quartile 1) had 2.02 (95% confidence interval [CI] 1.67-2.46) times greater odds of at least four care fragmentation occurrences (vs 0) and 2.37 (95% CI 1.73-3.25) times greater odds of at least two social needs (vs 0). CONCLUSION: Residence in areas of high combined adverse environmental burden and social vulnerability was associated with significantly greater odds of care fragmentation and social needs. Addressing structural factors and emerging environmental threats contributing to these adverse exposures is essential to reduce rheumatic disease care inequities.

Classifying Individuals With Rheumatic Conditions as Financially Insecure Using Electronic Health Record Data and Natural Language Processing: Algorithm Derivation and Validation.

OBJECTIVE: We aimed to examine the feasibility of applying natural language processing (NLP) to unstructured electronic health record (EHR) documents to detect the presence of financial insecurity among patients with rheumatologic disease enrolled in an integrated care management program (iCMP). METHODS: We incorporated supervised, rule-based NLP and statistical methods to identify financial insecurity among patients with rheumatic conditions enrolled in an iCMP (n = 20,395) in a multihospital EHR system. We constructed a lexicon for financial insecurity using data from available knowledge sources and then reviewed EHR notes from 538 randomly selected individuals (training cohort n = 366, validation cohort n = 172). We manually categorized records as having "definite," "possible," or "no" mention of financial insecurity. All available notes were processed using Narrative Information Linear Extraction, a rule-based version of NLP. Models were trained using the NLP features for financial insecurity using logistic, least absolute shrinkage operator (LASSO), and random forest performance characteristic and were compared with the reference standard. RESULTS: A total of 245,142 notes were processed from 538 individual patient records. Financial insecurity was present among 100 (27%) individuals in the training cohort and 63 (37%) in the validation cohort. The LASSO and random forest models performed identically and slightly better than logistic regression, with positive predictive values of 0.90, sensitivities of 0.29, and specificities of 0.98. CONCLUSION: The development of a context-driven lexicon used with rule-based NLP to extract data that identify financial insecurity is feasible for use and improved the capture for presence of financial insecurity with high accuracy. In the absence of a standard lexicon and construct definition for financial insecurity status, additional studies are needed to optimize the sensitivity of algorithms to categorize financial insecurity with construct validity.

Cost-Related Medication Behaviors for Patients With and Without Systemic Autoimmune Rheumatic Diseases.

OBJECTIVE: Medication nonadherence challenges the management of systemic autoimmune rheumatic diseases (SARDs). We investigated cost-related medication behaviors among patients with SARDs, and compared them to those of patients without SARDs, in a large diverse cohort across the United States. METHODS: As part of the All of Us (version 7), a nationwide diverse adult cohort with linked electronic health records begun in 2017, participants completed questionnaires concerning cost-related medication behaviors. Chi-square tests compared responses between patients with SARDs, by disease and medication type, and to those without SARDs. Logistic regression analyses were used to calculate odds ratios (95% confidence intervals [CIs]). RESULTS: We analyzed data from 3,997 patients with SARDs and 73,990 participants without SARDs. After adjustment, patients with versus without SARDs had 1.56 times increased odds of reporting unaffordability of prescription medicines (95% CI 1.43-1.70), 1.43 times increased odds of cost-related medication nonadherence (95% CI 1.31-1.56), and 1.23 times increased odds of using cost-reducing strategies (95% CI 1.14-1.32). Patients with SARDs who reported unaffordability were 16.5% less likely to receive a disease-modifying drug (95% CI 0.70-0.99) but 18.1% more likely to receive glucocorticoids (95% CI 0.99-1.42). In addition, unaffordability of prescription medicines was likely to have 1.27 times increased odds of one to two emergency room visits per year (95% CI 1.03-1.57) and 1.38-fold increased odds of three or more emergency room visits per year (95% CI 0.96-1.99). CONCLUSION: In this large diverse cohort, patients with versus without SARDs had more self-reported cost-related medication behaviors, and those who reported medication unaffordability received fewer disease-modifying drugs and had more emergency room visits.