We evaluated neural tube defect (NTD) risk associated with prescription opioid analgesic use during early pregnancy. We conducted a cohort study of liveborn singletons during 2001-2014 among 9 U.S. health plans. Risks of medical chart-confirmed primary NTD (anencephaly and select spina bifida) and any NTD (primary and other NTD) were compared among singletons with maternal new use of opioids during 18-56 days after last menstrual period to those with no maternal use during this period or pre-pregnancy. In a sensitivity analysis, singletons exposed during the second or third trimesters only ("negative control exposure") were compared to those never exposed. The main analysis adjusted odds ratio (OR) associated with exposure was 3.0 (95% CI, 0.7-12.7) for primary NTD and 2.3 (95% CI, 0.8-6.7) for any NTD. The sensitivity analysis ORs were similar 2.4 and 1.6, respectively. Our results do not support or refute the hypothesis that prescription opioid analgesic exposure in early pregnancy increases risk of NTD. Although ORs were increased in the main analysis, the sensitivity analysis suggested the presence of residual confounding. Future studies that are sufficiently powered and adjust for confounders beyond those in the present study are warranted but likely challenging.

PURPOSE:Information from electronic health records (EHRs) may be incorporated into computable phenotype algorithms in efforts to overcome inaccuracies of algorithms based on administrative claims data alone. However, such efforts can be resource-intensive and unsuccessful. Assessing the feasibility of computable phenotyping for a health outcome of interest (HOI) before proceeding is therefore recommended. METHODS:We developed a systematic fitness-for-purpose (FFP) assessment process to implement concepts outlined in a previously described general framework for computable phenotyping incorporating EHR data. Our process includes verifying the HOI is well-defined, reviewing clinical information about the HOI, identifying existing algorithms and their performance, evaluating HOI clinical and data complexity, and determining an overall FFP conclusion and recommendation. We applied this process to 10 HOIs lacking high-performing claims-based algorithms, selecting HOIs of public health importance that varied in clinical and data complexity, including neutropenia, pericardial effusion, and drug-induced liver injury. RESULTS:HOIs assessed as having moderate (vs. easy) overall difficulty had characteristics such as the need for natural language processing, integration of multiple laboratory test results, or longitudinal EHR data. HOIs assessed as having high difficulty required using data from multiple EHR sources, ruling out many other potential causes, or relying on low-sensitivity diagnostic tests. Input from experts in EHR data and clinical care was crucial. CONCLUSION:EHR data have the potential to enhance the accuracy of defining certain HOIs for research and surveillance compared to administrative claims data. The process and tools we created will support others in assessing FFP of HOIs for computable phenotyping.

BACKGROUND:Accurate algorithms are needed to support real-world studies of medication safety in pregnancy. Kharbanda et al. developed and validated algorithms for congenital malformations that incorporate death data and diagnosis codes from the ICD-9-CM and ICD-10-CM coding eras. OBJECTIVES:To modify an EHR-based malformation algorithm for use with claims data and quantify the impact of missing death by calculating prevalence and sensitivity. METHODS:Using the MarketScan Commercial Database (2007-2022), we established a linked cohort of birthing parents and their liveborn infants, and a claims subcohort restricted to years with inpatient death information. We established a cohort of liveborn infants in Kaiser Permanente Washington (KPWA) integrated EHR/claims data (2007-2022) that included comprehensive death information. We applied the validated algorithm to identify 22 malformations in MarketScan and 7 in KPWA (those with a prevalence ≥ 10 per 10,000 live births in MarketScan). In MarketScan, we calculated malformation prevalence with and without death information. We assessed the contribution of death on malformation identification by calculating sensitivity (with death as the gold standard). RESULTS:Among 2,203,328 infants in the MarketScan cohort, malformation prevalence was 201.3 per 10,000 live births. In the MarketScan subcohort (n = 1,287,384), prevalence was 198.2 and 199.1 per 10,000 live births without and with death information, respectively. Among the most prevalent malformations, estimated sensitivity ranged from 95.8% for severe cardiac defects to 100.0% for intestinal atresia or stenosis, pyloric stenosis and limb deficiency (claims/EHR cohort) and from 98.6% for severe cardiac defects to 100.0% for intestinal atresia or stenosis and pyloric stenosis (claims subcohort). Limitations include the use of an imperfect gold standard and a lack of chart review. CONCLUSIONS:We adapted a validated malformation algorithm for use with claims data. Omitting death information did not meaningfully impact sensitivity, suggesting this algorithm can be applied to data sources lacking death information.

BACKGROUND:Choosing antibiotics for prenatal urinary tract infections (UTIs) is challenging because fetal safety data are inconclusive and must be weighed against resistance. Recent information on prescribing patterns and antimicrobial susceptibility is sparse; we aimed to address this gap. STUDY DESIGN:We used electronic health record data from two US health systems (Kaiser Permanente Washington [KPWA] and Vanderbilt University Medical Center [VUMC]) to ascertain pregnancies from January 1, 2006, to August 31, 2023, to individuals aged 15-49. We included outpatient-treated prenatal UTIs with an oral UTI antibiotic plus a diagnosis or positive urine culture. We described patterns of antibiotic utilization and susceptibility overall (2006-2023) and by year (2010-2022). RESULTS:We identified 10,734 eligible UTIs. The most common antibiotics were nitrofurantoin (KPWA: 42%; VUMC: 65%) and first generation cephalosporins (KPWA: 23%; VUMC: 15%). Nitrofurantoin decreased substantially from 2010 to 2022 (KPWA: from 50% to 27%; VUMC: 65-49%), while first generation cephalosporins increased (KPWA: 12 to 45%; VUMC: 14 to 21%). Escherichia coli was susceptible to nitrofurantoin in 98-99% of UTIs. Susceptibility to other antibiotics was higher at KPWA than VUMC (first generation cephalosporins: KPWA: 97%, VUMC: 89%; amoxicillin-clavulanate: KPWA: 93%, VUMC: 85%). 93-95% were treated with an appropriate antibiotic based on susceptibility results. CONCLUSIONS:In two health systems in different regions, nitrofurantoin decreased substantially while first generation cephalosporins increased, despite better nitrofurantoin susceptibility. Other concerns, like malformation risk, may have influenced prescribing. Further research and guideline development are needed to weigh risks versus benefits of different antibiotics for prenatal UTIs.

Benzodiazepines and sedative hypnotics such as zolpidem (“z-drugs”) are commonly prescribed for anxiety and sleep disorders. Epidemiologic evidence links their use to increased risk of venous thromboembolism. We investigated venous thromboembolism risk among concomitant users of individual benzodiazepines/z-drugs (examined separately) with other prescription medications to generate data-driven hypotheses about drug interactions resulting in clinically meaningful harm to inform future etiologic studies of specific drug combinations. We conducted a series of self-controlled case series studies within a 50

OBJECTIVE:Insufficient sleep is linked to adverse health outcomes, and less than one-third of high schoolers achieve sufficient sleep. This study sought to identify factors associated with sufficient sleep in New Mexico high schoolers over the last decade, with attention to trends over time. METHODS:We used individual-level data from a representative sample of New Mexico high schoolers from five waves of the New Mexico Youth Risk and Resiliency Survey. We generated figures depicting how the presence/absence of risk and resiliency factors related to reported achievement of sufficient sleep. We also used logistic regression models including survey weights to examine adjusted associations. RESULTS:We included New Mexico high schoolers from 2015 (n = 11,310), 2017 (n = 13,371), 2019 (n = 14,440), 2021 (n = 12,317) and 2023 (n = 10,806). High schoolers with resiliency factors were consistently more likely to report sufficient sleep, and those with risk factors were consistently less likely to report sufficient sleep. Across all waves, feeling sad/hopeless (estimates ranged from prevalence odds ratio [POR] 0.53, 95%CI 0.45, 0.62 to POR 0.67, 95%CI 0.57, 0.80) and current alcohol use (POR 0.69, 95%CI 0.55, 0.86 to 0.86, 95%CI 0.74, 0.99) were associated with lower odds of sufficient sleep. Conversely, social support from adults (POR 1.22, 95%CI 1.01, 1.46 to POR 1.48, 95%CI 1.25, 1.75) and eating breakfast daily (POR 1.90, 95%CI 1.69, 2.14 to POR 2.07, 95%CI 1.80, 2.38) were associated with greater odds of sufficient sleep. CONCLUSIONS:The modifiable risk and resiliency factors identified could be meaningful targets for future interventions to improve adolescent sleep duration.

The eRADAR study is a randomized controlled trial in two United States healthcare systems testing targeted dementia screening in primary care. High-risk individuals ≥ age 65 were identified using a predictive algorithm—eRADAR (electronic health record Risk of Alzheimer’s and Dementia Assessment Rule)—and invited for a “brain health” screening visit (BHV). Patients and their primary care providers (PCPs) were given visit findings (possible dementia, possible mild cognitive impairment [MCI], or normal cognition) and follow-up recommendations. A subset of participants and their care partners (CPs) were invited to complete in-depth telephone interviews based on a purposeful sampling that prioritized participants who screened positive for possible dementia or MCI. We conducted interviews representing 40 BHVs (38 patients and 19 CPs). Forty-eight percent of patients screened positive for possible dementia, 38% for MCI, and 15% were cognitively normal. Patient reactions to the BHV were positive overall. Very few participants reported that the BHV caused distress or anxiety, but they also did not feel the BHV provided them with new information. Patients and CPs often downplayed results suggestive of cognitive impairment or attributed them to other causes, and many were reluctant to follow up with their PCP. Interview feedback indicates the BHV did not consistently motivate people to seek further evaluation or care for cognitive issues. Similar screening programs should consider more direct options to motivate follow-up for positive findings (e.g., directly scheduling a follow-up with the PCP, direct referral to specialty care, screening immediately before a PCP appointment).

Approximately half of people with dementia and 90% with mild cognitive impairment (MCI) are undiagnosed, making it harder to access appropriate care. eRADAR is a validated algorithm that uses structured electronic health record (EHR) data to identify patients at risk of undiagnosed dementia (C statistics: 0.79 to 0.84). We performed embedded, pragmatic clinical trials at Kaiser Permanente Washington (KPWA) and the University of California, San Francisco (UCSF) to assess the impact of targeted dementia screening using eRADAR on dementia detection and patient outcomes. Study participants were aged 65+ without documented dementia (diagnoses or current dementia-related medication) who had eRADAR scores in the top 15% (20% in Black/African American patients at UCSF to ensure equivalent sensitivity). They were randomized at the primary care provider (PCP) level to usual care or intervention. The intervention consisted of a “brain health” assessment visit with a clinical research interventionist embedded in the primary care clinic. Participants were encouraged to include a care partner. Visits could be in-person or by video or phone. At UCSF, visits were offered in Spanish, Mandarin, and Cantonese in addition to English. The interventionist assessed functional status (independence in instrumental activities of daily living), depressive symptoms (PHQ-2/9), and cognitive function (Montreal Cognitive Assessment) and shared the findings with the participant and their PCP. The PCP independently followed up and made diagnoses at their discretion. 3417 KPWA and 1271 UCSF patients were identified as “high risk” based on their eRADAR scores and randomized to targeted outreach or usual care. Of those who received outreach, 29% at KPWA and 31% at UCSF completed an assessment visit. Visit completion rates were 52% to 74% lower in those who were older vs. younger and Black or Asian (both sites) or Spanish-speaking (UCSF only) vs. non-Hispanic White. Of those who completed the visit, approximately half had results consistent with undiagnosed cognitive impairment (KPWA: 17% possible dementia, 32% MCI; UCSF: 21% possible dementia, 31% MCI). Among older adults identified as high-risk using eRADAR who agreed to an assessment visit, about half appeared to have undiagnosed cognitive impairment. Additional studies are needed to explore disparities in visit acceptance.

Approximately half of people with dementia are undiagnosed, yet little is known about patient responses to real-world screening for dementia in primary care. As part of a pragmatic trial, we evaluated patient-reported acceptability and satisfaction with a primary care-based, targeted cognitive screening visit. Study sites were Kaiser Permanente Washington (KPWA) and University of California, San Francisco (UCSF) healthcare systems. Participants identified as high-risk for having undiagnosed dementia were invited for a “brain health” visit that assessed independence in daily activities, depressive symptoms and cognitive impairment. Participants were encouraged to include a care partner at the visit. Afterward, participants were given a post-visit evaluation survey that included structured and open-ended questions about visit satisfaction. Using a mixed methods approach, proportions of “agree” responses were calculated for structured questions, and free text responses were analyzed using Rapid Group Analysis Process (Rap-GAP). N  = 426 ( n  = 327/64.6% KPWA, n  = 99/50.3% UCSF) completed a post-visit survey. KPWA survey respondents were 43% female, 91% Non-Hispanic White, and 71% had a college degree or higher; UCSF respondents were 52% female, 66% Non-Hispanic White, and 76% had a college degree or higher. Based on the visit, 13% of respondents were referred to their PCP for follow-up regarding memory concerns. From structured questions (Figure), >90% of participants reported they felt comfortable discussing brain health and found the visit helpful for understanding their current brain health. Qualitative themes (Table) confirmed these positive findings, but also surfaced some important critiques. Participant concerns included not understanding the next steps recommended for cognitive evaluation, highlighting the need to share next steps clearly and, when appropriate, to share summary recommendations with a care partner directly. Few participants reported feeling upset by the visit (∼11% of participants). Qualitative themes suggested that when upset occurred, reasons included concern about their results, perceptions that the assessment of their functioning might be inaccurate due to hearing loss or difficulty understanding the instructions, or finding questions about their daily functioning patronizing. High-risk patients participating in primary care-based cognitive screening visits found them acceptable. Targeted screening approaches could provide a supportive pathway to improve dementia detection in primary care.

To ensure new dementia research findings and treatments benefit all populations, researchers must include non-English-speaking patients in studies. However, in the U.S., these populations are often excluded due to the increased complexity created for research teams. This study describes challenges and learnings related to recruitment and engagement with non-English speakers in a clinical cognitive assessment study. The electronic health record Risk of Alzheimer's and Dementia Assessment Rule (eRADAR) trial is a pragmatic randomized controlled trial using an electronic health record (EHR)-based algorithm to identify older adults at high risk of undiagnosed dementia and invite them to participate in a brain health assessment visit (BHV). Recruitment at one site included Chinese (Cantonese and Mandarin) and Spanish speakers alongside English speakers. We followed evidence-based approaches to language translation, recruitment, and engagement of non-English speakers in clinical research, including consultation with bilingual patient advisors. Research staff documented detailed comments about recruitment calls and BHVs in a tracking database. When enrollment was complete, three research staff reviewed all tracking comments ( n  = 574 participants) for language relevance and iteratively developed key themes in consultation with the larger research team. Supporting examples were summarized for each theme in a coding memo. The team identified language-relevant tracking comments for 83 participants: developing four recruitment outreach themes with six sub-themes and three BHV-specific themes (Table). Discrepancies between the EHR-documented language and participants' preferred language impacted recruitment. For some participants, hearing and cognitive difficulties compounded the communication challenges across a language barrier. Care partners were crucial to recruitment, and we frequently needed to work with them to reach participants, which was not always successful. During BHVs, complexities arose from differences in care partner and participant language fluency, participant bilingualism, and selecting the appropriate version of the cognitive screening tool (Montreal Cognitive Assessment [MoCA]). In the eRADAR study, we followed evidence-based approaches to inclusion of non-English speakers in research and still uncovered important areas that required additional attention to the language needs of participants and their care partners. Other researchers should anticipate similar challenges and implement robust strategies to ensure inclusive and equitable participation in dementia studies.

ABSTRACTPurposeStudies using insurance claims data to identify pregnancies are rarely able to directly assess the validity of the pregnancy/delivery. Inpatient versus outpatient delivery claims may provide different levels of evidence, but more stringent requirements could result in exclusion of true pregnancies. We identified delivery codes from the inpatient and outpatient settings and examined possible confirmatory evidence suggesting that a delivery truly occurred.MethodsUsing a US commercial insurance database (2006–2021), we identified potential pregnancies by presence of delivery claims from a provider and/or facility. We classified deliveries as inpatient (claim date during inpatient admission) or outpatient (claim date not during inpatient admission). We identified possible confirmatory evidence for each delivery including: (1) Presence of both provider and facility delivery codes; (2) presence of both diagnosis and procedure delivery codes; (3) labor and delivery revenue codes; (4) gestational age diagnosis codes; (5) pregnancy‐related care codes; (6) linkage to an infant claim; and (7) infant insurance enrollment and linkage to a birthing parent. We quantified the proportion of deliveries with confirmatory evidence by delivery setting. Among deliveries with ≥ 1 piece of confirmatory evidence, we compared patient characteristics by apparent delivery setting.ResultsAmong 4 084 474 delivery episodes, 96.4% were classified as inpatient and 3.6% outpatient. 99.9% of inpatient and 94.0% of outpatient deliveries had ≥ 1 piece of confirmatory evidence. Pregnancy‐related care codes were the most common type of confirmatory evidence (99.0% inpatient, 85.7% outpatient). Deliveries classified as inpatient occurred among patients who were older and more clinically complex (i.e., more pregnancy complications, chronic diseases, and prescription medications).ConclusionsThe vast majority of deliveries had confirmatory evidence regardless of apparent setting. Patient characteristics differed by delivery setting. Inclusion of apparent outpatient deliveries may increase the sample size of the study population and improve the generalizability of study results.

Importance:Clinical guidelines recommend screening and treating bacteriuria in early pregnancy given that urinary tract infections (UTIs) can cause serious maternal and neonatal consequences. Evidence regarding antibiotic exposure during early pregnancy and risk of congenital malformations is limited and inconsistent. Objective:To compare the risk of congenital malformations following first-trimester exposure to different antibiotic agents used to treat UTI. Design, Setting, and Participants:This population-based cohort study included commercially insured pregnant individuals aged 15 to 49 years who were treated for UTI and linked liveborn infants in the Merative MarketScan Commercial Database (2006-2022). Exposure:First-trimester antibiotic prescription fill of nitrofurantoin, trimethoprim-sulfamethoxazole (TMP-SMX), fluoroquinolones (ciprofloxacin, levofloxacin, ofloxacin), and β-lactams to treat UTI. Main Outcomes and Measures:Congenital malformations (any and by organ system) were identified using validated algorithms based on diagnosis codes up to 365 days after birth. Log-binomial regression models were used to estimate propensity score-weighted risk ratios (RRs) and risk differences. Results:The cohort of 71 604 eligible pregnancies (median maternal [IQR] age, 30 [27-34] years) included 42 402 (59.2%) nitrofurantoin-exposed, 3494 (4.9%) TMP-SMX-exposed, 3663 (5.1%) fluoroquinolone-exposed, and 22 045 (30.8%) β-lactam-exposed individuals. Median (IQR) gestational age differed by antibiotic (nitrofurantoin, 62 [45-77] days; TMP-SMX, 26 [13-59] days; fluoroquinolones, 18 [9-27] days; β-lactams, 63 [48-77] days). The absolute risk of any malformation was 19.8 (95% CI, 18.0-21.8) per 1000 infants for β-lactams, 21.2 (95% CI, 19.9-22.7) per 1000 infants for nitrofurantoin, 23.5 (95% CI, 18.8-28.9) per 1000 infants for fluoroquinolones, and 26.9 (95% CI, 21.8-32.8) per 1000 infants for TMP-SMX. After accounting for confounding, risk of any congenital malformation was higher for TMP-SMX (RR, 1.35; 95% CI, 1.04-1.75) but similar for nitrofurantoin (RR, 1.12; 95% CI, 1.00-1.26) and fluoroquinolones (RR, 1.18; 95% CI, 0.87-1.60) compared with β-lactams. TMP-SMX was associated with increased risk of severe cardiac malformations (RR, 2.09; 95% CI, 1.09-3.99), other cardiac malformations (RR, 1.52; 95% CI, 1.02-2.25), and cleft lip and palate (RR, 3.23; 95% CI, 1.44-7.22) compared with β-lactams; however, for these specific malformations, the corresponding risk difference estimates included the null. Risk of other malformation types did not differ by agent, although some estimates were imprecise. Results were generally consistent across sensitivity analyses. Conclusions and Relevance:In this cohort study of first-trimester antibiotic exposure, the risk of any malformation, severe cardiac malformation, other cardiac malformation, and cleft lip and palate was higher for infants exposed to TMP-SMX vs β-lactam antibiotics. No elevated risk was observed for nitrofurantoin.

Background: Urinary tract infections (UTIs) occur in 10-18% of United States pregnancies and may lead to maternal and neonatal complications. We examined differences in the incidence and treatment of UTI during pregnancy by race and ethnicity. Methods: We conducted a cohort study within an integrated health care system, including members aged 15-49 years with live births, stillbirths, spontaneous abortions, or terminations between January 2011 and August 2023. Self-reported race and ethnicity were documented in the electronic health record (EHR). UTIs were defined from diagnosis, medication, and laboratory data. We used modified poisson regression to estimate the cumulative incidence and adjusted risk ratios ([aRR]; with non-Hispanic [NH] White individuals as a reference) of the first UTI per pregnancy, adjusted for maternal age. Results: Among 63,029 pregnancies across eight racial and ethnic categories, 5,083 (8.1%) individuals experienced UTI during pregnancy. Cohort mean maternal age was 30.4 years and 75.9% of birth outcomes were live or stillbirths. UTI risk ranged from 7.2% among NH White individuals to 14.3% among NH Native Hawaiian/Pacific Islander [NHPI] individuals. Compared to NH White individuals, UTI risk was elevated among NH NHPI (aRR 1.80, 95% confidence interval [CI] 1.50-2.15), NH Black (aRR 1.32, 95% CI 1.19-1.46), NH Asian (aRR 1.13, 95% CI 1.04-1.24), and Hispanic (aRR 1.45, 95% CI 1.32-1.59) individuals. We observed no differences in UTI treatment by race or ethnicity. Conclusion: UTI burden during pregnancy is greater among racially minoritized groups, suggesting a need for more focus on upstream risk factors, screening, and alleviating the influence of structural racism on infection risk.

Importance:Modifiable risk factors are hypothesized to account for 30% to 40% of dementia; yet, few trials have demonstrated that risk-reduction interventions, especially multidomain, are efficacious. Objective:To determine if a personalized, multidomain risk reduction intervention improves cognition and dementia risk profile among older adults. Design, Setting, and Participants:The Systematic Multi-Domain Alzheimer Risk Reduction Trial was a randomized clinical trial with a 2-year personalized, risk-reduction intervention. A total of 172 adults at elevated risk for dementia (age 70-89 years and with ≥2 of 8 targeted risk factors) were recruited from primary care clinics associated with Kaiser Permanente Washington. Data were collected from August 2018 to August 2022 and analyzed from October 2022 to September 2023. Intervention:Participants were randomly assigned to the intervention (personalized risk-reduction goals with health coaching and nurse visits) or to a health education control. Main Outcomes and Measures:The primary outcome was change in a composite modified Neuropsychological Test Battery; preplanned secondary outcomes were change in risk factors and quality of life (QOL). Outcomes were assessed at baseline and 6, 12, 18, and 24 months. Linear mixed models were used to compare, by intention to treat, average treatment effects (ATEs) from baseline over follow-up. The intervention and outcomes were initially in person but then, due to onset of the COVID-19 pandemic, were remote. Results:The 172 total participants had a mean (SD) age of 75.7 (4.8) years, and 108 (62.8%) were women. After 2 years, compared with the 90 participants in the control group, the 82 participants assigned to intervention demonstrated larger improvements in the composite cognitive score (ATE of SD, 0.14; 95% CI, 0.03-0.25; P = .02; a 74% improvement compared with the change in the control group), better composite risk factor score (ATE of SD, 0.11; 95% CI, 0.01-0.20; P = .03), and improved QOL (ATE, 0.81 points; 95% CI, -0.21 to 1.84; P = .12). There were no between-group differences in serious adverse events (24 in the intervention group and 23 in the control group; P = .59), but the intervention group had greater treatment-related adverse events such as musculoskeletal pain (14 in the intervention group vs 0 in the control group; P < .001). Conclusions and Relevance:In this randomized clinical trial, a 2-year, personalized, multidomain intervention led to modest improvements in cognition, dementia risk factors, and QOL. Modifiable risk-reduction strategies should be considered for older adults at risk for dementia. Trial Registration:ClinicalTrials.gov Identifier: NCT03683394.

Abstract Nearly 7 million people in the U.S. are living with dementia, but approximately half are undiagnosed. We are performing a randomized, controlled trial in two healthcare systems (Kaiser Permanent Washington-KPWA, University of California San Francisco-UCSF) of targeted dementia screening. Among primary care patients ≥ age 65, high-risk individuals are identified using a predictive algorithm, eRADAR (electronic health record Risk of Alzheimer’s and Dementia Assessment Rule), and invited for a “brain health” screening visit (BHV). Among 1709 KPWA patients in the intervention group, 506 (30%) completed a BHV; ongoing UCSF recruitment is similar (N=68/221; 31%). Visit completion rates were lower for participants of color (Asian/Black/Latinx) at 18% (N=51/336) compared with non-Hispanic whites at 34% (N=446/1,313), adjusted OR 0.36, 95% CI 0.26-0.49. Completion rates were 33% (N=30/90) among people aged 65-74 vs. 16% (N=44/268) for those aged 90+, adjusted OR per decade 0.71, 95% CI 0.56-0.90. Prior diagnoses of memory problems/concerns or mild cognitive impairment (MCI) were associated with higher odds of completion (adjusted OR 1.49; CI 1.13-1.95), and so was lower comorbidity score (adjusted OR 1.01; CI 1.00-1.03). eRADAR score and gender were not associated with BHV completion. The eRADAR study participation rate is similar to other dementia screening studies. Younger, white patients and those who might be aware of their cognitive deficits–as demonstrated by prior diagnoses–appear more likely to participate. As disease-modifying treatments increasingly become available, further research with communities of color to understand facilitators and barriers to engagement in primary-care dementia assessment is vital to prevent care disparities.

The association between current use of oral contraceptives (OCs) among women younger than 50 years (n = 306 541), and hormone therapy (HT) among women aged 50 years or older (n = 323 203), and coronavirus 2019 (COVID-19) infection and hospitalization was evaluated in this population-based cohort. Current OC/HT use was recorded monthly using prescription dispensing data. COVID-19 infections were identified from March 2020 through February 2021. COVID-19 infections and hospitalizations were identified through diagnosis codes and laboratory tests. We used weighted generalized estimating equations models to estimate multivariable adjusted odds ratios (aORs) for COVID-19 infection associated with time-varying OC/HT use. Among women with COVID-19, logistic regression models were used to evaluate OC/HT use and COVID-19 hospitalization. Over 12 months, 11 727 (3.8%) women younger than 50 years and 8661 (2.7%) women aged 50 years or older experienced COVID-19 infections. There was no evidence of an association between OC use and infection (aOR = 1.05; 95% CI, 0.97-1.12). There was a modest association between HT use and infection (aOR = 1.19; 95% CI, 1.03-1.38). Women using OCs had a 39% lower risk of hospitalization (aOR = 0.61; 95% CI, 0.38-1.00), but there was no association of HT use with hospitalization (aOR = 0.89; 95% CI, 0.51-1.53). These findings do not suggest a meaningfully greater risk of COVID-19 infection associated with OC or HT use. OC use may be associated with lower COVID-19 hospitalization risk.

BACKGROUND:Stopping or reducing risky or unneeded medications ("deprescribing") could improve older adults' health. Electronic health data can support observational and intervention studies of deprescribing, but there are no standardized measures for key variables, and healthcare systems have differing data types and availability. We developed definitions for chronic medication use and discontinuation based on electronic health data and applied them in a case study of benzodiazepines and Z-drugs in five diverse US healthcare systems. METHODS:We conducted a retrospective cohort study of adults age 65+ from 2017 to 2019 with chronic benzodiazepine or Z-drug use. We determined whether sites had access to medication orders and/or dispensings. We developed definitions for chronic use and discontinuation using both data types. Discontinuation definitions were based on (1) gaps in medication availability during follow-up or (2) not having medication available at a fixed time point. We examined the impact of varying the gap length and requiring a 30-day period without orders/dispensings ("halo") around the fixed time point. We compared results derived from orders versus dispensings at one site. RESULTS:Approximately 1.6%-2.6% of older adults had chronic benzodiazepine/Z-drug use (total N = 6775, ranging from 431 to 2122 across sites). Depending on the definition and site, the proportion discontinuing use during 12 months ranged from 6% to 49%. Requiring a longer gap or a 30-day "halo" resulted in lower estimates. At one site, only 56% of those with chronic use defined from orders also qualified based on dispensings, and the discontinuation rate at 180 days was 20% from orders versus 32% from dispensings. CONCLUSIONS:Requiring a gap of ≥90 days or a "halo" around a time point may more accurately capture discontinuation than using a shorter gap or no halo. Orders data underestimate discontinuation compared to dispensings. Work is needed to adapt these definitions for other drug classes and settings.