On- and Off-Label Atypical Antipsychotic Prescription Trends Across a Nine-Year Period Among Adolescents Pre- to Post-COVID-19.

OBJECTIVE: This study examined atypical antipsychotic prescribing by Food and Drug Administration (FDA) approved-use (on-label) status for adolescents before and during the COVID-19 pandemic. METHODS: Retrospective data were collected from electronic health records (EHRs) of adolescents aged 10-17 years in Kaiser Permanente Northern California. New outpatient atypical antipsychotic prescription orders during 2013-2021 were evaluated. Prescriptions were categorized as on-label if linked in EHRs to autism, psychosis, bipolar disorder, or Tourette's diagnoses; otherwise, they were potentially off-label (herein, off-label). Trend analysis of monthly prescribing rates assessed slope change at pandemic onset for the cohort and by sex and age groups. RESULTS: Among 5828 patients, 74.5% of new antipsychotic orders were off-label in 2021. Overall prescribing decreased significantly until early 2020 (slope = -0.045, P < .01) but then significantly increased through 2021 (post-March 2020 slope change = 0.211, P = .01). Off-label prescriptions increased at a similar rate during the COVID-19 time period, but on-label prescriptions did not change significantly. Males and younger adolescents (ages 10-14 years) showed significant decreases until early 2020, while females and older adolescents (ages 15-17 years) did not. Females and younger adolescents exhibited significant increases in overall and off-label prescribing rates following pandemic onset; older adolescents exhibited increases in overall prescriptions while males had no detectable changes. CONCLUSIONS: Antipsychotic prescribing declined slightly but then increased significantly following COVID-19 onset for overall and off-label prescriptions. Pandemic onset differentially impacted antipsychotic prescribing by sex and age, with overall and off-label prescribing driven by increases among female and younger adolescents.

Relative contribution of COVID-19 vaccination and SARS-CoV-2 infection to population-level seroprevalence of SARS-CoV-2 spike antibodies in a large integrated health system.

Background: Understanding the relative contributions of SARS-CoV-2 infection-induced and vaccine- induced seroprevalence is key to measuring overall population-level seroprevalence and help guide policy decisions. Methods: Using a series of six population-based cross-sectional surveys conducted among persons aged ≥7 years in a large health system with over 4.5 million members between May 2021 and April 2022, we combined data from the electronic health record (EHR), an electronic survey and SARS-CoV-2 spike antibody binding assay, to assess the relative contributions of infection and vaccination to population- level SARS-CoV-2 seroprevalence. EHR and survey data were incorporated to determine spike antibody positivity due to SARS-CoV-2 infection and COVID-19 vaccination. We used sampling and non-response weighting to create population-level estimates. Results: We enrolled 4,319 persons over six recruitment waves. SARS-CoV-2 spike antibody seroprevalence increased from 83.3% (CI 77.0-88.9) in May 2021 to 93.5% (CI 89.5-97.5) in April 2022. By April 2022, 68.5% (CI 61.9-74.3) of the population was seropositive from COVID-19 vaccination only, 13.9% (10.7-17.9) from COVID-19 vaccination and prior diagnosed SARS-CoV-2 infection, 8.2% (CI 4.5- 14.5) from prior diagnosed SARS-CoV-2 infection only and 2.9% (CI 1.1-7.6) from prior undiagnosed SARS-CoV-2 infection only. We found high agreement (≥97%) between EHR and survey data for ascertaining COVID-19 vaccination and SARS-CoV-2 infection status. Conclusions: By April 2022, 93.5% of persons had detectable SARS-CoV-2 spike antibody, predominantly from COVID-19 vaccination. In this highly vaccinated population and over 18 months into the pandemic, SARS-CoV-2 infection without COVID-19 vaccination was a small contributor to overall population-level seroprevalence. Article summary: By April 2022, >93% of people had antibodies to SARS-CoV-2 with COVID-19 vaccination as the main driver of overall population-level seroprevalence in our healthcare system. SARS-CoV-2 infection without vaccination made a small contribution to population-level seroprevalence in our healthcare system.

Blurring cluster randomized trials and observational studies: Two-Stage TMLE for subsampling, missingness, and few independent units.

Cluster randomized trials (CRTs) often enroll large numbers of participants; yet due to resource constraints, only a subset of participants may be selected for outcome assessment, and those sampled may not be representative of all cluster members. Missing data also present a challenge: if sampled individuals with measured outcomes are dissimilar from those with missing outcomes, unadjusted estimates of arm-specific endpoints and the intervention effect may be biased. Further, CRTs often enroll and randomize few clusters, limiting statistical power and raising concerns about finite sample performance. Motivated by SEARCH-TB, a CRT aimed at reducing incident tuberculosis infection, we demonstrate interlocking methods to handle these challenges. First, we extend Two-Stage targeted minimum loss-based estimation to account for three sources of missingness: (i) subsampling; (ii) measurement of baseline status among those sampled; and (iii) measurement of final status among those in the incidence cohort (persons known to be at risk at baseline). Second, we critically evaluate the assumptions under which subunits of the cluster can be considered the conditionally independent unit, improving precision and statistical power but also causing the CRT to behave like an observational study. Our application to SEARCH-TB highlights the real-world impact of different assumptions on measurement and dependence; estimates relying on unrealistic assumptions suggested the intervention increased the incidence of TB infection by 18% (risk ratio [RR]=1.18, 95% confidence interval [CI]: 0.85-1.63), while estimates accounting for the sampling scheme, missingness, and within community dependence found the intervention decreased the incident TB by 27% (RR=0.73, 95% CI: 0.57-0.92).

Severe Acute Respiratory Syndrome Coronavirus 2 Delta Variant Genomic Variation Associated With Breakthrough Infection in Northern California: A Retrospective Cohort Study.

BACKGROUND: The association between severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) genomic variation and breakthrough infection is not well defined among persons with Delta variant SARS-CoV-2 infection. METHODS: In a retrospective cohort, we assessed whether individual nonlineage defining mutations and overall genomic variation (including low-frequency alleles) were associated with breakthrough infection, defined as SARS-CoV-2 infection after coronavirus disease 2019 primary vaccine series. We identified all nonsynonymous single-nucleotide polymorphisms, insertions, and deletions in SARS-CoV-2 genomes with ≥5% allelic frequency and population frequency of ≥5% and ≤95%. Using Poisson regression, we assessed the association with breakthrough infection for each individual mutation and a viral genomic risk score. RESULTS: Thirty-six mutations met our inclusion criteria. Among 12 744 persons infected with Delta variant SARS-CoV-2, 5949 (47%) were vaccinated and 6795 (53%) were unvaccinated. Viruses with a viral genomic risk score in the highest quintile were 9% more likely to be associated with breakthrough infection than viruses in the lowest quintile, but including the risk score improved overall predictive model performance (measured by C statistic) by only +0.0006. CONCLUSIONS: Genomic variation within SARS-CoV-2 Delta variant was weakly associated with breakthrough infection, but several potential nonlineage defining mutations were identified that might contribute to immune evasion by SARS-CoV-2.

A Demonstration of Modified Treatment Policies to Evaluate Shifts in Mobility and COVID-19 Case Rates in US Counties.

Mixed evidence exists of associations between mobility data and coronavirus disease 2019 (COVID-19) case rates. We aimed to evaluate the county-level impact of reducing mobility on new COVID-19 cases in summer/fall of 2020 in the United States and to demonstrate modified treatment policies to define causal effects with continuous exposures. Specifically, we investigated the impact of shifting the distribution of 10 mobility indexes on the number of newly reported cases per 100,000 residents 2 weeks ahead. Primary analyses used targeted minimum loss-based estimation with Super Learner to avoid parametric modeling assumptions during statistical estimation and flexibly adjust for a wide range of confounders, including recent case rates. We also implemented unadjusted analyses. For most weeks, unadjusted analyses suggested strong associations between mobility indexes and subsequent new case rates. However, after confounder adjustment, none of the indexes showed consistent associations under mobility reduction. Our analysis demonstrates the utility of this novel distribution-shift approach to defining and estimating causal effects with continuous exposures in epidemiology and public health.

A mid-level health manager intervention to promote uptake of isoniazid preventive therapy among people with HIV in Uganda: a cluster randomised trial.

BACKGROUND: Despite longstanding guidelines endorsing isoniazid preventive therapy (IPT) for people with HIV, uptake is low across sub-Saharan Africa. Mid-level health managers oversee IPT programmes nationally; interventions aimed at this group have not been tested. We aimed to establish whether providing structured leadership and management training and facilitating subregional collaboration and routine data feedback to mid-level managers could increase IPT initiation among people with HIV compared with standard practice. METHODS: We conducted a cluster randomised trial in Uganda among district-level health managers. We randomly assigned clusters of between four and seven managers in a 1:1 ratio to intervention or control groups. Our intervention convened managers into mini-collaboratives facilitated by Ugandan experts in tuberculosis and HIV, and provided business leadership and management training, SMS platform access, and data feedback. The control was standard practice. Participants were not masked to trial group, but study statisticians were masked until trial completion. The primary outcome was IPT initiation rates among adults with HIV in facilities overseen by participants over a period of 2 years (2019-21). We conducted prespecified analyses that excluded the third quarter of 2019 (Q3-2019) to understand intervention effects independent of a national 100-day IPT push tied to a financial contingency during Q3-2019. This trial is registered with ClinicalTrials.gov (NCT03315962), and is ongoing. FINDINGS: Between Nov 15, 2017, and March 14, 2018, managers from 82 of 82 eligible districts (61% of Uganda's 135 districts) were enrolled and randomised: 43 districts to intervention, 39 to control. Intervention delivery took place between Dec 6, 2017, and Feb 2, 2022. Over 2 years, IPT initiation rates were 0·74 versus 0·65 starts per person-year in intervention versus control groups (incidence rate ratio [IRR] 1·14, 95% CI 0·88-1·46; p=0·16). Excluding Q3-2019, IPT initiation was higher in the intervention group versus the control group: 0·32 versus 0·25 starts per person-year (IRR 1·27, 95% CI 1·00-1·61; p=0·026). INTERPRETATION: Following an intervention targeting managers in more than 60% of Uganda's districts, IPT initiation rates were not significantly higher in intervention than control groups. After accounting for large increases in IPT from a 100-day push in both groups, the intervention led to significantly increased IPT rates, sustained after the push and during the COVID-19 pandemic. Our findings suggest that interventions centred on mid-level health managers can improve IPT implementation on a large, subnational scale, and merit further exploration to address key public health challenges for which strong evidence exists but implementation remains suboptimal. FUNDING: National Institute of Allergy and Infectious Diseases.

Type I error control for cluster randomized trials under varying small sample structures.

BACKGROUND: Linear mixed models (LMM) are a common approach to analyzing data from cluster randomized trials (CRTs). Inference on parameters can be performed via Wald tests or likelihood ratio tests (LRT), but both approaches may give incorrect Type I error rates in common finite sample settings. The impact of different combinations of cluster size, number of clusters, intraclass correlation coefficient (ICC), and analysis approach on Type I error rates has not been well studied. Reviews of published CRTs find that small sample sizes are not uncommon, so the performance of different inferential approaches in these settings can guide data analysts to the best choices. METHODS: Using a random-intercept LMM stucture, we use simulations to study Type I error rates with the LRT and Wald test with different degrees of freedom (DF) choices across different combinations of cluster size, number of clusters, and ICC. RESULTS: Our simulations show that the LRT can be anti-conservative when the ICC is large and the number of clusters is small, with the effect most pronouced when the cluster size is relatively large. Wald tests with the between-within DF method or the Satterthwaite DF approximation maintain Type I error control at the stated level, though they are conservative when the number of clusters, the cluster size, and the ICC are small. CONCLUSIONS: Depending on the structure of the CRT, analysts should choose a hypothesis testing approach that will maintain the appropriate Type I error rate for their data. Wald tests with the Satterthwaite DF approximation work well in many circumstances, but in other cases the LRT may have Type I error rates closer to the nominal level.