Two-Stage TMLE to reduce bias and improve efficiency in cluster randomized trials.

Cluster randomized trials (CRTs) randomly assign an intervention to groups of individuals (e.g., clinics or communities) and measure outcomes on individuals in those groups. While offering many advantages, this experimental design introduces challenges that are only partially addressed by existing analytic approaches. First, outcomes are often missing for some individuals within clusters. Failing to appropriately adjust for differential outcome measurement can result in biased estimates and inference. Second, CRTs often randomize limited numbers of clusters, resulting in chance imbalances on baseline outcome predictors between arms. Failing to adaptively adjust for these imbalances and other predictive covariates can result in efficiency losses. To address these methodological gaps, we propose and evaluate a novel two-stage targeted minimum loss-based estimator to adjust for baseline covariates in a manner that optimizes precision, after controlling for baseline and postbaseline causes of missing outcomes. Finite sample simulations illustrate that our approach can nearly eliminate bias due to differential outcome measurement, while existing CRT estimators yield misleading results and inferences. Application to real data from the SEARCH community randomized trial demonstrates the gains in efficiency afforded through adaptive adjustment for baseline covariates, after controlling for missingness on individual-level outcomes.

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).

Cluster randomized trial designs for modeling time-varying intervention effects.

Stepped-wedge cluster randomized trials (SW-CRTs) are typically analyzed assuming a constant intervention effect. In practice, the intervention effect may vary as a function of exposure time, leading to biased results. The estimation of time-on-intervention (TOI) effects specifies separate discrete intervention effects for each elapsed period of exposure time since the intervention was first introduced. It has been demonstrated to produce results with minimum bias and nominal coverage probabilities in the analysis of SW-CRTs. Due to the design's staggered crossover, TOI effect variances are heteroskedastic in a SW-CRT. Accordingly, we hypothesize that alternative CRT designs will be more efficient at modeling certain TOI effects. We derive and compare the variance estimators of TOI effects between a SW-CRT, parallel CRT (P-CRT), parallel CRT with baseline (PB-CRT), and novel parallel CRT with baseline and an all-exposed period (PBAE-CRT). We also prove that the time-averaged TOI effect variance and point estimators are identical to that of the constant intervention effect in both P-CRTs and PB-CRTs. We then use data collected from a hospital disinvestment study to simulate and compare the differences in TOI effect estimates between the different CRT designs. Our results reveal that the SW-CRT has the most efficient estimator for the early TOI effect, whereas the PB-CRT typically has the most efficient estimator for the long-term and time-averaged TOI effects. Overall, the PB-CRT with TOI effects can be a more appropriate choice of CRT design for modeling intervention effects that vary by exposure time.

A mixed model approach to estimate the survivor average causal effect in cluster-randomized trials.

In many medical studies, the outcome measure (such as quality of life, QOL) for some study participants becomes informatively truncated (censored, missing, or unobserved) due to death or other forms of dropout, creating a nonignorable missing data problem. In such cases, the use of a composite outcome or imputation methods that fill in unmeasurable QOL values for those who died rely on strong and untestable assumptions and may be conceptually unappealing to certain stakeholders when estimating a treatment effect. The survivor average causal effect (SACE) is an alternative causal estimand that surmounts some of these issues. While principal stratification has been applied to estimate the SACE in individually randomized trials, methods for estimating the SACE in cluster-randomized trials are currently limited. To address this gap, we develop a mixed model approach along with an expectation-maximization algorithm to estimate the SACE in cluster-randomized trials. We model the continuous outcome measure with a random intercept to account for intracluster correlations due to cluster-level randomization, and model the principal strata membership both with and without a random intercept. In simulations, we compare the performance of our approaches with an existing fixed-effects approach to illustrate the importance of accounting for clustering in cluster-randomized trials. The methodology is then illustrated using a cluster-randomized trial of telecare and assistive technology on health-related QOL in the elderly.

Robust analysis of stepped wedge trials using composite likelihood models.

Stepped wedge trials (SWTs) are a type of cluster randomized trial that involve repeated measures on clusters and design-induced confounding between time and treatment. Although mixed models are commonly used to analyze SWTs, they are susceptible to misspecification particularly for cluster-longitudinal designs such as SWTs. Mixed model estimation leverages both "horizontal" or within-cluster information and "vertical" or between-cluster information. To use horizontal information in a mixed model, both the mean model and correlation structure must be correctly specified or accounted for, since time is confounded with treatment and measurements are likely correlated within clusters. Alternative non-parametric methods have been proposed that use only vertical information; these are more robust because between-cluster comparisons in a SWT preserve randomization, but these non-parametric methods are not very efficient. We propose a composite likelihood method that focuses on vertical information, but has the flexibility to recover efficiency by using additional horizontal information. We compare the properties and performance of various methods, using simulations based on COVID-19 data and a demonstration of application to the LIRE trial. We found that a vertical composite likelihood model that leverages baseline data is more robust than traditional methods, and more efficient than methods that use only vertical information. We hope that these results demonstrate the potential value of model-based vertical methods for SWTs with a large number of clusters, and that these new tools are useful to researchers who are concerned about misspecification of traditional models.

Planning stepped wedge cluster randomized trials to detect treatment effect heterogeneity.

Stepped wedge design is a popular research design that enables a rigorous evaluation of candidate interventions by using a staggered cluster randomization strategy. While analytical methods were developed for designing stepped wedge trials, the prior focus has been solely on testing for the average treatment effect. With a growing interest on formal evaluation of the heterogeneity of treatment effects across patient subpopulations, trial planning efforts need appropriate methods to accurately identify sample sizes or design configurations that can generate evidence for both the average treatment effect and variations in subgroup treatment effects. To fill in that important gap, this article derives novel variance formulas for confirmatory analyses of treatment effect heterogeneity, that are applicable to both cross-sectional and closed-cohort stepped wedge designs. We additionally point out that the same framework can be used for more efficient average treatment effect analyses via covariate adjustment, and allows the use of familiar power formulas for average treatment effect analyses to proceed. Our results further sheds light on optimal design allocations of clusters to maximize the weighted precision for assessing both the average and heterogeneous treatment effects. We apply the new methods to the Lumbar Imaging with Reporting of Epidemiology Trial, and carry out a simulation study to validate our new methods.

Power calculation for detecting interaction effect in cross-sectional stepped-wedge cluster randomized trials: an important tool for disparity research.

BACKGROUND: The stepped-wedge cluster randomized trial (SW-CRT) design has become popular in healthcare research. It is an appealing alternative to traditional cluster randomized trials (CRTs) since the burden of logistical issues and ethical problems can be reduced. Several approaches for sample size determination for the overall treatment effect in the SW-CRT have been proposed. However, in certain situations we are interested in examining the heterogeneity in treatment effect (HTE) between groups instead. This is equivalent to testing the interaction effect. An important example includes the aim to reduce racial disparities through healthcare delivery interventions, where the focus is the interaction between the intervention and race. Sample size determination and power calculation for detecting an interaction effect between the intervention status variable and a key covariate in the SW-CRT study has not been proposed yet for binary outcomes. METHODS: We utilize the generalized estimating equation (GEE) method for detecting the heterogeneity in treatment effect (HTE). The variance of the estimated interaction effect is approximated based on the GEE method for the marginal models. The power is calculated based on the two-sided Wald test. The Kauermann and Carroll (KC) and the Mancl and DeRouen (MD) methods along with GEE (GEE-KC and GEE-MD) are considered as bias-correction methods. RESULTS: Among three approaches, GEE has the largest simulated power and GEE-MD has the smallest simulated power. Given cluster size of 120, GEE has over 80% statistical power. When we have a balanced binary covariate (50%), simulated power increases compared to an unbalanced binary covariate (30%). With intermediate effect size of HTE, only cluster sizes of 100 and 120 have more than 80% power using GEE for both correlation structures. With large effect size of HTE, when cluster size is at least 60, all three approaches have more than 80% power. When we compare an increase in cluster size and increase in the number of clusters based on simulated power, the latter has a slight gain in power. When the cluster size changes from 20 to 40 with 20 clusters, power increases from 53.1% to 82.1% for GEE; 50.6% to 79.7% for GEE-KC; and 48.1% to 77.1% for GEE-MD. When the number of clusters changes from 20 to 40 with cluster size of 20, power increases from 53.1% to 82.1% for GEE; 50.6% to 81% for GEE-KC; and 48.1% to 79.8% for GEE-MD. CONCLUSIONS: We propose three approaches for cluster size determination given the number of clusters for detecting the interaction effect in SW-CRT. GEE and GEE-KC have reasonable operating characteristics for both intermediate and large effect size of HTE.

Hierarchical Bayesian modeling of heterogeneous outcome variance in cluster randomized trials.

BACKGROUND: Heterogeneous outcome correlations across treatment arms and clusters have been increasingly acknowledged in cluster randomized trials with binary endpoints, where analytical methods have been developed to study such heterogeneity. However, cluster-specific outcome variances and correlations have yet to be studied for cluster randomized trials with continuous outcomes. METHODS: This article proposes models fitted in the Bayesian setting with hierarchical variance structure to quantify heterogeneous variances across clusters and explain it with cluster-level covariates when the outcome is continuous. The models can also be extended to analyzing heterogeneous variances in individually randomized group treatment trials, with arm-specific cluster-level covariates, or in partially nested designs. Simulation studies are carried out to validate the performance of the newly introduced models across different settings. RESULTS: Simulations showed that overall the newly introduced models have good performance, reporting low bias and approximately 95% coverage for the intraclass correlation coefficients and regression parameters in the variance model. When variances are heterogeneous, our proposed models had improved model fit over models with homogeneous variances. When used to analyze data from the Kerala Diabetes Prevention Program study, our models identified heterogeneous variances and intraclass correlation coefficients across clusters and examined cluster-level characteristics associated with such heterogeneity. CONCLUSION: We proposed new hierarchical Bayesian variance models to accommodate cluster-specific variances in cluster randomized trials. The newly developed methods inform the understanding of how an intervention strategy is implemented and disseminated differently across clusters and can help improve future trial design.

The symbolic two-step method applied to cancer care delivery research: Safeguarding against designing an underpowered cluster randomized trial with a continuous outcome by accounting for the imprecision in the within- and between-center variation.

BACKGROUND: Knowing the predictive factors of the variation in a center-level continuous outcome of interest is valuable in the design and analysis of parallel-arm cluster randomized trials. The symbolic two-step method for sample size planning that we present incorporates this knowledge while simultaneously accounting for patient-level characteristics. Our approach is illustrated through application to cluster randomized trials in cancer care delivery research. The required number of centers (clusters) depends on the between- and within-center variance; the within-center variance is a function of estimates obtained by regressing the log within-center variance on predictive factors. Obtaining accurate estimates of the components needed to characterize the within-center variation is challenging. METHODS: Using our previously derived sample size formula, our objective in the current research is to directly account for the imprecision in these estimates, using a Bayesian approach, to safeguard against designing an underpowered study when using the symbolic two-step method. Using estimates of the required components, including the number of centers that contribute to those estimates, we make formal allowance for the imprecision in these estimates on which a sample size will be based. RESULTS: The mean of the distribution for power is consistently smaller than the single point estimate that the sample size formula yields. The reduction in power is more pronounced in the presence of increased uncertainty about the estimates with the reduction becoming more attenuated with increased numbers of centers that contribute to the estimates. CONCLUSIONS: Accounting for imprecision in the estimates of the components required for sample size estimation using the symbolic two-step method in the design of a cluster randomized trial yields conservative estimates of power.

Multilevel Intervention Stepped Wedge Designs (MLI-SWDs).

Multilevel interventions (MLIs) hold promise for reducing health inequities by intervening at multiple types of social determinants of health consistent with the socioecological model of health. In spite of their potential, methodological challenges related to study design compounded by a lack of tools for sample size calculation inhibit their development. We help address this gap by proposing the Multilevel Intervention Stepped Wedge Design (MLI-SWD), a hybrid experimental design which combines cluster-level (CL) randomization using a Stepped Wedge design (SWD) with independent individual-level (IL) randomization. The MLI-SWD is suitable for MLIs where the IL intervention has a low risk of interference between individuals in the same cluster, and it enables estimation of the component IL and CL treatment effects, their interaction, and the combined intervention effect. The MLI-SWD accommodates cross-sectional and cohort designs as well as both incomplete (clusters are not observed in every study period) and complete observation patterns. We adapt recent work using generalized estimating equations for SWD sample size calculation to the multilevel setting and provide an R package for power and sample size calculation. Furthermore, motivated by our experiences with the ongoing NC Works 4 Health study, we consider how to apply the MLI-SWD when individuals join clusters over the course of the study. This situation arises when unemployment MLIs include IL interventions that are delivered while the individual is unemployed. This extension requires carefully considering whether the study interventions will satisfy additional causal assumptions but could permit randomization in new settings.

Bayesian optimal stepped wedge design.

Recently, there has been a growing interest in designing cluster trials using stepped wedge design (SWD). An SWD is a type of cluster-crossover design in which clusters of individuals are randomized unidirectional from a control to an intervention at certain time points. The intraclass correlation coefficient (ICC) that measures the dependency of subject within a cluster plays an important role in design and analysis of stepped wedge trials. In this paper, we discuss a Bayesian approach to address the dependency of SWD on the ICC and robust Bayesian SWDs are proposed. Bayesian design is shown to be more robust against the misspecification of the parameter values compared to the locally optimal design. Designs are obtained for the various choices of priors assigned to the ICC. A detailed sensitivity analysis is performed to assess the robustness of proposed optimal designs. The power superiority of Bayesian design against the commonly used balanced design is demonstrated numerically using hypothetical as well as real scenarios.

Information content of stepped wedge designs under the working independence assumption.

The stepped wedge design is increasingly popular in pragmatic trials and implementation science research studies for evaluating system-level interventions that are perceived to be beneficial to patient populations. An important step in planning a stepped wedge design is to understand the efficiency of the treatment effect estimator and hence the power of the study. We develop several novel analytical results for designing stepped wedge cluster randomized trials analyzed through generalized estimating equations under a misspecified working independence correlation structure. We first contribute a general variance expression of the treatment effect estimator when data collection is scheduled for each cluster-period. Because resource and patient-centered considerations may intentionally call for an incomplete design with outcome data being omitted for certain cluster-periods, we further derive the information content based on the robust sandwich variance to identify data elements that may be preferentially omitted with minimum loss of precision in estimating the treatment effect. We prove that centrosymmetric pairs of cluster-periods, treatment sequences and periods have identical information content and thus contribute equally to the treatment effect estimation, as long as the true covariance structure for the cluster-period means remains centrosymmetric. Finally, we provide an example of how to obtain an incomplete stepped wedge design that admits a more efficient independence GEE estimator but requires less data collection effort. Our results elegantly extend existing ones from linear mixed models coupled with model-based variances to accommodate a misspecified independence working correlation structure through the robust sandwich variances.

A Bayesian Approach for Estimating the Survivor Average Causal Effect When Outcomes Are Truncated by Death in Cluster-Randomized Trials.

Many studies encounter clustering due to multicenter enrollment and nonmortality outcomes, such as quality of life, that are truncated due to death-that is, missing not at random and nonignorable. Traditional missing-data methods and target causal estimands are suboptimal for statistical inference in the presence of these combined issues, which are especially common in multicenter studies and cluster-randomized trials (CRTs) carried out among the elderly or seriously ill. Using principal stratification, we developed a Bayesian estimator that jointly identifies the always-survivor principal stratum in a clustered/hierarchical data setting and estimates the average treatment effect among them (i.e., the survivor average causal effect (SACE)). In simulations, we observed low bias and good coverage with our method. In a motivating CRT, the SACE and the estimate from complete-case analysis differed in magnitude, but both were small, and neither was incompatible with a null effect. However, the SACE estimate has a clear causal interpretation. The option to assess the rigorously defined SACE estimand in studies with informative truncation and clustering can provide additional insight into an important subset of study participants. Based on the simulation study and CRT reanalysis, we provide practical recommendations for using the SACE in CRTs and software code to support future research.

Marginal modeling of cluster-period means and intraclass correlations in stepped wedge designs with binary outcomes.

Stepped wedge cluster randomized trials (SW-CRTs) with binary outcomes are increasingly used in prevention and implementation studies. Marginal models represent a flexible tool for analyzing SW-CRTs with population-averaged interpretations, but the joint estimation of the mean and intraclass correlation coefficients (ICCs) can be computationally intensive due to large cluster-period sizes. Motivated by the need for marginal inference in SW-CRTs, we propose a simple and efficient estimating equations approach to analyze cluster-period means. We show that the quasi-score for the marginal mean defined from individual-level observations can be reformulated as the quasi-score for the same marginal mean defined from the cluster-period means. An additional mapping of the individual-level ICCs into correlations for the cluster-period means further provides a rigorous justification for the cluster-period approach. The proposed approach addresses a long-recognized computational burden associated with estimating equations defined based on individual-level observations, and enables fast point and interval estimation of the intervention effect and correlations. We further propose matrix-adjusted estimating equations to improve the finite-sample inference for ICCs. By providing a valid approach to estimate ICCs within the class of generalized linear models for correlated binary outcomes, this article operationalizes key recommendations from the CONSORT extension to SW-CRTs, including the reporting of ICCs.

Assessing exposure-time treatment effect heterogeneity in stepped-wedge cluster randomized trials.

A stepped-wedge cluster randomized trial (CRT) is a unidirectional crossover study in which timings of treatment initiation for clusters are randomized. Because the timing of treatment initiation is different for each cluster, an emerging question is whether the treatment effect depends on the exposure time, namely, the time duration since the initiation of treatment. Existing approaches for assessing exposure-time treatment effect heterogeneity either assume a parametric functional form of exposure time or model the exposure time as a categorical variable, in which case the number of parameters increases with the number of exposure-time periods, leading to a potential loss in efficiency. In this article, we propose a new model formulation for assessing treatment effect heterogeneity over exposure time. Rather than a categorical term for each level of exposure time, the proposed model includes a random effect to represent varying treatment effects by exposure time. This allows for pooling information across exposure-time periods and may result in more precise average and exposure-time-specific treatment effect estimates. In addition, we develop an accompanying permutation test for the variance component of the heterogeneous treatment effect parameters. We conduct simulation studies to compare the proposed model and permutation test to alternative methods to elucidate their finite-sample operating characteristics, and to generate practical guidance on model choices for assessing exposure-time treatment effect heterogeneity in stepped-wedge CRTs.

Sample size calculation for randomized trials via inverse probability of response weighting when outcome data are missing at random.

Randomized trials are an established method to evaluate the causal effects of interventions. Despite concerted efforts to retain all trial participants, some missing outcome data are often inevitable. It is unclear how best to account for missing outcome data in sample size calculations. A standard approach is to inflate the sample size by the inverse of one minus the anticipated dropout probability. However, the performance of this approach in the presence of informative outcome missingness has not been well-studied. We investigate sample size calculation when outcome data are missing at random given the randomized intervention group and fully observed baseline covariates under an inverse probability of response weighted (IPRW) estimating equations approach. Using M-estimation theory, we derive sample size formulas for both individually randomized and cluster randomized trials (CRTs). We illustrate the proposed method by calculating a sample size for a CRT designed to detect a difference in HIV testing strategies under an IPRW approach. We additionally develop an R shiny app to facilitate implementation of the sample size formulas.

Defining and estimating effects in cluster randomized trials: A methods comparison.

Across research disciplines, cluster randomized trials (CRTs) are commonly implemented to evaluate interventions delivered to groups of participants, such as communities and clinics. Despite advances in the design and analysis of CRTs, several challenges remain. First, there are many possible ways to specify the causal effect of interest (eg, at the individual-level or at the cluster-level). Second, the theoretical and practical performance of common methods for CRT analysis remain poorly understood. Here, we present a general framework to formally define an array of causal effects in terms of summary measures of counterfactual outcomes. Next, we provide a comprehensive overview of CRT estimators, including the t-test, generalized estimating equations (GEE), augmented-GEE, and targeted maximum likelihood estimation (TMLE). Using finite sample simulations, we illustrate the practical performance of these estimators for different causal effects and when, as commonly occurs, there are limited numbers of clusters of different sizes. Finally, our application to data from the Preterm Birth Initiative (PTBi) study demonstrates the real-world impact of varying cluster sizes and targeting effects at the cluster-level or at the individual-level. Specifically, the relative effect of the PTBi intervention was 0.81 at the cluster-level, corresponding to a 19% reduction in outcome incidence, and was 0.66 at the individual-level, corresponding to a 34% reduction in outcome risk. Given its flexibility to estimate a variety of user-specified effects and ability to adaptively adjust for covariates for precision gains while maintaining Type-I error control, we conclude TMLE is a promising tool for CRT analysis.

Comparing analytical strategies for balancing site-level characteristics in stepped-wedge cluster randomized trials: a simulation study.

BACKGROUND: Stepped-wedge cluster randomized trials (SWCRTs) are a type of cluster-randomized trial in which clusters are randomized to cross-over to the active intervention sequentially at regular intervals during the study period. For SWCRTs, sequential imbalances of cluster-level characteristics across the random sequence of clusters may lead to biased estimation. Our study aims to examine the effects of balancing cluster-level characteristics in SWCRTs. METHODS: To quantify the level of cluster-level imbalance, a novel imbalance index was developed based on the Spearman correlation and rank regression of the cluster-level characteristic with the cross-over timepoints. A simulation study was conducted to assess the impact of sequential cluster-level imbalances across different scenarios varying the: number of sites (clusters), sample size, number of cross-over timepoints, site-level intra-cluster correlation coefficient (ICC), and effect sizes. SWCRTs assumed either an immediate "constant" treatment effect, or a gradual "learning" treatment effect which increases over time after crossing over to the active intervention. Key performance metrics included the relative root mean square error (RRMSE) and relative mean bias. RESULTS: Fully-balanced designs almost always had the highest efficiency, as measured by the RRMSE, regardless of the number of sites, ICC, effect size, or sample sizes at each time for SWCRTs with learning effect. A consistent decreasing trend of efficiency was observed by increasing RRMSE as imbalance increased. For example, for a 12-site study with 20 participants per site/timepoint and ICC of 0.10, between the most balanced and least balanced designs, the RRMSE efficiency loss ranged from 52.5% to 191.9%. In addition, the RRMSE was decreased for larger sample sizes, larger number of sites, smaller ICC, and larger effect sizes. The impact of pre-balancing diminished when there was no learning effect. CONCLUSION: The impact of pre-balancing on preventing efficiency loss was easily observed when there was a learning effect. This suggests benefit of pre-balancing with respect to impacting factors of treatment effects.

Constructing within and between hospital physician social networks for modeling physician research participation.

BACKGROUND: Physician participation in clinical trials is essential for the progress of modern medicine. However, the demand for physician research partners is outpacing physicians' interest in participating in scientific studies. Understanding the factors that influence physician participation in research is crucial to addressing this gap. METHODS: In this study, we used a physician's social network, as constructed from patient billing data, to study if the research choices of a physician's immediate peers influence their likelihood to participate in scientific research. We analyzed data from 348 physicians across 40 hospitals. We used logistic regression models to examine the relationship between a physician's participation in clinical trials and the participation of their social network peers, adjusting for age, years of employment, and influences from other hospital facilities. RESULTS: We found that the likelihood of a physician participating in clinical trials increased dramatically with the proportion of their social network-defined colleagues at their primary hospital who were participating ([Formula: see text] for a 1% increase in the proportion of participating peers, [Formula: see text]). Additionally, physicians who work regularly at multiple facilities were more likely to participate ([Formula: see text], [Formula: see text]) and increasingly so as the extent to which they have social network ties to colleagues at hospitals other than their primary hospital increases ([Formula: see text], [Formula: see text]). These findings suggest an inter-hospital peer participation process. CONCLUSION: Our study provides evidence that the social structure of a physician's work-life is associated with their decision to participate in scientific research. The results suggest that interventions aimed at increasing physician participation in clinical trials could leverage the social networks of physicians to encourage participation. By identifying factors that influence physician participation in research, we can work towards closing the gap between the demand for physician research partners and the number of physicians willing to participate in scientific studies.

Considerations for Subgroup Analyses in Cluster-Randomized Trials Based on Aggregated Individual-Level Predictors.

In research assessing the effect of an intervention or exposure, a key secondary objective often involves assessing differential effects of this intervention or exposure in subgroups of interest; this is often referred to as assessing effect modification or heterogeneity of treatment effects (HTE). Observed HTE can have important implications for policy, including intervention strategies (e.g., will some patients benefit more from intervention than others?) and prioritizing resources (e.g., to reduce observed health disparities). Analysis of HTE is well understood in studies where the independent unit is an individual. In contrast, in studies where the independent unit is a cluster (e.g., a hospital or school) and a cluster-level outcome is used in the analysis, it is less well understood how to proceed if the HTE analysis of interest involves an individual-level characteristic (e.g., self-reported race) that must be aggregated at the cluster level. Through simulations, we show that only individual-level models have power to detect HTE by individual-level variables; if outcomes must be defined at the cluster level, then there is often low power to detect HTE by the corresponding aggregated variables. We illustrate the challenges inherent to this type of analysis in a study assessing the effect of an intervention on increasing COVID-19 booster vaccination rates at long-term care centers.