Designing individually randomized group treatment trials with repeated outcome measurements using generalized estimating equations.

Individually randomized group treatment (IRGT) trials, in which the clustering of outcome is induced by group-based treatment delivery, are increasingly popular in public health research. IRGT trials frequently incorporate longitudinal measurements, of which the proper sample size calculations should account for correlation structures reflecting both the treatment-induced clustering and repeated outcome measurements. Given the relatively sparse literature on designing longitudinal IRGT trials, we propose sample size procedures for continuous and binary outcomes based on the generalized estimating equations approach, employing the block exchangeable correlation structures with different correlation parameters for the treatment arm and for the control arm, and surveying five marginal mean models with different assumptions of time effect: no-time constant treatment effect, linear-time constant treatment effect, categorical-time constant treatment effect, linear time by treatment interaction, and categorical time by treatment interaction. Closed-form sample size formulas are derived for continuous outcomes, which depends on the eigenvalues of the correlation matrices; detailed numerical sample size procedures are proposed for binary outcomes. Through simulations, we demonstrate that the empirical power agrees well with the predicted power, for as few as eight groups formed in the treatment arm, when data are analyzed using the matrix-adjusted estimating equations for the correlation parameters with a bias-corrected sandwich variance estimator.

A Community-Based Intervention for Managing Hypertension in Rural South Asia.

BACKGROUND: The burden of hypertension is escalating, and control rates are poor in low- and middle-income countries. Cardiovascular mortality is high in rural areas. METHODS: We conducted a cluster-randomized, controlled trial in rural districts in Bangladesh, Pakistan, and Sri Lanka. A total of 30 communities were randomly assigned to either a multicomponent intervention (intervention group) or usual care (control group). The intervention involved home visits by trained government community health workers for blood-pressure monitoring and counseling, training of physicians, and care coordination in the public sector. A total of 2645 adults with hypertension were enrolled. The primary outcome was reduction in systolic blood pressure at 24 months. Follow-up at 24 months was completed for more than 90% of the participants. RESULTS: At baseline, the mean systolic blood pressure was 146.7 mm Hg in the intervention group and 144.7 mm Hg in the control group. At 24 months, the mean systolic blood pressure fell by 9.0 mm Hg in the intervention group and by 3.9 mm Hg in the control group; the mean reduction was 5.2 mm Hg greater with the intervention (95% confidence interval [CI], 3.2 to 7.1; P<0.001). The mean reduction in diastolic blood pressure was 2.8 mm Hg greater in the intervention group than in the control group (95% CI, 1.7 to 3.9). Blood-pressure control (<140/90 mm Hg) was achieved in 53.2% of the participants in the intervention group, as compared with 43.7% of those in the control group (relative risk, 1.22; 95% CI, 1.10 to 1.35). All-cause mortality was 2.9% in the intervention group and 4.3% in the control group. CONCLUSIONS: In rural communities in Bangladesh, Pakistan, and Sri Lanka, a multicomponent intervention that was centered on proactive home visits by trained government community health workers who were linked with existing public health care infrastructure led to a greater reduction in blood pressure than usual care among adults with hypertension. (Funded by the Joint Global Health Trials scheme; COBRA-BPS ClinicalTrials.gov number, NCT02657746.).

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.

Telemedicine Referral to Improve Access to Specialty Care for Preschool Children in Rural Alaska: A Cluster-Randomized Controlled Trial.

OBJECTIVES: Preschool programs provide essential preventive services, such as hearing screening, but in rural regions, limited access to specialists and loss to follow-up compound rural health disparities. We conducted a parallel-arm cluster-randomized controlled trial to evaluate telemedicine specialty referral for preschool hearing screening. The goal of this trial was to improve timely identification and treatment of early childhood infection-related hearing loss, a preventable condition with lifelong implications. We hypothesized that telemedicine specialty referral would improve time to follow-up and the number of children receiving follow-up compared with the standard primary care referral. DESIGN: We conducted a cluster-randomized controlled trial in K-12 schools in 15 communities over two academic years. Community randomization occurred within four strata using location and school size. In the second academic year (2018-2019), an ancillary trial was performed in the 14 communities that had preschools to compare telemedicine specialty referral (intervention) to standard primary care referral (comparison) for preschool hearing screening. Randomization of communities from the main trial was used for this ancillary trial. All children enrolled in preschool were eligible. Masking was not possible because of timing in the second year of the main trial, but referral assignment was not openly disclosed. Study team members and school staff were masked throughout data collection, and statisticians were blinded to allocation during analysis. Preschool screening occurred once, and children who were referred for possible hearing loss or ear disease were monitored for follow-up for 9 months from the screening date. The primary outcome was time to ear/hearing-related follow-up from the date of screening. The secondary outcome was any ear/hearing follow-up from screening to 9 months. Analyses were conducted using an intention-to-treat approach. RESULTS: A total of 153 children were screened between September 2018 and March 2019. Of the 14 communities, 8 were assigned to the telemedicine specialty referral pathway (90 children), and 6 to the standard primary care referral pathway (63 children). Seventy-one children (46.4%) were referred for follow-up: 39 (43.3%) in the telemedicine specialty referral communities and 32 (50.8%) in the standard primary care referral communities. Of children referred, 30 (76.9%) children in telemedicine specialty referral communities and 16 (50.0%) children in standard primary care referral communities received follow-up within 9 months (Risk Ratio = 1.57; 95% confidence interval [CI], 1.22 to 2.01). Among children who received follow-up, median time to follow-up was 28 days (interquartile range [IQR]: 15 to 71) in telemedicine specialty referral communities compared with 85 days (IQR: 26 to 129) in standard primary care referral communities. Mean time to follow-up for all referred children was 4.5 (event time ratio = 4.5; 95% CI, 1.8 to 11.4; p = 0.045) times faster in telemedicine specialty referral communities compared with standard primary care referral communities in the 9-month follow-up time frame. CONCLUSIONS: Telemedicine specialty referral significantly improved follow-up and reduced time to follow-up after preschool hearing screening in rural Alaska. Telemedicine referrals could extend to other preventive school-based services to improve access to specialty care for rural preschool children.

Improving sandwich variance estimation for marginal Cox analysis of cluster randomized trials.

Cluster randomized trials (CRTs) frequently recruit a small number of clusters, therefore necessitating the application of small-sample corrections for valid inference. A recent systematic review indicated that CRTs reporting right-censored, time-to-event outcomes are not uncommon and that the marginal Cox proportional hazards model is one of the common approaches used for primary analysis. While small-sample corrections have been studied under marginal models with continuous, binary, and count outcomes, no prior research has been devoted to the development and evaluation of bias-corrected sandwich variance estimators when clustered time-to-event outcomes are analyzed by the marginal Cox model. To improve current practice, we propose nine bias-corrected sandwich variance estimators for the analysis of CRTs using the marginal Cox model and report on a simulation study to evaluate their small-sample properties. Our results indicate that the optimal choice of bias-corrected sandwich variance estimator for CRTs with survival outcomes can depend on the variability of cluster sizes and can also slightly differ whether it is evaluated according to relative bias or type I error rate. Finally, we illustrate the new variance estimators in a real-world CRT where the conclusion about intervention effectiveness differs depending on the use of small-sample bias corrections. The proposed sandwich variance estimators are implemented in an R package CoxBcv.

When Shepherds Shed: Trajectories of Weight-Related Behaviors in a Holistic Health Intervention Tailored for US Christian Clergy.

Maintaining healthy behaviors is challenging. Based upon previous reports that in North Carolina (NC), USA, overweight/obese clergy lost weight during a two-year religiously tailored health intervention, we described trajectories of diet, physical activity, and sleep. We investigated whether behavior changes were associated with weight and use of health-promoting theological messages. Improvements were observed in sleep, calorie-dense food intake, and physical activity, with the latter two associated with weight loss. While theological messages were well-retained, their relationship with behaviors depended on the specific message, behavior, and timing. Findings offer insights into weight loss mechanisms, including the role of theological messages in religiously tailored health interventions.

Constrained randomization and statistical inference for multi-arm parallel cluster randomized controlled trials.

A practical limitation of cluster randomized controlled trials (cRCTs) is that the number of available clusters may be small, resulting in an increased risk of baseline imbalance under simple randomization. Constrained randomization overcomes this issue by restricting the allocation to a subset of randomization schemes where sufficient overall covariate balance across comparison arms is achieved. However, for multi-arm cRCTs, several design and analysis issues pertaining to constrained randomization have not been fully investigated. Motivated by an ongoing multi-arm cRCT, we elaborate the method of constrained randomization and provide a comprehensive evaluation of the statistical properties of model-based and randomization-based tests under both simple and constrained randomization designs in multi-arm cRCTs, with varying combinations of design and analysis-based covariate adjustment strategies. In particular, as randomization-based tests have not been extensively studied in multi-arm cRCTs, we additionally develop most-powerful randomization tests under the linear mixed model framework for our comparisons. Our results indicate that under constrained randomization, both model-based and randomization-based analyses could gain power while preserving nominal type I error rate, given proper analysis-based adjustment for the baseline covariates. Randomization-based analyses, however, are more robust against violations of distributional assumptions. The choice of balance metrics and candidate set sizes and their implications on the testing of the pairwise and global hypotheses are also discussed. Finally, we caution against the design and analysis of multi-arm cRCTs with an extremely small number of clusters, due to insufficient degrees of freedom and the tendency to obtain an overly restricted randomization space.

Healthy eating and all-cause mortality among Chinese aged 80 years or older.

BACKGROUND: There is little evidence of the influence of dietary patterns on mortality risk among adults 80 years or older ("oldest-old"). We evaluated the association between the Simplified Healthy Eating index (SHE-index) and mortality among Chinese oldest-old. METHODS: Population-based cohort study from the Chinese Longitudinal Healthy Longevity Survey (CLHLS 1998-2014, n = 35 927), conducted in 22 Chinese provinces, were pooled for analysis. The first seven waves of the CLHLS (1998, 2000, 2002, 2005, 2008-09, 2011-12, and 2013-2014) were utilized, with follow-up to the last wave (2018) (range 0-21 years). The SHE-index was collected in each wave, and was constructed from intake frequency of nine dietary variables, with a higher score indicating better diet quality. Cox proportional hazards model with dietary patterns as a time-varying exposure was employed to analyze the relationship between SHE-index and mortality. RESULTS: At baseline, the median age of all participants was 92 years (25th percentile, 85 years; 75th percentile, 100 years). In multivariable models, the hazard ratios (95% confidence intervals) for SHE-index quartile 2, quartile 3 and quartile 4 versus quartile1 were 0.91 (0.88, 0.93), 0.89 (0.86, 0.92) and 0.82 (0.78, 0.85), respectively. Results were generally consistent for men and women and in a large number of sensitivity analyses. CONCLUSIONS: Healthier eating patterns were associated with a significant reduction in the risk of all-cause mortality among Chinese oldest-old, lending support to the importance of life-long adherence to healthy diet into advanced old age.

Sauti ya Vijana (SYV; The Voice of Youth): Longitudinal Outcomes of an Individually Randomized Group Treatment Pilot Trial for Young People Living with HIV in Tanzania.

Sauti ya Vijana is a mental health and life skills intervention delivered by young adult group leaders for the improvement of HIV outcomes in young people living with HIV in Tanzania. This pilot randomized controlled trial estimated exploratory intervention effectiveness compared to standard of care. YPLWH (N = 105) were randomized to receive intervention or SOC. The mean age of participants was 18.1 years and 53% were female. Mean scores on mental health measures (Patient Health Questionnaire [PHQ-9], Strengths and Difficulties Questionnaire [SDQ], UCLA Trauma) were asymptomatic to mild in both study arms through 30-month follow-up with a non-significant fluctuation of 1-2 points. The mean self-reported adherence was higher in the intervention arm across all time points (but the confidence interval contained the null at all time points except 6 months). Risk ratio of virologic suppression (HIV RNA < 400 copies/mL) in the intervention arm compared to SOC was 1.15 [95% CI = 0.95, 1.39]) at 6-months, 1.17 [95% CI: 0.92, 1.48] at 12-months, and 0.99 [95% CI 0.76, 1.31] at 18-months. Though these findings were not powered for statistical significance, the trends in HIV outcomes suggest that SYV holds promise for improving antiretroviral therapy (ART) adherence and virologic suppression in YPLWH.

power swgee: GEE-based power calculations in stepped wedge cluster randomized trials.

Stepped wedge cluster randomized trials are increasingly being used to evaluate interventions in medical, public health, educational, and social science contexts. With the longitudinal and crossover nature of a SW-CRT, complex analysis techniques are often needed which makes appropriately powering SW-CRTs challenging. In this paper, we introduce a newly-developed SW-CRT power calculator, embedded within the power command in Stata. The power calculator assumes a marginal model (i.e., generalized estimating equations [GEE]) for the primary analysis of SW-CRTs, for which other currently available SW-CRT power calculators may not be suitable. The program accommodates complete cross-sectional and closed-cohort designs, and includes multilevel correlation structures appropriate for such designs. We discuss the methods and formulae underlying our SW-CRT calculator, and provide illustrative examples of the use of power swgee. We provide suggestions about the choice of parameters in power swgee, and conclude by discussing areas of future research which may improve the program.

Power considerations for generalized estimating equations analyses of four-level cluster randomized trials.

In this article, we develop methods for sample size and power calculations in four-level intervention studies when intervention assignment is carried out at any level, with a particular focus on cluster randomized trials (CRTs). CRTs involving four levels are becoming popular in healthcare research, where the effects are measured, for example, from evaluations (level 1) within participants (level 2) in divisions (level 3) that are nested in clusters (level 4). In such multilevel CRTs, we consider three types of intraclass correlations between different evaluations to account for such clustering: that of the same participant, that of different participants from the same division, and that of different participants from different divisions in the same cluster. Assuming arbitrary link and variance functions, with the proposed correlation structure as the true correlation structure, closed-form sample size formulas for randomization carried out at any level (including individually randomized trials within a four-level clustered structure) are derived based on the generalized estimating equations approach using the model-based variance and using the sandwich variance with an independence working correlation matrix. We demonstrate that empirical power corresponds well with that predicted by the proposed method for as few as eight clusters, when data are analyzed using the matrix-adjusted estimating equations for the correlation parameters with a bias-corrected sandwich variance estimator, under both balanced and unbalanced designs.

Impact of unequal cluster sizes for GEE analyses of stepped wedge cluster randomized trials with binary outcomes.

The stepped wedge (SW) design is a type of unidirectional crossover design where cluster units switch from control to intervention condition at different prespecified time points. While a convention in study planning is to assume the cluster-period sizes are identical, SW cluster randomized trials (SW-CRTs) involving repeated cross-sectional designs frequently have unequal cluster-period sizes, which can impact the efficiency of the treatment effect estimator. In this paper, we provide a comprehensive investigation of the efficiency impact of unequal cluster sizes for generalized estimating equation analyses of SW-CRTs, with a focus on binary outcomes as in the Washington State Expedited Partner Therapy trial. Several major distinctions between our work and existing work include the following: (i) we consider multilevel correlation structures in marginal models with binary outcomes; (ii) we study the implications of both the between-cluster and within-cluster imbalances in sizes; and (iii) we provide a comparison between the independence working correlation versus the true working correlation and detail the consequences of ignoring correlation estimation in SW-CRTs with unequal cluster sizes. We conclude that the working independence assumption can lead to substantial efficiency loss and a large sample size regardless of cluster-period size variability in SW-CRTs, and recommend accounting for correlations in the analysis. To improve study planning, we additionally provide a computationally efficient search algorithm to estimate the sample size in SW-CRTs accounting for unequal cluster-period sizes, and conclude by illustrating the proposed approach in the context of the Washington State study.

Effectiveness of a primary care-based integrated mobile health intervention for stroke management in rural China (SINEMA): A cluster-randomized controlled trial.

BACKGROUND: Managing noncommunicable diseases through primary healthcare has been identified as the key strategy to achieve universal health coverage but is challenging in most low- and middle-income countries. Stroke is the leading cause of death and disability in rural China. This study aims to determine whether a primary care-based integrated mobile health intervention (SINEMA intervention) could improve stroke management in rural China. METHODS AND FINDINGS: Based on extensive barrier analyses, contextual research, and feasibility studies, we conducted a community-based, two-arm cluster-randomized controlled trial with blinded outcome assessment in Hebei Province, rural Northern China including 1,299 stroke patients (mean age: 65.7 [SD:8.2], 42.6% females, 71.2% received education below primary school) recruited from 50 villages between June 23 and July 21, 2017. Villages were randomly assigned (1:1) to either the intervention or control arm (usual care). In the intervention arm, village doctors who were government-sponsored primary healthcare providers received training, conducted monthly follow-up visits supported by an Android-based mobile application, and received performance-based payments. Participants received monthly doctor visits and automatically dispatched daily voice messages. The primary outcome was the 12-month change in systolic blood pressure (BP). Secondary outcomes were predefined, including diastolic BP, health-related quality of life, physical activity level, self-reported medication adherence (antiplatelet, statin, and antihypertensive), and performance in "timed up and go" test. Analyses were conducted in the intention-to-treat framework at the individual level with clusters and stratified design accounted for by following the prepublished statistical analysis plan. All villages completed the 12-month follow-up, and 611 (intervention) and 615 (control) patients were successfully followed (3.4% lost to follow-up among survivors). The program was implemented with high fidelity, and the annual program delivery cost per capita was US$24.3. There was a significant reduction in systolic BP in the intervention as compared with the control group with an adjusted mean difference: -2.8 mm Hg (95% CI -4.8, -0.9; p = 0.005). The intervention was significantly associated with improvements in 6 out of 7 secondary outcomes in diastolic BP reduction (p < 0.001), health-related quality of life (p = 0.008), physical activity level (p < 0.001), adherence in statin (p = 0.003) and antihypertensive medicines (p = 0.039), and performance in "timed up and go" test (p = 0.022). We observed reductions in all exploratory outcomes, including stroke recurrence (4.4% versus 9.3%; risk ratio [RR] = 0.46, 95% CI 0.32, 0.66; risk difference [RD] = 4.9 percentage points [pp]), hospitalization (4.4% versus 9.3%; RR = 0.45, 95% CI 0.32, 0.62; RD = 4.9 pp), disability (20.9% versus 30.2%; RR = 0.65, 95% CI 0.53, 0.79; RD = 9.3 pp), and death (1.8% versus 3.1%; RR = 0.52, 95% CI 0.28, 0.96; RD = 1.3 pp). Limitations include the relatively short study duration of only 1 year and the generalizability of our findings beyond the study setting. CONCLUSIONS: In this study, a primary care-based mobile health intervention integrating provider-centered and patient-facing technology was effective in reducing BP and improving stroke secondary prevention in a resource-limited rural setting in China. TRIAL REGISTRATION: ClinicalTrials.gov NCT03185858.

Effectiveness of Group Problem Management Plus, a brief psychological intervention for adults affected by humanitarian disasters in Nepal: A cluster randomized controlled trial.

BACKGROUND: Globally, 235 million people are impacted by humanitarian emergencies worldwide, presenting increased risk of experiencing a mental disorder. Our objective was to test the effectiveness of a brief group psychological treatment delivered by trained facilitators without prior professional mental health training in a disaster-prone setting. METHODS AND FINDINGS: We conducted a cluster randomized controlled trial (cRCT) from November 25, 2018 through September 30, 2019. Participants in both arms were assessed at baseline, midline (7 weeks post-baseline, which was approximately 1 week after treatment in the experimental arm), and endline (20 weeks post-baseline, which was approximately 3 months posttreatment). The intervention was Group Problem Management Plus (PM+), a psychological treatment of 5 weekly sessions, which was compared with enhanced usual care (EUC) consisting of a family psychoeducation meeting with a referral option to primary care providers trained in mental healthcare. The setting was 72 wards (geographic unit of clustering) in eastern Nepal, with 1 PM+ group per ward in the treatment arm. Wards were eligible if they were in disaster-prone regions and residents spoke Nepali. Wards were assigned to study arms based on covariate constrained randomization. Eligible participants were adult women and men 18 years of age and older who met screening criteria for psychological distress and functional impairment. Outcomes were measured at the participant level, with assessors blinded to group assignment. The primary outcome was psychological distress assessed with the General Health Questionnaire (GHQ-12). Secondary outcomes included depression symptoms, posttraumatic stress disorder (PTSD) symptoms, "heart-mind" problems, social support, somatic symptoms, and functional impairment. The hypothesized mediator was skill use aligned with the treatment's mechanisms of action. A total of 324 participants were enrolled in the control arm (36 wards) and 319 in the Group PM+ arm (36 wards). The overall sample (N = 611) had a median age of 45 years (range 18-91 years), 82% of participants were female, 50% had recently experienced a natural disaster, and 31% had a chronic physical illness. Endline assessments were completed by 302 participants in the control arm (36 wards) and 303 participants in the Group PM+ arm (36 wards). At the midline assessment (immediately after Group PM+ in the experimental arm), mean GHQ-12 total score was 2.7 units lower in Group PM+ compared to control (95% CI: 1.7, 3.7, p < 0.001), with standardized mean difference (SMD) of -0.4 (95% CI: -0.5, -0.2). At 3 months posttreatment (primary endpoint), mean GHQ-12 total score was 1.4 units lower in Group PM+ compared to control (95% CI: 0.3, 2.5, p = 0.014), with SMD of -0.2 (95% CI: -0.4, 0.0). Among the secondary outcomes, Group PM+ was associated with endline with a larger proportion attaining more than 50% reduction in depression symptoms (29.9% of Group PM+ arm versus 17.3% of control arm, risk ratio = 1.7, 95% CI: 1.2, 2.4, p = 0.002). Fewer participants in the Group PM+ arm continued to have "heart-mind" problems at endline (58.8%) compared to the control arm (69.4%), risk ratio = 0.8 (95% CI, 0.7, 1.0, p = 0.042). Group PM+ was not associated with lower PTSD symptoms or functional impairment. Use of psychosocial skills at midline was estimated to explain 31% of the PM+ effect on endline GHQ-12 scores. Adverse events in the control arm included 1 suicide death and 1 reportable incidence of domestic violence; in the Group PM+ arm, there was 1 death due to physical illness. Study limitations include lack of power to evaluate gender-specific effects, lack of long-term outcomes (e.g., 12 months posttreatment), and lack of cost-effectiveness information. CONCLUSIONS: In this study, we found that a 5-session group psychological treatment delivered by nonspecialists modestly reduced psychological distress and depression symptoms in a setting prone to humanitarian emergencies. Benefits were partly explained by the degree of psychosocial skill use in daily life. To improve the treatment benefit, future implementation should focus on approaches to enhance skill use by PM+ participants. TRIAL REGISTRATION: ClinicalTrials.gov NCT03747055.

Intra-cluster correlations from the CLustered OUtcome Dataset bank to inform the design of longitudinal cluster trials.

BACKGROUND: Sample size calculations for longitudinal cluster randomised trials, such as crossover and stepped-wedge trials, require estimates of the assumed correlation structure. This includes both within-period intra-cluster correlations, which importantly differ from conventional intra-cluster correlations by their dependence on period, and also cluster autocorrelation coefficients to model correlation decay. There are limited resources to inform these estimates. In this article, we provide a repository of correlation estimates from a bank of real-world clustered datasets. These are provided under several assumed correlation structures, namely exchangeable, block-exchangeable and discrete-time decay correlation structures. METHODS: Longitudinal studies with clustered outcomes were collected to form the CLustered OUtcome Dataset bank. Forty-four available continuous outcomes from 29 datasets were obtained and analysed using each correlation structure. Patterns of within-period intra-cluster correlation coefficient and cluster autocorrelation coefficients were explored by study characteristics. RESULTS: The median within-period intra-cluster correlation coefficient for the discrete-time decay model was 0.05 (interquartile range: 0.02-0.09) with a median cluster autocorrelation of 0.73 (interquartile range: 0.19-0.91). The within-period intra-cluster correlation coefficients were similar for the exchangeable, block-exchangeable and discrete-time decay correlation structures. Within-period intra-cluster correlation coefficients and cluster autocorrelations were found to vary with the number of participants per cluster-period, the period-length, type of cluster (primary care, secondary care, community or school) and country income status (high-income country or low- and middle-income country). The within-period intra-cluster correlation coefficients tended to decrease with increasing period-length and slightly decrease with increasing cluster-period sizes, while the cluster autocorrelations tended to move closer to 1 with increasing cluster-period size. Using the CLustered OUtcome Dataset bank, an RShiny app has been developed for determining plausible values of correlation coefficients for use in sample size calculations. DISCUSSION: This study provides a repository of intra-cluster correlations and cluster autocorrelations for longitudinal cluster trials. This can help inform sample size calculations for future longitudinal cluster randomised trials.

Essential Ingredients and Innovations in the Design and Analysis of Group-Randomized Trials.

This article reviews the essential ingredients and innovations in the design and analysis of group-randomized trials. The methods literature for these trials has grown steadily since they were introduced to the biomedical research community in the late 1970s, and we summarize those developments. We review, in addition to the group-randomized trial, methods for two closely related designs, the individually randomized group treatment trial and the stepped-wedge group-randomized trial. After describing the essential ingredients for these designs, we review the most important developments in the evolution of their methods using a new bibliometric tool developed at the National Institutes of Health. We then discuss the questions to be considered when selecting from among these designs or selecting the traditional randomized controlled trial. We close with a review of current methods for the analysis of data from these designs, a case study to illustrate each design, and a brief summary.

Advancing successful implementation of task-shifted mental health care in low-resource settings (BASIC): protocol for a stepped wedge cluster randomized trial.

BACKGROUND: The mental health treatment gap-the difference between those with mental health need and those who receive treatment-is high in low- and middle-income countries. Task-shifting has been used to address the shortage of mental health professionals, with a growing body of research demonstrating the effectiveness of mental health interventions delivered through task-shifting. However, very little research has focused on how to embed, support, and sustain task-shifting in government-funded systems with potential for scale up. The goal of the Building and Sustaining Interventions for Children (BASIC) study is to examine implementation policies and practices that predict adoption, fidelity, and sustainment of a mental health intervention in the education sector via teacher delivery and the health sector via community health volunteer delivery. METHODS: BASIC is a Hybrid Type II Implementation-Effectiveness trial. The study design is a stepped wedge, cluster randomized trial involving 7 sequences of 40 schools and 40 communities surrounding the schools. Enrollment consists of 120 teachers, 120 community health volunteers, up to 80 site leaders, and up to 1280 youth and one of their primary guardians. The evidence-based mental health intervention is a locally adapted version of Trauma-focused Cognitive Behavioral Therapy, called Pamoja Tunaweza. Lay counselors are trained and supervised in Pamoja Tunaweza by local trainers who are experienced in delivering the intervention and who participated in a Train-the-Trainer model of skills transfer. After the first sequence completes implementation, in-depth interviews are conducted with initial implementing sites' counselors and leaders. Findings are used to inform delivery of implementation facilitation for subsequent sequences' sites. We use a mixed methods approach including qualitative comparative analysis to identify necessary and sufficient implementation policies and practices that predict 3 implementation outcomes of interest: adoption, fidelity, and sustainment. We also examine child mental health outcomes and cost of the intervention in both the education and health sectors. DISCUSSION: The BASIC study will provide knowledge about how implementation of task-shifted mental health care can be supported in government systems that already serve children and adolescents. Knowledge about implementation policies and practices from BASIC can advance the science of implementation in low-resource contexts. TRIAL REGISTRATION: Trial Registration: ClinicalTrials.gov Identifier: NCT03243396. Registered 9th August 2017, https://clinicaltrials.gov/ct2/show/NCT03243396.

A group-based mental health intervention for young people living with HIV in Tanzania: results of a pilot individually randomized group treatment trial.

BACKGROUND: Increasing numbers of young people living with HIV (YPLWH) have unaddressed mental health challenges. Such challenges are associated with poor antiretroviral therapy (ART) adherence and high mortality. Few evidence-based mental health interventions exist to improve HIV outcomes among YPLWH. METHODS: This pilot group treatment trial individually randomized YPLWH from two clinical sites in Tanzania, evaluated acceptability, feasibility, and preliminary effectiveness of a mental health intervention, Sauti ya Vijana (SYV; The Voice of Youth), was compared to the local standard-of-care (SOC) for improving ART adherence and virologic suppression. Enrolled YPLWH were 12-24 years of age and responded to mental health and stigma questionnaires, self-reported adherence, objective adherence measures (ART concentration in hair), and HIV RNA at baseline and 6-months (post-intervention). Feasibility and acceptability were evaluated, and potential effectiveness was assessed by comparing outcomes between arms using mixed effects modeling. RESULTS: Between June 2016 and July 2017, 128 YPLWH enrolled; 105 were randomized and 93 (55 in SYV) followed-up at 6-months and were thereby included in this analysis. Mean age was 18.1 years; 51% were female; and 84% were HIV-infected perinatally. Attendance to intervention sessions was 86%; 6-month follow-up was 88%, and fidelity to the protocol approached 100%. Exploratory analyses of effectiveness demonstrated self-reported adherence improved by 7.3 percentage points (95% CI: 2.2, 12.3); and the pooled standard deviation for all ART concentration values increased by 0.17 units (95% CI: - 0.52, 0.85) in the SYV arm compared to SOC. Virologic suppression rates (HIV RNA < 400 copies/mL) at baseline were 65% in both arms but increased to 75% in the SYV arm while staying the same in the SOC arm (RR 1.13; 95% CI: 0.94, 1.36). CONCLUSIONS: YPLWH often have poor HIV outcomes, making interventions to improve outcomes in this population critical. This pilot trial of the Tanzania-based SYV intervention demonstrated trends towards improvement in ART adherence and virologic outcomes among YPLWH, supporting efforts to scale the intervention into a fully-powered effectiveness trial. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT02888288 . Registered August 9, 2016. Retrospectively registered as first participant enrolled June 16, 2016.

Maternal depression in rural Pakistan: the protective associations with cultural postpartum practices.

BACKGROUND: Traditional postpartum practices are intended to provide care to mothers, but there is mixed evidence concerning their impact on postpartum depression (PPD). It remains unknown if there is a unique impact of postpartum practices on PPD separately from other types of social support, or if practices differentially affect those with existing prenatal depression. In Pakistan, chilla (چله) is a traditional postpartum practice in which women receive relief from household work, additional familial support, and supplemental food for up to 40 days postpartum. This study aims to understand if chilla protects against PPD independent of other support and whether this relationship varies by prenatal depression status. METHODS: Data come from the Bachpan cohort study in rural Pakistan. Chilla participation and social support (Multidimensional Scale of Perceived Social Support) were assessed at 3 months postpartum. Women were assessed for major depressive episodes (MDE) with the Structured Clinical Interview, DSM-IV and for depression symptom severity with the Patient Health Questionnaire (PHQ-9) in their third trimester and at 6 months postpartum. Adjusted linear mixed models were used to assess the relationship between chilla participation and PPD. RESULTS: Eighty-nine percent of women (N = 786) participated in chilla and almost 70% of those that participated took part in all of chilla's aspects. In adjusted models, chilla participation was inversely related to MDE (OR = 0.56;95%CI = 0.31,1.03) and symptom severity (Mean Difference (MD) = - 1.54;95%CI: - 2.94,-0.14). Chilla participation was associated with lower odds of MDE (OR = 0.44;95%CI = 0.20,0.97) among those not prenatally depressed and with lower symptom severity among those prenatally depressed (MD = -2.05;95%CI:-3.81,-0.49). CONCLUSIONS: Chilla is inversely associated with both MDE and symptom severity at 6 months postpartum above and beyond social support. Specifically, chilla is inversely associated with MDE among those not prenatally depressed and with lower symptom severity among those prenatally depressed. This relationship signals an opportunity for interventions aimed at preventing and treating PPD in this region to draw upon chilla and similar traditional postpartum practices in creating community-based, low-cost, sustainable interventions for maternal mental health. TRIAL REGISTRATION: NCT02111915. Registered 18 September 2015. NCT02658994. Registered 22 January 2016. Both trials were prospectively registered.

xtgeebcv: A command for bias-corrected sandwich variance estimation for GEE analyses of cluster randomized trials.

Cluster randomized trials, where clusters (for example, schools or clinics) are randomized to comparison arms but measurements are taken on individuals, are commonly used to evaluate interventions in public health, education, and the social sciences. Analysis is often conducted on individual-level outcomes, and such analysis methods must consider that outcomes for members of the same cluster tend to be more similar than outcomes for members of other clusters. A popular individual-level analysis technique is generalized estimating equations (GEE). However, it is common to randomize a small number of clusters (for example, 30 or fewer), and in this case, the GEE standard errors obtained from the sandwich variance estimator will be biased, leading to inflated type I errors. Some bias-corrected standard errors have been proposed and studied to account for this finite-sample bias, but none has yet been implemented in Stata. In this article, we describe several popular bias corrections to the robust sandwich variance. We then introduce our newly created command, xtgeebcv, which will allow Stata users to easily apply finite-sample corrections to standard errors obtained from GEE models. We then provide examples to demonstrate the use of xtgeebcv. Finally, we discuss suggestions about which finite-sample corrections to use in which situations and consider areas of future research that may improve xtgeebcv.