Randomized Trial of Early Detection and Treatment of Postpartum Hemorrhage.

BACKGROUND: Delays in the detection or treatment of postpartum hemorrhage can result in complications or death. A blood-collection drape can help provide objective, accurate, and early diagnosis of postpartum hemorrhage, and delayed or inconsistent use of effective interventions may be able to be addressed by a treatment bundle. METHODS: We conducted an international, cluster-randomized trial to assess a multicomponent clinical intervention for postpartum hemorrhage in patients having vaginal delivery. The intervention included a calibrated blood-collection drape for early detection of postpartum hemorrhage and a bundle of first-response treatments (uterine massage, oxytocic drugs, tranexamic acid, intravenous fluids, examination, and escalation), supported by an implementation strategy (intervention group). Hospitals in the control group provided usual care. The primary outcome was a composite of severe postpartum hemorrhage (blood loss, ≥1000 ml), laparotomy for bleeding, or maternal death from bleeding. Key secondary implementation outcomes were the detection of postpartum hemorrhage and adherence to the treatment bundle. RESULTS: A total of 80 secondary-level hospitals across Kenya, Nigeria, South Africa, and Tanzania, in which 210,132 patients underwent vaginal delivery, were randomly assigned to the intervention group or the usual-care group. Among hospitals and patients with data, a primary-outcome event occurred in 1.6% of the patients in the intervention group, as compared with 4.3% of those in the usual-care group (risk ratio, 0.40; 95% confidence interval [CI], 0.32 to 0.50; P<0.001). Postpartum hemorrhage was detected in 93.1% of the patients in the intervention group and in 51.1% of those in the usual-care group (rate ratio, 1.58; 95% CI, 1.41 to 1.76), and the treatment bundle was used in 91.2% and 19.4%, respectively (rate ratio, 4.94; 95% CI, 3.88 to 6.28). CONCLUSIONS: Early detection of postpartum hemorrhage and use of bundled treatment led to a lower risk of the primary outcome, a composite of severe postpartum hemorrhage, laparotomy for bleeding, or death from bleeding, than usual care among patients having vaginal delivery. (Funded by the Bill and Melinda Gates Foundation; E-MOTIVE ClinicalTrials.gov number, NCT04341662.).

Interpretation of statistical findings in randomised trials: a survey of statisticians using thematic analysis of open-ended questions.

BACKGROUND: Dichotomisation of statistical significance, rather than interpretation of effect sizes supported by confidence intervals, is a long-standing problem. METHODS: We distributed an online survey to clinical trial statisticians across the UK, Australia and Canada asking about their experiences, perspectives and practices with respect to interpretation of statistical findings from randomised trials. We report a descriptive analysis of the closed-ended questions and a thematic analysis of the open-ended questions. RESULTS: We obtained 101 responses across a broad range of career stages (24% professors; 51% senior lecturers; 22% junior statisticians) and areas of work (28% early phase trials; 44% drug trials; 38% health service trials). The majority (93%) believed that statistical findings should be interpreted by considering (minimal) clinical importance of treatment effects, but many (61%) said quantifying clinically important effect sizes was difficult, and fewer (54%) followed this approach in practice. Thematic analysis identified several barriers to forming a consensus on the statistical interpretation of the study findings, including: the dynamics within teams, lack of knowledge or difficulties in communicating that knowledge, as well as external pressures. External pressures included the pressure to publish definitive findings and statistical review which can sometimes be unhelpful but can at times be a saving grace. However, the concept of the minimally important difference was identified as a particularly poorly defined, even nebulous, construct which lies at the heart of much disagreement and confusion in the field. CONCLUSION: The majority of participating statisticians believed that it is important to interpret statistical findings based on the clinically important effect size, but report this is difficult to operationalise. Reaching a consensus on the interpretation of a study is a social process involving disparate members of the research team along with editors and reviewers, as well as patients who likely have a role in the elicitation of minimally important differences.

Leveraging baseline covariates to analyze small cluster-randomized trials with a rare binary outcome.

Cluster-randomized trials (CRTs) involve randomizing entire groups of participants-called clusters-to treatment arms but are often comprised of a limited or fixed number of available clusters. While covariate adjustment can account for chance imbalances between treatment arms and increase statistical efficiency in individually randomized trials, analytical methods for individual-level covariate adjustment in small CRTs have received little attention to date. In this paper, we systematically investigate, through extensive simulations, the operating characteristics of propensity score weighting and multivariable regression as two individual-level covariate adjustment strategies for estimating the participant-average causal effect in small CRTs with a rare binary outcome and identify scenarios where each adjustment strategy has a relative efficiency advantage over the other to make practical recommendations. We also examine the finite-sample performance of the bias-corrected sandwich variance estimators associated with propensity score weighting and multivariable regression for quantifying the uncertainty in estimating the participant-average treatment effect. To illustrate the methods for individual-level covariate adjustment, we reanalyze a recent CRT testing a sedation protocol in 31 pediatric intensive care units.

Intensified screening for SARS-CoV-2 in 18 emergency departments in the Paris metropolitan area, France (DEPIST-COVID): A cluster-randomized, two-period, crossover trial.

BACKGROUND: Asymptomatic and paucisymptomatic infections account for a substantial portion of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) transmissions. The value of intensified screening strategies, especially in emergency departments (EDs), in reaching asymptomatic and paucisymptomatic patients and helping to improve detection and reduce transmission has not been documented. The objective of this study was to evaluate in EDs whether an intensified SARS-CoV-2 screening strategy combining nurse-driven screening for asymptomatic/paucisymptomatic patients with routine practice (intervention) could contribute to higher detection of SARS-CoV-2 infections compared to routine practice alone, including screening for symptomatic or hospitalized patients (control). METHODS AND FINDINGS: We conducted a cluster-randomized, two-period, crossover trial from February 2021 to May 2021 in 18 EDs in the Paris metropolitan area, France. All adults visiting the EDs were eligible. At the start of the first period, 18 EDs were randomized to the intervention or control strategy by balanced block randomization with stratification, with the alternative condition being applied in the second period. During the control period, routine screening for SARS-CoV-2 included screening for symptomatic or hospitalized patients. During the intervention period, in addition to routine screening practice, a questionnaire about risk exposure and symptoms and a SARS-CoV-2 screening test were offered by nurses to all remaining asymptomatic/paucisymptomatic patients. The primary outcome was the proportion of newly diagnosed SARS-CoV-2-positive patients among all adults visiting the 18 EDs. Primary analysis was by intention-to-treat. The primary outcome was analyzed using a generalized linear mixed model (Poisson distribution) with the center and center by period as random effects and the strategy (intervention versus control) and period (modeled as a weekly categorical variable) as fixed effects with additional adjustment for community incidence. During the intervention and control periods, 69,248 patients and 69,104 patients, respectively, were included for a total of 138,352 patients. Patients had a median age of 45.0 years [31.0, 63.0], and women represented 45.7% of the patients. During the intervention period, 6,332 asymptomatic/paucisymptomatic patients completed the questionnaire; 4,283 were screened for SARS-CoV-2 by nurses, leading to 224 new SARS-CoV-2 diagnoses. A total of 1,859 patients versus 2,084 patients were newly diagnosed during the intervention and control periods, respectively (adjusted analysis: 26.7/1,000 versus 26.2/1,000, adjusted relative risk: 1.02 (95% confidence interval (CI) [0.94, 1.11]; p = 0.634)). The main limitation of this study is that it was conducted in a rapidly evolving epidemiological context. CONCLUSIONS: The results of this study showed that intensified screening for SARS-CoV-2 in EDs was unlikely to identify a higher proportion of newly diagnosed patients. TRIAL REGISTRATION: Trial registration number: ClinicalTrials.gov NCT04756609.

Permutation-based multiple testing corrections for P $$ P $$ -values and confidence intervals for cluster randomized trials.

In this article, we derive and compare methods to derive P-values and sets of confidence intervals with strong control of the family-wise error rates and coverage for estimates of treatment effects in cluster randomized trials with multiple outcomes. There are few methods for P-value corrections and deriving confidence intervals, limiting their application in this setting. We discuss the methods of Bonferroni, Holm, and Romano and Wolf and adapt them to cluster randomized trial inference using permutation-based methods with different test statistics. We develop a novel search procedure for confidence set limits using permutation tests to produce a set of confidence intervals under each method of correction. We conduct a simulation-based study to compare family-wise error rates, coverage of confidence sets, and the efficiency of each procedure in comparison to no correction using both model-based standard errors and permutation tests. We show that the Romano-Wolf type procedure has nominal error rates and coverage under non-independent correlation structures and is more efficient than the other methods in a simulation-based study. We also compare results from the analysis of a real-world trial.

A re-analysis of 150 women's health trials to investigate how the Bayesian approach may offer a solution to the misinterpretation of statistical findings.

OBJECTIVE: To investigate whether a Bayesian interpretation might help prevent misinterpretation of statistical findings and support authors to differentiate evidence of no effect from statistical uncertainty. DESIGN: A Bayesian re-analysis to determine posterior probabilities of clinically important effects (e.g., a large effect is set at a 4 percentage point difference and a trivial effect to be within a 0.5 percentage point difference). Posterior probabilities greater than 95% are considered as strong statistical evidence, and less than 95% as inconclusive. SAMPLE: 150 major women's health trials with binary outcomes. MAIN OUTCOME MEASURES: Posterior probabilities of large, moderate, small and trivial effects. RESULTS: Under frequentist methods, 48 (32%) were statistically significant (p-value ≤ 0.05) and 102 (68%) statistically non-significant. The frequentist and Bayesian point estimates and confidence intervals showed strong concordance. Of the statistically non-significant trials (n = 102), the Bayesian approach classified the majority (94, 92%) as inconclusive, neither able to confirm or refute effectiveness. A small number of statistically non-significant findings (8, 8%) were classified as having strong statistical evidence of an effect. CONCLUSIONS: Whilst almost all trials report confidence intervals, in practice most statistical findings are interpreted on the basis of statistical significance, mostly concluding evidence of no effect. Findings here suggest the majority are likely uncertain. A Bayesian approach could help differentiate evidence of no effect from statistical uncertainty.

Minimisation for the design of parallel cluster-randomised trials: An evaluation of balance in cluster-level covariates and numbers of clusters allocated to each arm.

BACKGROUND: Cluster-randomised trials often use some form of restricted randomisation, such as stratified- or covariate-constrained randomisation. Minimisation has the potential to balance on more covariates than blocked stratification and can be implemented sequentially unlike covariate-constrained randomisation. Yet, unlike stratification, minimisation has no inbuilt guard to maintain close to a 1:1 allocation. A departure from a 1:1 allocation can be unappealing in a setting with a small number of allocation units such as cluster randomisation which typically include about 30 clusters. METHODS: Using simulation (10,000 per scenario), we evaluate the performance of a range of minimisation procedures on the likelihood of a 1:1 allocation of clusters (10-80 clusters) to treatment arms, along with its performance on covariate imbalance. The range of minimisation procedures includes varying: the proportion of clusters allocated to the least imbalanced arm (known as the stochastic element) - between 0.7 and 1, percentage of first clusters allocated completely at random (known as the bed-in period) - between 0% and 20% and adding 'number of clusters allocated to each arm' as a covariate in the minimisation algorithm. We additionally include a comparison of stratifying and then minimising within key strata (such as country within a multi country cluster trial) as a potential aid to increasing balance. RESULTS: Minimisation is unlikely to result in an exact 1:1 allocation unless the stochastic element is set higher than 0.9. For example, with 20 clusters, 2 binary covariates and setting the stochastic element to 0.7: only 41% of the possible randomisations over the 10,000 simulations achieved a 1:1 allocation. While typical sizes of imbalance were small (a difference of two clusters per arm), allocations as extreme as of 10:10 were observed. Adding the 'number of clusters' into the minimisation algorithm reduces this risk slightly, but covariate imbalance increases slightly. Stratifying and then minimising within key strata improve balance within strata but increase imbalance across all clusters, both on the number of clusters and covariate imbalance. CONCLUSION: In cluster trials, where there are typically about 30 allocation units, when using minimisation, unless the stochastic element is set very high, there is a high risk of not achieving a 1:1 allocation, and a small but nonetheless real risk of an extreme departure from a 1:1 allocation. Stratification with minimisation within key strata (such as country) improves the balance within strata although compromises overall balance.

Rapid intrapartum test for maternal group B streptococcal colonisation and its effect on antibiotic use in labouring women with risk factors for early-onset neonatal infection (GBS2): cluster randomised trial with nested test accuracy study.

BACKGROUND: Mother-to-baby transmission of group B Streptococcus (GBS) is the main cause of early-onset infection. We evaluated whether, in women with clinical risk factors for early neonatal infection, the use of point-of-care rapid intrapartum test to detect maternal GBS colonisation reduces maternal antibiotic exposure compared with usual care, where antibiotics are administered due to those risk factors. We assessed the accuracy of the rapid test in diagnosing maternal GBS colonisation, against the reference standard of selective enrichment culture. METHODS: We undertook a parallel-group cluster randomised trial, with nested test accuracy study and microbiological sub-study. UK maternity units were randomised to a strategy of rapid test (GeneXpert GBS system, Cepheid) or usual care. Within units assigned to rapid testing, vaginal-rectal swabs were taken from women with risk factors for vertical GBS transmission in established term labour. The trial primary outcome was the proportion of women receiving intrapartum antibiotics to prevent neonatal early-onset GBS infection. The accuracy of the rapid test was compared against the standard of selective enrichment culture in diagnosing maternal GBS colonisation. Antibiotic resistance profiles were determined in paired maternal and infant samples. RESULTS: Twenty-two maternity units were randomised and 20 were recruited. A total of 722 mothers (749 babies) participated in rapid test units; 906 mothers (951 babies) were in usual care units. There was no evidence of a difference in the rates of intrapartum antibiotic prophylaxis (relative risk 1.16, 95% CI 0.83 to 1.64) between the rapid test (41%, 297/716) and usual care (36%, 328/906) units. No serious adverse events were reported. The sensitivity and specificity measures of the rapid test were 86% (95% CI 81 to 91%) and 89% (95% CI 85 to 92%), respectively. Babies born to mothers who carried antibiotic-resistant Escherichia coli were more likely to be colonised with antibiotic-resistant strains than those born to mothers with antibiotic-susceptible E. coli. CONCLUSION: The use of intrapartum rapid test to diagnose maternal GBS colonisation did not reduce the rates of antibiotics administered for preventing neonatal early-onset GBS infection than usual care, although with considerable uncertainty. The accuracy of the rapid test is within acceptable limits. TRIAL REGISTRATION: ISRCTN74746075 . Prospectively registered on 16 April 2015.

Mind the gap: covariate constrained randomisation can protect against substantial power loss in parallel cluster randomised trials.

BACKGROUND: Cluster randomised trials often randomise a small number of units, putting them at risk of poor balance of covariates across treatment arms. Covariate constrained randomisation aims to reduce this risk by removing the worst balanced allocations from consideration. This is known to provide only a small gain in power over that averaged under simple randomisation and is likely influenced by the number and prognostic effect of the covariates. We investigated the performance of covariate constrained randomisation in comparison to the worst balanced allocations, and considered the impact on the power of the prognostic effect and number of covariates adjusted for in the analysis. METHODS: Using simulation, we examined the Monte Carlo type I error rate and power of cross-sectional, two-arm parallel cluster-randomised trials with a continuous outcome and four binary cluster-level covariates, using either simple or covariate constrained randomisation. Data were analysed using a small sample corrected linear mixed-effects model, adjusted for some or all of the binary covariates. We varied the number of clusters, intra-cluster correlation, number and prognostic effect of covariates balanced in the randomisation and adjusted in the analysis, and the size of the candidate set from which the allocation was selected. For each scenario, 20,000 simulations were conducted. RESULTS: When compared to the worst balanced allocations, covariate constrained randomisation with an adjusted analysis provided gains in power of up to 20 percentage points. Even with analysis-based adjustment for those covariates balanced in the randomisation, the type I error rate was not maintained when the intracluster correlation is very small (0.001). Generally, greater power was achieved when more prognostic covariates are restricted in the randomisation and as the size of the candidate set decreases. However, adjustment for weakly prognostic covariates lead to a loss in power of up to 20 percentage points. CONCLUSIONS: When compared to the worst balanced allocations, covariate constrained randomisation provides moderate to substantial improvements in power. However, the prognostic effect of the covariates should be carefully considered when selecting them for inclusion in the randomisation.

Bayesian analysis of a systematic review of early versus late tracheostomy in ICU patients.

BACKGROUND: A recent systematic review and meta-analysis of RCTs of early vs late tracheostomy in mechanically ventilated patients suggest that early tracheostomy reduces the duration of ICU stay and mechanical ventilation, but does not reduce short-term mortality or ventilator-associated pneumonia (VAP). Meta-analysis of randomised trials is typically performed using a frequentist approach, and although reporting confidence intervals, interpretation is usually based on statistical significance. To provide a robust basis for clinical decision-making, we completed the search used from the previous review and analysed the data using Bayesian methods to estimate posterior probabilities of the effect of early tracheostomy on clinical outcomes. METHODS: The search was completed for RCTS comparing early vs late tracheostomy in the databases PubMed, EMBASE, and Cochrane library in June 2022. Effect estimates and 95% confidence intervals were calculated for the outcomes short-term mortality, VAP, duration of ICU stay, and mechanical ventilation. A Bayesian meta-analysis was performed with uninformative priors. Risk ratios (RRs) and standardised mean differences (SMDs) with 95% credible intervals were reported alongside posterior probabilities for any benefit (RR<1; SMD<0), a small benefit (number needed to treat, 200; SMD<-0.5), or modest benefit (number needed to treat, 100; SMD<-1). RESULTS: Nineteen RCTs with 3508 patients were included. Comparing patients with early vs late tracheostomy, the posterior probabilities for any benefit, small benefit, and modest benefit, respectively, were: 99%, 99%, and 99% for short-term mortality; 94%, 78%, and 51% for VAP; 97%, 43%, and 1% for duration of mechanical ventilation; and 97%, 75%, and 27% and for length of ICU stay. CONCLUSIONS: Bayesian meta-analysis suggests a high probability that early tracheostomy compared with delayed tracheostomy has at least some benefit across all clinical outcomes considered.

Effects on childhood infections of promoting safe and hygienic complementary-food handling practices through a community-based programme: A cluster randomised controlled trial in a rural area of The Gambia.

BACKGROUND: The Gambia has high rates of under-5 mortality from diarrhoea and pneumonia, peaking during complementary-feeding age. Community-based interventions may reduce complementary-food contamination and disease rates. METHODS AND FINDINGS: A public health intervention using critical control points and motivational drivers, delivered February-April 2015 in The Gambia, was evaluated in a cluster randomised controlled trial at 6- and 32-month follow-up in September-October 2015 and October-December 2017, respectively. After consent for trial participation and baseline data were collected, 30 villages (clusters) were randomly assigned to intervention or control, stratified by population size and geography. The intervention included a community-wide campaign on days 1, 2, 17, and 25, a reminder visit at 5 months, plus informal community-volunteer home visits. It promoted 5 key complementary-food and 1 key drinking-water safety and hygiene behaviours through performing arts, public meetings, and certifications delivered by a team from local health and village structures to all villagers who attended the activities, to which mothers of 6- to 24-month-old children were specifically invited. Control villages received a 1-day campaign on domestic-garden water use. The background characteristics of mother and clusters (villages) were balanced between the trial arms. Outcomes were measured at 6 and 32 months in a random sample of 21-26 mothers per cluster. There were no intervention or research team visits to villages between 6 and 32 months. The primary outcome was a composite outcome of the number of times key complementary-food behaviours were observed as a proportion of the number of opportunities to perform the behaviours during the observation period at 6 months. Secondary outcomes included the rate of each recommended behaviour; microbiological growth from complementary food and drinking water (6 months only); and reported acute respiratory infections, diarrhoea, and diarrhoea hospitalisation. Analysis was by intention-to-treat analysis adjusted by clustering. (Registration: PACTR201410000859336). We found that 394/571 (69%) of mothers with complementary-feeding children in the intervention villages were actively involved in the campaign. No villages withdrew, and there were no changes in the implementation of the intervention. The intervention improved behaviour adoption significantly. For the primary outcome, the rate was 662/4,351(incidence rate [IR] = 0.15) in control villages versus 2,861/4,378 (IR = 0.65) in intervention villages (adjusted incidence rate ratio [aIRR] = 4.44, 95% CI 3.62-5.44, p < 0.001), and at 32 months the aIRR was 1.17 (95% CI 1.07-1.29, p = 0.001). Secondary health outcomes also improved with the intervention: (1) mother-reported diarrhoea at 6 months, with adjusted relative risk (aRR) = 0.39 (95% CI 0.32-0.48, p < 0.001), and at 32 months, with aRR = 0.68 (95% CI 0.48-0.96, p = 0.027); (2) mother-reported diarrhoea hospitalisation at 6 months, with aRR = 0.35 (95% CI 0.19-0.66, p = 0.001), and at 32 months, with aRR = 0.38 (95% CI 0.18-0.80, p = 0.011); and (3) mother-reported acute respiratory tract infections at 6 months, with aRR = 0.67 (95% CI 0.53-0.86, p = 0.001), though at 32 months improvement was not significant (p = 0.200). No adverse events were reported. The main limitations were that only medium to small rural villages were involved. Obtaining laboratory cultures from food at 32 months was not possible, and no stool microorganisms were investigated. CONCLUSIONS: We found that low-cost and culturally embedded behaviour change interventions were acceptable to communities and led to short- and long-term improvements in complementary-food safety and hygiene practices, and reported diarrhoea and acute respiratory tract infections. TRIAL REGISTRATION: The trial was registered on the 17th October 2014 with the Pan African Clinical Trial Registry in South Africa with number (PACTR201410000859336) and 32-month follow-up as an amendment to the trial.

The evidence for cognitive behavioural therapy in any condition, population or context: a meta-review of systematic reviews and panoramic meta-analysis.

The majority of psychological treatment research is dedicated to investigating the effectiveness of cognitive behavioural therapy (CBT) across different conditions, population and contexts. We aimed to summarise the current systematic review evidence and evaluate the consistency of CBT's effect across different conditions. We included reviews of CBT randomised controlled trials in any: population, condition, format, context, with any type of comparator and published in English. We searched DARE, Cochrane, MEDLINE, EMBASE, PsycINFO, CINAHL, CDAS, and OpenGrey between 1992 and January 2019. Reviews were quality assessed, their data extracted and summarised. The effects upon health-related quality of life (HRQoL) were pooled, within-condition groups. If the across-condition heterogeneity was I2 < 75%, we pooled effects using a random-effect panoramic meta-analysis. We summarised 494 reviews (221 128 participants), representing 14/20 physical and 13/20 mental conditions (World Health Organisation's International Classification of Diseases). Most reviews were lower-quality (351/494), investigated face-to-face CBT (397/494), and in adults (378/494). Few reviews included trials conducted in Asia, South America or Africa (45/494). CBT produced a modest benefit across-conditions on HRQoL (standardised mean difference 0.23; 95% confidence intervals 0.14-0.33, I2 = 32%). The effect's associated prediction interval -0.05 to 0.50 suggested CBT will remain effective in conditions for which we do not currently have available evidence. While there remain some gaps in the completeness of the evidence base, we need to recognise the consistent evidence for the general benefit which CBT offers.

Contamination: How much can an individually randomized trial tolerate?

Cluster randomization results in an increase in sample size compared to individual randomization, referred to as an efficiency loss. This efficiency loss is typically presented under an assumption of no contamination in the individually randomized trial. An alternative comparator is the sample size needed under individual randomization to detect the attenuated treatment effect due to contamination. A general framework is provided for determining the extent of contamination that can be tolerated in an individually randomized trial before a cluster randomized design yields a larger sample size. Results are presented for a variety of cluster trial designs including parallel arm, stepped-wedge and cluster crossover trials. Results reinforce what is expected: individually randomized trials can tolerate a surprisingly large amount of contamination before they become less efficient than cluster designs. We determine the point at which the contamination means an individual randomized design to detect an attenuated effect requires a larger sample size than cluster randomization under a nonattenuated effect. This critical rate is a simple function of the design effect for clustering and the design effect for multiple periods as well as design effects for stratification or repeated measures under individual randomization. These findings are important for pragmatic comparisons between a novel treatment and usual care as any bias due to contamination will only attenuate the true treatment effect. This is a bias that operates in a predictable direction. Yet, cluster randomized designs with post-randomization recruitment without blinding, are at high risk of bias due to the differential recruitment across treatment arms. This sort of bias operates in an unpredictable direction. Thus, with knowledge that cluster randomized trials are generally at a greater risk of biases that can operate in a nonpredictable direction, results presented here suggest that even in situations where there is a risk of contamination, individual randomization might still be the design of choice.

Design and analysis of three-arm parallel cluster randomized trials with small numbers of clusters.

In this article, we review and evaluate a number of methods used in the design and analysis of small three-arm parallel cluster randomized trials. We conduct a simulation-based study to evaluate restricted randomization methods including covariate-constrained randomization and a novel method for matched-group cluster randomization. We also evaluate the appropriate modelling of the data and small sample inferential methods for a variety of treatment effects relevant to three-arm trials. Our results indicate that small-sample corrections are required for high (0.05) but not low (0.001) values of the intraclass correlation coefficient and their performance can depend on trial design, number of clusters, and the nature of the hypothesis being tested. The Satterthwaite correction generally performed best at an ICC of 0.05 with a nominal type I error rate for single-period trials, and in trials with repeated measures type I error rates were between 0.04 and 0.06. Restricted randomization methods produce little benefit in trials with repeated measures but in trials with single post-intervention design can provide relatively large gains in power when compared to the most unbalanced possible allocations. Matched-group randomization improves power but is not as effective as covariate-constrained randomization. For model-based analysis, adjusting for fewer covariates than were used in a restricted randomization process under any design can produce non-nominal type I error rates and reductions in power. Where comparisons to two-arm cluster trials are possible, the performance of the methods is qualitatively very similar.

Reducing Medical Admissions and Presentations Into Hospital through Optimising Medicines (REMAIN HOME): a stepped wedge, cluster randomised controlled trial.

OBJECTIVE: To investigate whether integrating pharmacists into general practices reduces the number of unplanned re-admissions of patients recently discharged from hospital. DESIGN, SETTING: Stepped wedge, cluster randomised trial in 14 general practices in southeast Queensland. PARTICIPANTS: Adults discharged from one of seven study hospitals during the seven days preceding recruitment (22 May 2017 - 14 March 2018) and prescribed five or more long term medicines, or having a primary discharge diagnosis of congestive heart failure or exacerbation of chronic obstructive pulmonary disease. INTERVENTION: Comprehensive face-to-face medicine management consultation with an integrated practice pharmacist within seven days of discharge, followed by a consultation with their general practitioner and further pharmacist consultations as needed. MAJOR OUTCOMES: Rates of unplanned, all-cause hospital re-admissions and emergency department (ED) presentations 12 months after hospital discharge; incremental net difference in overall costs. RESULTS: By 12 months, there had been 282 re-admissions among 177 control patients (incidence rate [IR], 1.65 per person-year) and 136 among 129 intervention patients (IR, 1.09 per person-year; fully adjusted IR ratio [IRR], 0.79; 95% CI, 0.52-1.18). ED presentation incidence (fully adjusted IRR, 0.46; 95% CI, 0.22-0.94) and combined re-admission and ED presentation incidence (fully adjusted IRR, 0.69; 95% CI, 0.48-0.99) were significantly lower for intervention patients. The estimated incremental net cost benefit of the intervention was $5072 per patient, with a benefit-cost ratio of 31:1. CONCLUSION: A collaborative pharmacist-GP model of post-hospital discharge medicines management can reduce the incidence of hospital re-admissions and ED presentations, achieving substantial cost savings to the health system. TRIAL REGISTRATION: Australian New Zealand Clinical Trials Registry, ACTRN12616001627448 (prospective).

Reopening schools safely in the face of COVID-19: Can cluster randomized trials help?

The COVID-19 pandemic has highlighted the challenges of evidence-based health policymaking, as critical precautionary decisions, such as school closures, had to be made urgently on the basis of little evidence. As primary and secondary schools once again close in the face of surging infections, there is an opportunity to rigorously study their reopening. School-aged children appear to be less affected by COVID-19 than adults, yet schools may drive community transmission of the virus. Given the impact of school closures on both education and the economy, schools cannot remain closed indefinitely. But when and how can they be reopened safely? We argue that a cluster randomized trial is a rigorous and ethical way to resolve these uncertainties. We discuss key scientific, ethical, and resource considerations both to inform trial design of school reopenings and to prompt discussion of the merits and feasibility of conducting such a trial.

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.

An opportunistic evaluation of a routine service improvement project to reduce falls in hospital.

BACKGROUND: Preventing falls in hospital is a perennial patient safety issue. The University Hospital Coventry and Warwickshire initiated a programme to train ward staff in accordance with guidelines. The National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care West Midlands was asked to expedite an independent evaluation of the initiative. We set out to describe the intervention to implement the guidelines and to evaluate it by means of a step-wedge cluster study using routinely collected data. METHODS: The evaluation was set up as a partially randomised, step-wedge cluster study, but roll-out across wards was more rapid than planned. The study was therefore analysed as a time-series. Primary outcome was rate of falls per 1000 Occupied Bed Days (OBDs) collected monthly using routine data. Data was analysed using a mixed-effects Poisson regression model, with a fixed effect for intervention, time and post-intervention time. We allowed for random variations across clusters in initial fall rate, pre-intervention slope and post-intervention slope. RESULTS: There was an average of 6.62 falls per 1000 OBDs in the control phase, decreasing to an average of 5.89 falls per 1000 OBDs in the period after implementation to the study end. Regression models showed no significant step change in fall rates (IRR: 1.02, 95% CI: 0.92-1.14). However, there was a gradual decrease, of approximately 3%, after the intervention was introduced (IRR: 0.97 per month, 95% CI: 0.95-0.99). CONCLUSION: The intervention was associated with a small but statistically significantly improvement in falls rates. Expedited roll-out of an intervention may vitiate a step-wedge cluster design, but the intervention can still be studied using a time-series analysis. Assuming that there is some value in time series analyses, this is better than no evaluation at all. However, care is needed in making causal inferences given the non-experimental nature of the design.

Effect of a Sedation and Ventilator Liberation Protocol vs Usual Care on Duration of Invasive Mechanical Ventilation in Pediatric Intensive Care Units: A Randomized Clinical Trial.

Importance: There is limited evidence on the optimal strategy for liberating infants and children from invasive mechanical ventilation in the pediatric intensive care unit. Objective: To determine if a sedation and ventilator liberation protocol intervention reduces the duration of invasive mechanical ventilation in infants and children anticipated to require prolonged mechanical ventilation. Design, Setting, and Participants: A pragmatic multicenter, stepped-wedge, cluster randomized clinical trial was conducted that included 17 hospital sites (18 pediatric intensive care units) in the UK sequentially randomized from usual care to the protocol intervention. From February 2018 to October 2019, 8843 critically ill infants and children anticipated to require prolonged mechanical ventilation were recruited. The last date of follow-up was November 11, 2019. Interventions: Pediatric intensive care units provided usual care (n = 4155 infants and children) or a sedation and ventilator liberation protocol intervention (n = 4688 infants and children) that consisted of assessment of sedation level, daily screening for readiness to undertake a spontaneous breathing trial, a spontaneous breathing trial to test ventilator liberation potential, and daily rounds to review sedation and readiness screening and set patient-relevant targets. Main Outcomes and Measures: The primary outcome was the duration of invasive mechanical ventilation from initiation of ventilation until the first successful extubation. The primary estimate of the treatment effect was a hazard ratio (with a 95% CI) adjusted for calendar time and cluster (hospital site) for infants and children anticipated to require prolonged mechanical ventilation. Results: There were a total of 8843 infants and children (median age, 8 months [interquartile range, 1 to 46 months]; 42% were female) who completed the trial. There was a significantly shorter median time to successful extubation for the protocol intervention compared with usual care (64.8 hours vs 66.2 hours, respectively; adjusted median difference, -6.1 hours [interquartile range, -8.2 to -5.3 hours]; adjusted hazard ratio, 1.11 [95% CI, 1.02 to 1.20], P = .02). The serious adverse event of hypoxia occurred in 9 (0.2%) infants and children for the protocol intervention vs 11 (0.3%) for usual care; nonvascular device dislodgement occurred in 2 (0.04%) vs 7 (0.1%), respectively. Conclusions and Relevance: Among infants and children anticipated to require prolonged mechanical ventilation, a sedation and ventilator liberation protocol intervention compared with usual care resulted in a statistically significant reduction in time to first successful extubation. However, the clinical importance of the effect size is uncertain. Trial Registration: isrctn.org Identifier: ISRCTN16998143.

Effectiveness of polypill for primary and secondary prevention of cardiovascular diseases (PolyIran): a pragmatic, cluster-randomised trial.

BACKGROUND: A fixed-dose combination therapy (polypill strategy) has been proposed as an approach to reduce the burden of cardiovascular disease, especially in low-income and middle-income countries (LMICs). The PolyIran study aimed to assess the effectiveness and safety of a four-component polypill including aspirin, atorvastatin, hydrochlorothiazide, and either enalapril or valsartan for primary and secondary prevention of cardiovascular disease. METHODS: The PolyIran study was a two-group, pragmatic, cluster-randomised trial nested within the Golestan Cohort Study (GCS), a cohort study with 50 045 participants aged 40-75 years from the Golestan province in Iran. Clusters (villages) were randomly allocated (1:1) to either a package of non-pharmacological preventive interventions alone (minimal care group) or together with a once-daily polypill tablet (polypill group). Randomisation was stratified by three districts (Gonbad, Aq-Qala, and Kalaleh), with the village as the unit of randomisation. We used a balanced randomisation algorithm, considering block sizes of 20 and balancing for cluster size or natural log of the cluster size (depending on the skewness within strata). Randomisation was done at a fixed point in time (Jan 18, 2011) by statisticians at the University of Birmingham (Birmingham, UK), independent of the local study team. The non-pharmacological preventive interventions (including educational training about healthy lifestyle-eg, healthy diet with low salt, sugar, and fat content, exercise, weight control, and abstinence from smoking and opium) were delivered by the PolyIran field visit team at months 3 and 6, and then every 6 months thereafter. Two formulations of polypill tablet were used in this study. Participants were first prescribed polypill one (hydrochlorothiazide 12·5 mg, aspirin 81 mg, atorvastatin 20 mg, and enalapril 5 mg). Participants who developed cough during follow-up were switched by a trained study physician to polypill two, which included valsartan 40 mg instead of enalapril 5 mg. Participants were followed up for 60 months. The primary outcome-occurrence of major cardiovascular events (including hospitalisation for acute coronary syndrome, fatal myocardial infarction, sudden death, heart failure, coronary artery revascularisation procedures, and non-fatal and fatal stroke)-was centrally assessed by the GCS follow-up team, who were masked to allocation status. We did intention-to-treat analyses by including all participants who met eligibility criteria in the two study groups. The trial was registered with ClinicalTrials.gov, number NCT01271985. FINDINGS: Between Feb 22, 2011, and April 15, 2013, we enrolled 6838 individuals into the study-3417 (in 116 clusters) in the minimal care group and 3421 (in 120 clusters) in the polypill group. 1761 (51·5%) of 3421 participants in the polypill group were women, as were 1679 (49·1%) of 3417 participants in the minimal care group. Median adherence to polypill tablets was 80·5% (IQR 48·5-92·2). During follow-up, 301 (8·8%) of 3417 participants in the minimal care group had major cardiovascular events compared with 202 (5·9%) of 3421 participants in the polypill group (adjusted hazard ratio [HR] 0·66, 95% CI 0·55-0·80). We found no statistically significant interaction with the presence (HR 0·61, 95% CI 0·49-0·75) or absence of pre-existing cardiovascular disease (0·80; 0·51-1·12; pinteraction=0·19). When restricted to participants in the polypill group with high adherence, the reduction in the risk of major cardiovascular events was even greater compared with the minimal care group (adjusted HR 0·43, 95% CI 0·33-0·55). The frequency of adverse events was similar between the two study groups. 21 intracranial haemorrhages were reported during the 5 years of follow-up-ten participants in the polypill group and 11 participants in the minimal care group. There were 13 physician-confirmed diagnoses of upper gastrointestinal bleeding in the polypill group and nine in the minimal care group. INTERPRETATION: Use of polypill was effective in preventing major cardiovascular events. Medication adherence was high and adverse event numbers were low. The polypill strategy could be considered as an additional effective component in controlling cardiovascular diseases, especially in LMICs. FUNDING: Tehran University of Medical Sciences, Barakat Foundation, and Alborz Darou.