Interim data monitoring in cluster randomised trials: Practical issues and a case study.

BACKGROUND: There is an abundance of guidance for the interim monitoring of individually randomised trials. While methodological literature exists on how to extend these methods to cluster randomised trials, there is little guidance on practical implementation. Cluster trials have many features which make their monitoring needs different. We outline the methodological and practical challenges of interim monitoring of cluster trials; and apply these considerations to a case study. CASE STUDY: The E-MOTIVE study is an 80-cluster randomised trial of a bundle of interventions to treat postpartum haemorrhage. The proposed data monitoring plan includes (1) monitor sample size assumptions, (2) monitor for evidence of selection bias, and (3) an interim assessment of the primary outcome, as well as monitoring data completeness. The timing of the sample size monitoring is chosen with both consideration of statistical precision and to allow time to recruit more clusters. Monitoring for selection bias involves comparing individual-level characteristics and numbers recruited between study arms to identify any post-randomisation participant identification bias. An interim analysis of outcomes presented with 99.9% confidence intervals using the Haybittle-Peto approach should mitigate any concern regarding the inflation of type-I error. The pragmatic nature of the trial means monitoring for adherence is not relevant, as it is built into a process evaluation. CONCLUSIONS: The interim analyses of cluster trials have a number of important differences to monitoring individually randomised trials. In cluster trials, there will often be a greater need to monitor nuisance parameters, yet there will often be considerable uncertainty in their estimation. This means the utility of sample size re-estimation can be questionable particularly when there are practical or funding difficulties associated with making any changes to planned sample sizes. Perhaps most importantly interim monitoring has the potential to identify selection bias, particularly in trials with post-randomisation identification or recruitment. Finally, the pragmatic nature of cluster trials might mean that the utility of methods to allow for interim monitoring of outcomes based on statistical testing, or monitoring for adherence to study interventions, are less relevant. Our intention is to facilitate the planning of future cluster randomised trials and to promote discussion and debate to improve monitoring of these studies.

The accuracy of cell-free fetal DNA-based non-invasive prenatal testing in singleton pregnancies: a systematic review and bivariate meta-analysis.

BACKGROUND: Cell-free fetal DNA (cffDNA) non-invasive prenatal testing (NIPT) is rapidly expanding, and is being introduced at varying rates depending on country and condition. OBJECTIVES: Determine accuracy of cffDNA-based NIPT for all conditions. Evaluate influence of other factors on test performance. SEARCH STRATEGY: Medline, Embase, CINAHL, Cochrane Library, from 1997 to April 2015. SELECTION CRITERIA: Cohort studies reporting cffDNA-based NIPT performance in singleton pregnancies. DATA COLLECTION AND ANALYSIS: Bivariate or univariate meta-analysis and subgroup analysis performed to explore influence of test type and population risk. MAIN RESULTS: A total of 117 studies were included that analysed 18 conditions. Bivariate meta-analysis demonstrated sensitivities and specificities, respectively, for: fetal sex, 0.989 (95% CI 0.980-0.994) and 0.996 (95% CI 0.989-0.998), 11 179 tests; rhesus D, 0.993 (95% CI 0.982-0.997) and 0.984 (95% CI 0.964-0.993), 10 290 tests; trisomy 21, 0.994 (95% CI 0.983-0.998) and 0.999 (95% CI 0.999-1.000), 148 344 tests; trisomy 18, 0.977 (95% CI 0.952-0.989) and 0.999 (95% CI 0.998-1.000), 146 940 tests; monosomy X, 0.929 (95% CI 0.741-0.984) and 0.999 (95% CI 0.995-0.999), 6712 tests. Trisomy 13 was analysed by univariate meta-analysis, with a summary sensitivity of 0.906 (95% CI 0.823-0.958) and specificity of 1.00 (95% CI 0.999-0.100), from 134 691 tests. False and inconclusive results were poorly reported across all conditions. Although the test type affected both sensitivity and specificity, there was no evidence that population risk had any effect. CONCLUSION: Performance of cffDNA-based NIPT is affected by condition under investigation. For fetal sex and rhesus D status, NIPT can be considered diagnostic. For trisomy 21, 18, and 13, the lower sensitivity, specificity, and disease prevalence, combined with the biological influence of confined placental mosaicism, designates it a screening test. These factors must be considered when counselling patients and assessing the cost of introduction into routine care. TWEETABLE ABSTRACT: cffDNA NIPT accuracy high, can be diagnostic for fetal sex and rhesus D, but only screening test in aneuploidy.

Synthesis, conformation and antiproliferative activity of isothiazoloisoxazole 1,1-dioxides.

Sixteen new isothiazoloisoxazole 1,1-dioxides, one new isothiazolotriazole and one new isothiazolopyrazole have been synthesised by using 1,3-dipolar cycloadditions to isothiazole 1,1-dioxides. One sub-set of these isothiazoloisoxazoles showed low µM activity against a human breast carcinoma cell line, whilst a second sub-set plus the isothiazolotriazole demonstrated an interesting restricted rotation of sterically hindered bridgehead substituents. A thiazete 1,1-dioxide produced from one of the isothiazole 1,1-dioxides underwent conversion into an unknown 1,2,3-oxathiazolin-2-oxide upon treatment with Lewis acids, but was inert towards 1,3-dipoles and cyclopropenones. Six supporting crystal structures are included.

The Impact of a 24-h Ultra-Marathon on Circulatory Endotoxin and Cytokine Profile.

The study aimed to determine circulatory endotoxin concentration, cytokine profile, and gastrointestinal symptoms of ultra-endurance runners (UER, n=17) in response to a 24-h continuous ultra-marathon competition (total distance range: 122-208 km) conducted in temperate ambient conditions (0-20 °C) in mountainous terrain. Body mass and body temperature were measured, and venous blood samples were taken before and immediately after competition. Samples were analysed for gram-negative bacterial endotoxin, C-reactive protein, cytokine profile, and plasma osmolality. Gastrointestinal symptoms were also monitored throughout competition. Mean exercise-induced body mass loss was (mean±SD) 1.7±1.8% in UER. Pre- and post-competition plasma osmolality in UER was 286±11 mOsmol·kg(-1) and 286±9 mOsmol·kg(-1), respectively. Pre- to post-competition increases (p<0.05) were observed for endotoxin (37%), C-reactive protein (2832%), IL-6 (3 436%), IL-1ß (332%), TNF-α (35%), IL-10 (511%), and IL-8 (239%) concentrations in UER, with no change in the control group (CON; n=12) observed (p>0.05). Gastrointestinal symptoms were reported by 75% of UER, with no symptoms reported by CON. IL-10 (r=0.535) and IL-8 (r=0.503) were positively correlated with gastrointestinal symptoms. A 24-h continuous ultra-marathon competition in temperate ambient conditions resulted in a circulatory endotoxaemia and pro-inflammatory cytokinaemia, counteracted by a compensatory anti-inflammatory response.

ß-Adrenergic receptor antagonists and fracture risk: a meta-analysis of selectivity, gender, and site-specific effects.

UNLABELLED: By meta-analysis, the risk of fracture was 15% lower in patients treated with ß-adrenergic blockers compared to controls independent of gender, fracture site, and dose. This might be attributable to ß1-selective blockers. INTRODUCTION: The aim of this study is to determine by meta-analysis whether ß-adrenergic blockers (BBs) reduce fracture risk and whether the effect, if demonstrable, is dependent upon selectivity, dose, gender, or fracture site. METHODS: A literature search was performed in electronic databases MEDLINE, EMBASE, and reference sections of relevant articles to identify eligible studies. Adjusted estimates of fracture risk effect size (ES) were pooled across studies using fixed or random-effects (RE) meta-analysis as appropriate. Dose-related effects were evaluated using meta-regression. To explore the relative efficacy of ß1-selective blockers in comparison to nonselective BBs, adjusted indirect comparison was performed. RESULTS: A total of 16 studies (7 cohort and 9 case-control studies), involving 1,644,570 subjects, were identified. The risk of any fracture was found to be significantly reduced in subjects receiving BBs as compared to control subjects (16 studies, RE pooled ES = 0.86, 95% CI 0.78-0.93; I(2) = 87 %). In a sensitivity analysis limited to those studies deemed to be most robust, the BB effect to reduce fracture risk was sustained (four studies, pooled ES = 0.79, 95% CI 0.67-0.94; I(2) = 96%). The risk of a hip fracture was lower in both women and men receiving BBs (women: pooled ES = 0.86, 95% CI 0.80-0.91; I(2) = 1% and men: pooled ES = 0.80, 95% CI 0.71-0.90; I(2) = 0%). Similar risk reductions were found for clinical vertebral and forearm fractures, although statistical significance was not reached. The reduction in risk did not appear to be dose-related (test for a linear trend p value 0.150). Using adjusted indirect comparisons, it was estimated that ß1-selective agents were significantly more effective than nonselective BBs in reducing the risk of any fracture (six studies, ß1-selective blockers vs. nonselective BBs: RE pooled ES = 0.82, 95% CI = 0.69-0.97). CONCLUSIONS: The findings suggest that the risk of fracture is approximately 15% lower in patients treated with BBs compared to controls independent of gender, fracture site, and dose. This risk reduction might be associated with the effects of ß1-selective blockers.

The effectiveness of service delivery initiatives at improving patients' waiting times in clinical radiology departments: a systematic review.

We reviewed the literature for the impact of service delivery initiatives (SDIs) on patients' waiting times within radiology departments. We searched MEDLINE, EMBASE, CINAHL, INSPEC and The Cochrane Library for relevant articles published between 1995 and February, 2013. The Cochrane EPOC risk of bias tool was used to assess the risk of bias on studies that met specified design criteria. Fifty-seven studies met the inclusion criteria. The types of SDI implemented included extended scope practice (ESP, three studies), quality management (12 studies), productivity-enhancing technologies (PETs, 29 studies), multiple interventions (11 studies), outsourcing and pay-for-performance (one study each). The uncontrolled pre- and post-intervention and the post-intervention designs were used in 54 (95%) of the studies. The reporting quality was poor: many of the studies did not test and/or report the statistical significance of their results. The studies were highly heterogeneous, therefore meta-analysis was inappropriate. The following type of SDIs showed promising results: extended scope practice; quality management methodologies including Six Sigma, Lean methodology, and continuous quality improvement; productivity-enhancing technologies including speech recognition reporting, teleradiology and computerised physician order entry systems. We have suggested improved study design and the mapping of the definitions of patient waiting times in radiology to generic timelines as a starting point for moving towards a situation where it becomes less restrictive to compare and/or pool the results of future studies in a meta-analysis.

Sample size determination for external pilot cluster randomised trials with binary feasibility outcomes: a tutorial.

Justifying sample size for a pilot trial is a reporting requirement, but few pilot trials report a clear rationale for their chosen sample size. Unlike full-scale trials, pilot trials should not be designed to test effectiveness, and so, conventional sample size justification approaches do not apply. Rather, pilot trials typically specify a range of primary and secondary feasibility objectives. Often, these objectives relate to estimation of parameters that inform the sample size justification for the full-scale trial, many of which are binary. These binary outcomes are referred to as "feasibility outcomes" and include expected prevalence of the primary trial outcome, primary outcome availability, or recruitment or retention proportions.For pilot cluster trials, sample size calculations depend on the number of clusters, the cluster sizes, the anticipated intra-cluster correlation coefficient for the feasibility outcome and the anticipated proportion for that outcome. Of key importance is the intra-cluster correlation coefficient for the feasibility outcome. It has been suggested that correlations for feasibility outcomes are larger than for clinical outcomes measuring effectiveness. Yet, there is a dearth of information on realised values for these correlations.In this tutorial, we demonstrate how to justify sample size in external pilot cluster trials where the objective is to estimate a binary feasibility outcome. We provide sample size calculation formulae for a variety of scenarios, make available an R Shiny app for implementation, and compile a report of intra-cluster correlations for feasibility outcomes from a convenience sample. We demonstrate that unless correlations are very low, external pilot cluster trials can be made more efficient by including more clusters and fewer observations per cluster.

Inpatient electronic prescribing data can be used to identify 'lost' discharge codes for diabetes.

AIM: Accurate assessment of missed discharge codes for diabetes is critical for effective planning of hospital diabetes services. We wished to estimate the frequency of missed discharge diagnostic codes for diabetes and the impact missed codes would have on diabetes-related payments to the hospital. METHODS: We linked Patient Administration System data to the Prescribing Information and Communication System. We defined diabetes as those having a discharge code for diabetes in the Patient Administration System and those on anti-diabetic medication in the Prescribing Information and Communication System. Based on the two sources, we calculated the estimated missed discharge codes for diabetes using the capture-recapture technique. We generated the Healthcare Resource Group for a given admission before and after correction for the missed code to estimate the impact that correction would make on payments to the hospital. RESULTS: Among the 171 067 admissions linked, 22 412 (13.1%) had a code for diabetes at discharge. An additional 2706 admissions were classified as having diabetes based on prescription data. The capture-recapture technique estimated there were 4588 (2.7% of all admissions) admissions with diabetes missed by current coding, of which 2706 (60%) would be obtained from prescription data. After adding a diabetes diagnostic code, 12.8% of the missed admissions with diabetes resulted in a change to the Healthcare Resource Group tariff code and payment. CONCLUSION: The use of electronic prescription data is a simple solution to correct for missed discharge diagnostic codes.

Hypoglycaemia is associated with increased length of stay and mortality in people with diabetes who are hospitalized.

AIM: To study the length of stay and inpatient mortality of patients with diabetes who had an episode of hypoglycaemia in a non critical care setting at University Hospital Birmingham, UK. METHODS: Retrospective analysis of routinely available electronic data of 6374 admissions with a recording of either laboratory or point-of-care blood glucose value. Based on the lowest recorded blood glucose values, patients were categorized into a group without hypoglycaemia (> 3.9 mmol/l), a group with mild to moderate hypoglycaemia (2.3-3.9 mmol/l) and a group with severe hypoglycaemic (≤ 2.2 mmol/l). Length of stay and inpatient mortality were compared between the three groups, adjusting for age, gender, ethnicity, deprivation, admission type, use of insulin and modified Charlson co-morbidity score. RESULTS: There were 148 admissions (2.3%) with severe hypoglycaemia (≤ 2.2 mmol/l), 500 admissions (7.8%) with mild to moderate hypoglycaemia (2.2-3.9 mmol/l) and 5726 admissions with no recorded hypoglycaemic episode (> 3.9 mmol/l). After adjustment, length of stay, when compared with those without a recorded hypoglycaemic episode, was 1.51 (95% CI 1.35-1.68) times higher in the group with blood glucose values of 2.3-3.9 mmol/l and 2.33 (95% CI 1.91-2.84) higher in the group with blood glucose values ≤ 2.2 mmol/l. Adjusted odds ratio of inpatient mortality when compared with the group without hypoglycaemia was 1.62 (95% CI 1.16-2.27) in the group with blood glucose values of 2.3-3.9 mmol/l and 2.05 (95% CI 1.24-3.38) in the group with blood glucose values ≤ 2.2 mmol/l. CONCLUSION: Hypoglycaemia is associated with increased length of stay and inpatient mortality. Whilst causative evidence is lacking, our data are consistent with the need to avoid hypoglycaemia in our current and continued approach for optimal glycaemic control in people with diabetes admitted to hospital.

Antenatal preventative pelvic floor muscle exercise intervention led by midwives to reduce postnatal urinary incontinence (APPEAL): protocol for a feasibility and pilot cluster randomised controlled trial.

BACKGROUND: Antenatal pelvic floor muscle exercises (PFME) in women without prior urinary incontinence (UI) are effective in reducing postnatal UI; however, UK midwives often do not provide advice and information to women on undertaking PFME, with evidence that among women who do receive advice, many do not perform PFME. METHODS: The primary aim of this feasibility and pilot cluster randomised controlled trial is to provide a potential assessment of the feasibility of undertaking a future definitive trial of a midwifery-led antenatal intervention to support women to perform PFME in pregnancy and reduce UI postnatally. Community midwifery teams in participating NHS sites comprise trial clusters (n = 17). Midwives in teams randomised to the intervention will be trained on how to teach PFME to women and how to support them in undertaking PFME in pregnancy. Women whose community midwifery teams are allocated to control will receive standard antenatal care only. All pregnant women who give birth over a pre-selected sample month who receive antenatal care from participating community midwifery teams (clusters) will be sent a questionnaire at 10-12 weeks postpartum (around 1400-1500 women). Process evaluation data will include interviews with midwives to assess if the intervention could be implemented as planned. Interviews with women in both trial arms will explore their experiences of support from midwives to perform PFME during pregnancy. Data will be stored securely at the Universities of Birmingham and Exeter. Results will be disseminated through publications aimed at maternity service users, clinicians, and academics and inform a potential definitive trial of effectiveness. The West Midlands-Edgbaston Research Ethics Committee approved the study protocol. DISCUSSION: Trial outcomes will determine if criteria to progress to a definitive cluster trial are met. These include women's questionnaire return rates, prevalence of UI, and other health outcomes as reported by women at 10-12 weeks postpartum. Progress to a definitive trial however is likely to be prevented in the UK context by new perinatal pelvic health service, although this may be possible elsewhere. TRIAL REGISTRATION: https://doi.org/10.1186/ISRCTN10833250 . Registered 09/03/2020.

Fetal growth and birthweight standards as screening tools: methods for evaluating performance.

OBJECTIVE: To discuss different methods for evaluating fetal growth and population-based birthweight standards relevant to different uses: either in antenatal care or in epidemiology. DESIGN: Population-based cohort study. SETTING: Routinely collected data in Scotland. POPULATION: A total of 540,849 singletons born after 24 weeks between 1980 and 2003. METHODS: The performance of a fetal growth standard and a population-based birthweight standard are compared in two ways. First, we consider the accuracy of estimated risks of stillbirth at any point during the remaining pregnancy, a measure that is relevant in antenatal care. Second, the rates of stillbirth at each gestation, which are measures relevant in epidemiology, are compared with the actual rates. MAIN OUTCOME MEASURES: Standard measures of screening and diagnostic performance: sensitivity, specificity, and positive and negative predictive values. RESULTS: In clinical care, the evidence points towards using fetal growth standards: sensitivity at term is about 30%, increasing to 43% for preterm births (24-31 weeks of gestation), compared with about 29% across all ages under the birthweight standard. Positive predictive values are about 1% across gestations. For epidemiology, the evidence is not so clear-cut: preterm, the population birthweight standard has higher sensitivity and specificity, but this is not the case in the full-term weeks. CONCLUSIONS: The performance of fetal growth and birthweight standards should be evaluated in different ways, depending on whether they are intended for use in antenatal care or in epidemiological investigations.

Knowledge translation of prediction rules: methods to help health professionals understand their trade-offs.

Clinical prediction models are developed with the ultimate aim of improving patient outcomes, and are often turned into prediction rules (e.g. classifying people as low/high risk using cut-points of predicted risk) at some point during the development stage. Prediction rules often have reasonable ability to either rule-in or rule-out disease (or another event), but rarely both. When a prediction model is intended to be used as a prediction rule, conveying its performance using the C-statistic, the most commonly reported model performance measure, does not provide information on the magnitude of the trade-offs. Yet, it is important that these trade-offs are clear, for example, to health professionals who might implement the prediction rule. This can be viewed as a form of knowledge translation. When communicating information on trade-offs to patients and the public there is a large body of evidence that indicates natural frequencies are most easily understood, and one particularly well-received way of depicting the natural frequency information is to use population diagrams. There is also evidence that health professionals benefit from information presented in this way.Here we illustrate how the implications of the trade-offs associated with prediction rules can be more readily appreciated when using natural frequencies. We recommend that the reporting of the performance of prediction rules should (1) present information using natural frequencies across a range of cut-points to inform the choice of plausible cut-points and (2) when the prediction rule is recommended for clinical use at a particular cut-point the implications of the trade-offs are communicated using population diagrams. Using two existing prediction rules, we illustrate how these methods offer a means of effectively and transparently communicating essential information about trade-offs associated with prediction rules.

Comparison of small-sample standard-error corrections for generalised estimating equations in stepped wedge cluster randomised trials with a binary outcome: A simulation study.

Generalised estimating equations with the sandwich standard-error estimator provide a promising method of analysis for stepped wedge cluster randomised trials. However, they have inflated type-one error when used with a small number of clusters, which is common for stepped wedge cluster randomised trials. We present a large simulation study of binary outcomes comparing bias-corrected standard errors from Fay and Graubard; Mancl and DeRouen; Kauermann and Carroll; Morel, Bokossa, and Neerchal; and Mackinnon and White with an independent and exchangeable working correlation matrix. We constructed 95% confidence intervals using a t-distribution with degrees of freedom including clusters minus parameters (DFC-P), cluster periods minus parameters, and estimators from Fay and Graubard (DFFG), and Pan and Wall. Fay and Graubard and an approximation to Kauermann and Carroll (with simpler matrix inversion) were unbiased in a wide range of scenarios with an independent working correlation matrix and more than 12 clusters. They gave confidence intervals with close to 95% coverage with DFFG with 12 or more clusters, and DFC-P with 18 or more clusters. Both standard errors were conservative with fewer clusters. With an exchangeable working correlation matrix, approximated Kauermann and Carroll and Fay and Graubard had a small degree of under-coverage.

Is convalescent plasma futile in COVID-19? A Bayesian re-analysis of the RECOVERY randomized controlled trial.

BACKGROUND: Randomized trials are generally performed from a frequentist perspective, which can conflate absence of evidence with evidence of absence. The RECOVERY trial evaluated convalescent plasma for patients hospitalized with coronavirus disease 2019 (COVID-19) and concluded that there was no evidence of an effect. Re-analysis from a Bayesian perspective is warranted. METHODS: Outcome data were extracted from the RECOVERY trial by serostatus and time of presentation. A Bayesian re-analysis with a wide variety of priors (vague, optimistic, sceptical, and pessimistic) was performed, calculating the posterior probability for: any benefit, an absolute risk difference of 0.5% (small benefit, number needed to treat 200), and an absolute risk difference of one percentage point (modest benefit, number needed to treat 100). RESULTS: Across all patients, when analysed with a vague prior, the likelihood of any benefit or a modest benefit with convalescent plasma was estimated to be 64% and 18%, respectively. The estimated chance of any benefit was 95% if presenting within 7 days of symptoms, or 17% if presenting after this. In patients without a detectable antibody response at presentation, the chance of any benefit was 85%. However, it was only 20% in patients with a detectable antibody response at presentation. CONCLUSIONS: Bayesian re-analysis suggests that convalescent plasma reduces mortality by at least one percentage point among the 39% of patients who present within 7 days of symptoms, and that there is a 67% chance of the same mortality reduction in the 38% who are seronegative at the time of presentation. This is in contrast to the results in people who already have antibodies when they present. This biologically plausible finding bears witness to the advantage of Bayesian analyses over misuse of hypothesis tests to inform decisions.

Placental growth factor in assessment of women with suspected pre-eclampsia to reduce maternal morbidity: a stepped wedge cluster randomised control trial (PARROT Ireland).

OBJECTIVE: To determine whether the addition of placental growth factor (PlGF) measurement to current clinical assessment of women with suspected pre-eclampsia before 37 weeks' gestation would reduce maternal morbidity without increasing neonatal morbidity. DESIGN: Stepped wedge cluster randomised control trial from 29 June 2017 to 26 April 2019. SETTING: National multisite trial in seven maternity hospitals throughout the island of Ireland PARTICIPANTS: Women with a singleton pregnancy between 20+0 to 36+6 weeks' gestation, with signs or symptoms suggestive of evolving pre-eclampsia. Of the 5718 women screened, 2583 were eligible and 2313 elected to participate. INTERVENTION: Participants were assigned randomly to either usual care or to usual care plus the addition of point-of-care PlGF testing based on the randomisation status of their maternity hospital at the time point of enrolment. MAIN OUTCOMES MEASURES: Co-primary outcomes of composite maternal morbidity and composite neonatal morbidity. Analysis was on an individual participant level using mixed-effects Poisson regression adjusted for time effects (with robust standard errors) by intention-to-treat. RESULTS: Of the 4000 anticipated recruitment target, 2313 eligible participants (57%) were enrolled, of whom 2219 (96%) were included in the primary analysis. Of these, 1202 (54%) participants were assigned to the usual care group, and 1017 (46%) were assigned the intervention of additional point-of-care PlGF testing. The results demonstrate that the integration of point-of-care PlGF testing resulted in no evidence of a difference in maternal morbidity-457/1202 (38%) of women in the control group versus 330/1017 (32%) of women in the intervention group (adjusted risk ratio (RR) 1.01 (95% CI 0.76 to 1.36), P=0.92)-or in neonatal morbidity-527/1202 (43%) of neonates in the control group versus 484/1017 (47%) in the intervention group (adjusted RR 1.03 (0.89 to 1.21), P=0.67). CONCLUSIONS: This was a pragmatic evaluation of an interventional diagnostic test, conducted nationally across multiple sites. These results do not support the incorporation of PlGF testing into routine clinical investigations for women presenting with suspected preterm pre-eclampsia, but nor do they exclude its potential benefit. TRIAL REGISTRATION: ClinicalTrials.gov NCT02881073.

Meta-regression with partial information on summary trial or patient characteristics.

We present a model for meta-regression in the presence of missing information on some of the study level covariates, obtaining inferences using Bayesian methods. In practice, when confronted with missing covariate data in a meta-regression, it is common to carry out a complete case or available case analysis. We propose to use the full observed data, modelling the joint density as a factorization of a meta-regression model and a conditional factorization of the density for the covariates. With the inclusion of several covariates, inter-relations between these covariates are modelled. Under this joint likelihood-based approach, it is shown that the lesser assumption of the covariates being Missing At Random is imposed, instead of the more usual Missing Completely At Random (MCAR) assumption. The model is easily programmable in WinBUGS, and we examine, through the analysis of two real data sets, sensitivity and robustness of results to the MCAR assumption.

Impact of non-uniform correlation structure on sample size and power in multiple-period cluster randomised trials.

Stepped wedge and cluster randomised crossover trials are examples of cluster randomised designs conducted over multiple time periods that are being used with increasing frequency in health research. Recent systematic reviews of both of these designs indicate that the within-cluster correlation is typically taken account of in the analysis of data using a random intercept mixed model, implying a constant correlation between any two individuals in the same cluster no matter how far apart in time they are measured: within-period and between-period intra-cluster correlations are assumed to be identical. Recently proposed extensions allow the within- and between-period intra-cluster correlations to differ, although these methods require that all between-period intra-cluster correlations are identical, which may not be appropriate in all situations. Motivated by a proposed intensive care cluster randomised trial, we propose an alternative correlation structure for repeated cross-sectional multiple-period cluster randomised trials in which the between-period intra-cluster correlation is allowed to decay depending on the distance between measurements. We present results for the variance of treatment effect estimators for varying amounts of decay, investigating the consequences of the variation in decay on sample size planning for stepped wedge, cluster crossover and multiple-period parallel-arm cluster randomised trials. We also investigate the impact of assuming constant between-period intra-cluster correlations instead of decaying between-period intra-cluster correlations. Our results indicate that in certain design configurations, including the one corresponding to the proposed trial, a correlation decay can have an important impact on variances of treatment effect estimators, and hence on sample size and power. An R Shiny app allows readers to interactively explore the impact of correlation decay.