Group sequential designs for pragmatic clinical trials with early outcomes: methods and guidance for planning and implementation.

BACKGROUND: Group sequential designs are one of the most widely used methodologies for adaptive design in randomized clinical trials. In settings where early outcomes are available, they offer large gains in efficiency compared to a fixed design. However, such designs are underused and used predominantly in therapeutic areas where there is expertise and experience in implementation. One barrier to their greater use is the requirement to undertake simulation studies at the planning stage that require considerable knowledge, coding experience and additional costs. Based on some modest assumptions about the likely patterns of recruitment and the covariance structure of the outcomes, some simple analytic expressions are presented that negate the need to undertake simulations. METHODS: A model for longitudinal outcomes with an assumed approximate multivariate normal distribution and three contrasting simple recruitment models are described, based on fixed, increasing and decreasing rates. For assumed uniform and exponential correlation models, analytic expressions for the variance of the treatment effect and the effects of the early outcomes on reducing this variance at the primary outcome time-point are presented. Expressions for the minimum and maximum values show how the correlations and timing of the early outcomes affect design efficiency. RESULTS: Simulations showed how patterns of information accrual varied between correlation and recruitment models, and consequentially to some general guidance for planning a trial. Using a previously reported group sequential trial as an exemplar, it is shown how the analytic expressions given here could have been used as a quick and flexible planning tool, avoiding the need for extensive simulation studies based on individual participant data. CONCLUSIONS: The analytic expressions described can be routinely used at the planning stage of a putative trial, based on some modest assumptions about the likely number of outcomes and when they might occur and the expected recruitment patterns. Numerical simulations showed that these models behaved sensibly and allowed a range of design options to be explored in a way that would have been difficult and time-consuming if the previously described method of simulating individual trial participant data had been used.

Subacromial balloon spacer for irreparable rotator cuff tears of the shoulder (START:REACTS): a group-sequential, double-blind, multicentre randomised controlled trial.

BACKGROUND: New surgical procedures can expose patients to harm and should be carefully evaluated before widespread use. The InSpace balloon (Stryker, USA) is an innovative surgical device used to treat people with rotator cuff tears that cannot be repaired. We aimed to determine the effectiveness of the InSpace balloon for people with irreparable rotator cuff tears. METHODS: We conducted a double-blind, group-sequential, adaptive randomised controlled trial in 24 hospitals in the UK, comparing arthroscopic debridement of the subacromial space with biceps tenotomy (debridement only group) with the same procedure but including insertion of the InSpace balloon (debridement with device group). Participants had an irreparable rotator cuff tear, which had not resolved with conservative treatment, and they had symptoms warranting surgery. Eligibility was confirmed intraoperatively before randomly assigning (1:1) participants to a treatment group using a remote computer system. Participants and assessors were masked to group assignment. Masking was achieved by using identical incisions for both procedures, blinding the operation note, and a consistent rehabilitation programme was offered regardless of group allocation. The primary outcome was the Oxford Shoulder Score at 12 months. Pre-trial simulations using data from early and late timepoints informed stopping boundaries for two interim analyses. The primary analysis was on a modified intention-to-treat basis, adjusted for the planned interim analysis. The trial was registered with ISRCTN, ISRCTN17825590. FINDINGS: Between June 1, 2018, and July 30, 2020, we assessed 385 people for eligibility, of which 317 were eligible. 249 (79%) people consented for inclusion in the study. 117 participants were randomly allocated to a treatment group, 61 participants to the debridement only group and 56 to the debridement with device group. A predefined stopping boundary was met at the first interim analysis and recruitment stopped with 117 participants randomised. 43% of participants were female, 57% were male. We obtained primary outcome data for 114 (97%) participants. The mean Oxford Shoulder Score at 12 months was 34·3 (SD 11·1) in the debridement only group and 30·3 (10·9) in the debridement with device group (mean difference adjusted for adaptive design -4·2 [95% CI -8·2 to -0·26];p=0·037) favouring control. There was no difference in adverse events between the two groups. INTERPRETATION: In an efficient, adaptive trial design, our results favoured the debridement only group. We do not recommend the InSpace balloon for the treatment of irreparable rotator cuff tears. FUNDING: Efficacy and Mechanism Evaluation Programme, a Medical Research Council and National Institute for Health and Care Research partnership.

Adaptive enrichment designs with a continuous biomarker.

A popular design for clinical trials assessing targeted therapies is the two-stage adaptive enrichment design with recruitment in stage 2 limited to a biomarker-defined subgroup chosen based on data from stage 1. The data-dependent selection leads to statistical challenges if data from both stages are used to draw inference on treatment effects in the selected subgroup. If subgroups considered are nested, as when defined by a continuous biomarker, treatment effect estimates in different subgroups follow the same distribution as estimates in a group-sequential trial. This result is used to obtain tests controlling the familywise type I error rate (FWER) for six simple subgroup selection rules, one of which also controls the FWER for any selection rule. Two approaches are proposed: one based on multivariate normal distributions suitable if the number of possible subgroups, k, is small, and one based on Brownian motion approximations suitable for large k. The methods, applicable in the wide range of settings with asymptotically normal test statistics, are illustrated using survival data from a breast cancer trial.

Adjusting for treatment selection in phase II/III clinical trials with time to event data.

Phase II/III clinical trials are efficient two-stage designs that test multiple experimental treatments. In stage 1, patients are allocated to the control and all experimental treatments, with the data collected from them used to select experimental treatments to continue to stage 2. Patients recruited in stage 2 are allocated to the selected treatments and the control. Combined data of stage 1 and stage 2 are used for a confirmatory phase III analysis. Appropriate analysis needs to adjust for selection bias of the stage 1 data. Point estimators exist for normally distributed outcome data. Extending these estimators to time to event data is not straightforward because treatment selection is based on correlated treatment effects and stage 1 patients who do not get events in stage 1 are followed-up in stage 2. We have derived an approximately uniformly minimum variance conditional unbiased estimator (UMVCUE) and compared its biases and mean squared errors to existing bias adjusted estimators. In simulations, one existing bias adjusted estimator has similar properties as the practically unbiased UMVCUE while the others can have noticeable biases but they are less variable than the UMVCUE. For confirmatory phase II/III clinical trials where unbiased estimators are desired, we recommend the UMVCUE or the existing estimator with which it has similar properties.

Should the two-trial paradigm still be the gold standard in drug assessment?

Two significant pivotal trials are usually required for a new drug approval by a regulatory agency. This standard requirement is known as the two-trial paradigm. However, several authors have questioned why we need exactly two pivotal trials, what statistical error the regulators are trying to protect against, and potential alternative approaches. Therefore, it is important to investigate these questions to better understand the regulatory decision-making in the assessment of drugs' effectiveness. It is common that two identically designed trials are run solely to adhere to the two-trial rule. Previous work showed that combining the data from the two trials into a single trial (one-trial paradigm) would increase the power while ensuring the same level of type I error protection as the two-trial paradigm. However, this is true only under a specific scenario and there is little investigation on the type I error protection over the whole null region. In this article, we compare the two paradigms by considering scenarios in which the two trials are conducted in identical or different populations as well as with equal or unequal size. With identical populations, the results show that a single trial provides better type I error protection and higher power. Conversely, with different populations, although the one-trial rule is more powerful in some cases, it does not always protect against the type I error. Hence, there is the need for appropriate flexibility around the two-trial paradigm and the appropriate approach should be chosen based on the questions we are interested in.

Group sequential designs in pragmatic trials: feasibility and assessment of utility using data from a number of recent surgical RCTs.

BACKGROUND: Assessing the long term effects of many surgical interventions tested in pragmatic RCTs may require extended periods of participant follow-up to assess effectiveness and use patient-reported outcomes that require large sample sizes. Consequently the RCTs are often perceived as being expensive and time-consuming, particularly if the results show the test intervention is not effective. Adaptive, and particularly group sequential, designs have great potential to improve the efficiency and cost of testing new and existing surgical interventions. As a means to assess the potential utility of group sequential designs, we re-analyse data from a number of recent high-profile RCTs and assess whether using such a design would have caused the trial to stop early. METHODS: Many pragmatic RCTs monitor participants at a number of occasions (e.g. at 6, 12 and 24 months after surgery) during follow-up as a means to assess recovery and also to keep participants engaged with the trial process. Conventionally one of the outcomes is selected as the primary (final) outcome, for clinical reasons, with others designated as either early or late outcomes. In such settings, novel group sequential designs that use data from not only the final outcome but also from early outcomes at interim analyses can be used to inform stopping decisions. We describe data from seven recent surgical RCTs (WAT, DRAFFT, WOLLF, FASHION, CSAW, FIXDT, TOPKAT), and outline possible group sequential designs that could plausibly have been proposed at the design stage. We then simulate how these group sequential designs could have proceeded, by using the observed data and dates to replicate how information could have accumulated and decisions been made for each RCT. RESULTS: The results of the simulated group sequential designs showed that for two of the RCTs it was highly likely that they would have stopped for futility at interim analyses, potentially saving considerable time (15 and 23 months) and costs and avoiding patients being exposed to interventions that were either ineffective or no better than standard care. We discuss the characteristics of RCTs that are important in order to use the methodology we describe, particularly the value of early outcomes and the window of opportunity when early stopping decisions can be made and how it is related to the length of recruitment period and follow-up. CONCLUSIONS: The results for five of the RCTs tested showed that group sequential designs using early outcome data would have been feasible and likely to provide designs that were at least as efficient, and possibly more efficient, than the original fixed sample size designs. In general, the amount of information provided by the early outcomes was surprisingly large, due to the strength of correlations with the primary outcome. This suggests that the methods described here are likely to provide benefits more generally across the range of surgical trials and more widely in other application areas where trial designs, outcomes and follow-up patterns are structured and behave similarly.

Development of a model to demonstrate the impact of National Institute of Health and Care Excellence cost-effectiveness assessment on health utility for targeted medicines.

Advances in medical technology have led to a better understanding of heterogeneity of diseases and patients, and to the development of targeted medicines. This development is beneficial to society but can come at an increased cost to pharmaceutical manufacturers due to the costs associated with developing and manufacturing a diagnostic test. For such medicines, the conventional pricing structure, where a therapy is approved if it is deemed cost-effective, may not appropriately incentivize targeted drug development. We model the decision-making processes for both the healthcare provider and the pharmaceutical manufacturer, capturing their main priorities, and populate it with information from a recent appraisal by the National Institute of Health and Care Excellence. Healthcare providers prefer a stratified drug to be developed for a subgroup of the population when the drug is on average effective in the subgroup but with a detrimental effect in the complement. Whilst pharmaceutical manufacturers' preferences are similar, regions of disagreement exist. We show how preferences can be aligned by either penalizing the development of a non-stratified drug or rewarding the development of a stratified drug. The cost and position of alignment depends on the true value of health to the healthcare provider, among other parameters.

Effect of Noninvasive Respiratory Strategies on Intubation or Mortality Among Patients With Acute Hypoxemic Respiratory Failure and COVID-19: The RECOVERY-RS Randomized Clinical Trial.

Importance: Continuous positive airway pressure (CPAP) and high-flow nasal oxygen (HFNO) have been recommended for acute hypoxemic respiratory failure in patients with COVID-19. Uncertainty exists regarding the effectiveness and safety of these noninvasive respiratory strategies. Objective: To determine whether either CPAP or HFNO, compared with conventional oxygen therapy, improves clinical outcomes in hospitalized patients with COVID-19-related acute hypoxemic respiratory failure. Design, Setting, and Participants: A parallel group, adaptive, randomized clinical trial of 1273 hospitalized adults with COVID-19-related acute hypoxemic respiratory failure. The trial was conducted between April 6, 2020, and May 3, 2021, across 48 acute care hospitals in the UK and Jersey. Final follow-up occurred on June 20, 2021. Interventions: Adult patients were randomized to receive CPAP (n = 380), HFNO (n = 418), or conventional oxygen therapy (n = 475). Main Outcomes and Measures: The primary outcome was a composite of tracheal intubation or mortality within 30 days. Results: The trial was stopped prematurely due to declining COVID-19 case numbers in the UK and the end of the funded recruitment period. Of the 1273 randomized patients (mean age, 57.4 [95% CI, 56.7 to 58.1] years; 66% male; 65% White race), primary outcome data were available for 1260. Crossover between interventions occurred in 17.1% of participants (15.3% in the CPAP group, 11.5% in the HFNO group, and 23.6% in the conventional oxygen therapy group). The requirement for tracheal intubation or mortality within 30 days was significantly lower with CPAP (36.3%; 137 of 377 participants) vs conventional oxygen therapy (44.4%; 158 of 356 participants) (absolute difference, -8% [95% CI, -15% to -1%], P = .03), but was not significantly different with HFNO (44.3%; 184 of 415 participants) vs conventional oxygen therapy (45.1%; 166 of 368 participants) (absolute difference, -1% [95% CI, -8% to 6%], P = .83). Adverse events occurred in 34.2% (130/380) of participants in the CPAP group, 20.6% (86/418) in the HFNO group, and 13.9% (66/475) in the conventional oxygen therapy group. Conclusions and Relevance: Among patients with acute hypoxemic respiratory failure due to COVID-19, an initial strategy of CPAP significantly reduced the risk of tracheal intubation or mortality compared with conventional oxygen therapy, but there was no significant difference between an initial strategy of HFNO compared with conventional oxygen therapy. The study may have been underpowered for the comparison of HFNO vs conventional oxygen therapy, and early study termination and crossover among the groups should be considered when interpreting the findings. Trial Registration: isrctn.org Identifier: ISRCTN16912075.

Association of maternal vitamin B12 and folate levels in early pregnancy with gestational diabetes: a prospective UK cohort study (PRiDE study).

AIMS/HYPOTHESIS: The prevalence of gestational diabetes mellitus (GDM) is increasing worldwide in all ethnic groups. Low vitamin B12 and low/high folate levels may contribute to GDM risk, but there is conflicting evidence. Our aim is to assess the relationships of early pregnancy vitamin B12 and folate levels with the risk of GDM status at 26-28 weeks of gestation. METHODS: This was a prospective, multi-centre, multi-ethnic cohort study (n = 4746) in the UK. Participants who were eligible to be selectively screened as per the National Institute for Health and Care Excellence (NICE) criteria were included in the study. RESULTS: GDM prevalence was 12.5% by NICE and 14.7% by International Association of Diabetes and Pregnancy Study Groups (IADPSG) criteria. Folate deficiency (1.3%) was rare but B12 insufficiency (42.3% at <220 pmol/l) and folate excess (36.5%) were common in early pregnancy. Early pregnancy median B12 levels were lower, and folate levels higher, in women who were diagnosed with GDM at 26-28 weeks. B12 was negatively associated with fasting plasma glucose (1 SD: -0.06 mmol/l; 95% CI -0.04, -0.08; p < 0.0001) and 2 h plasma glucose levels (-0.07 mmol/l; 95% CI -0.02, -0.12; p = 0.004). Higher B12 was associated with 14.4% lower RR of IADPSG-GDM (0.856; 95% CI 0.786, 0.933; p = 0.0004) after adjusting for key confounders (age, parity, smoking status, ethnicity, family history, household income and folate status). Approximately half of this association was mediated through BMI. Folate was positively associated with 2 h plasma glucose levels (0.08 mmol/l; 95% CI 0.04, 0.13; p = 0.0005) but its relationship with fasting plasma glucose was U-shaped (quadratic ß: 0.011; p = 0.05). Higher folate was associated with 11% higher RR of IADPSG-GDM (adjusted RR 1.11; 95% CI 1.036, 1.182; p = 0.002) (age, parity, smoking status, ethnicity, family history, household income and B12 status). Although no interactions were observed for B12 and folate (as continuous variables) with glucose levels and GDM risk, a low B12-high folate combination was associated with higher blood glucose level and risk of IADPSG-GDM (adjusted RR 1.742; 95% CI 1.226, 2.437; p = 0.003). CONCLUSIONS/INTERPRETATION: B12 insufficiency and folate excess were common in early pregnancy. Low B12 and high folate levels in early pregnancy were associated with small but statistically significant changes in maternal blood glucose level and higher RR of GDM. Our findings warrant additional studies on the role of unmetabolised folic acid in glucose metabolism and investigating the effect of optimising early pregnancy or pre-conception B12 and folate levels on subsequent hyperglycaemia. TRIAL REGISTRATION: ClinicalTrials.gov NCT03008824.

A Randomized Trial of Epinephrine in Out-of-Hospital Cardiac Arrest.

BACKGROUND: Concern about the use of epinephrine as a treatment for out-of-hospital cardiac arrest led the International Liaison Committee on Resuscitation to call for a placebo-controlled trial to determine whether the use of epinephrine is safe and effective in such patients. METHODS: In a randomized, double-blind trial involving 8014 patients with out-of-hospital cardiac arrest in the United Kingdom, paramedics at five National Health Service ambulance services administered either parenteral epinephrine (4015 patients) or saline placebo (3999 patients), along with standard care. The primary outcome was the rate of survival at 30 days. Secondary outcomes included the rate of survival until hospital discharge with a favorable neurologic outcome, as indicated by a score of 3 or less on the modified Rankin scale (which ranges from 0 [no symptoms] to 6 [death]). RESULTS: At 30 days, 130 patients (3.2%) in the epinephrine group and 94 (2.4%) in the placebo group were alive (unadjusted odds ratio for survival, 1.39; 95% confidence interval [CI], 1.06 to 1.82; P=0.02). There was no evidence of a significant difference in the proportion of patients who survived until hospital discharge with a favorable neurologic outcome (87 of 4007 patients [2.2%] vs. 74 of 3994 patients [1.9%]; unadjusted odds ratio, 1.18; 95% CI, 0.86 to 1.61). At the time of hospital discharge, severe neurologic impairment (a score of 4 or 5 on the modified Rankin scale) had occurred in more of the survivors in the epinephrine group than in the placebo group (39 of 126 patients [31.0%] vs. 16 of 90 patients [17.8%]). CONCLUSIONS: In adults with out-of-hospital cardiac arrest, the use of epinephrine resulted in a significantly higher rate of 30-day survival than the use of placebo, but there was no significant between-group difference in the rate of a favorable neurologic outcome because more survivors had severe neurologic impairment in the epinephrine group. (Funded by the U.K. National Institute for Health Research and others; Current Controlled Trials number, ISRCTN73485024 .).

Identification of subgroup effect with an individual participant data meta-analysis of randomised controlled trials of three different types of therapist-delivered care in low back pain.

BACKGROUND: Proven treatments for low back pain, at best, only provide modest overall benefits. Matching people to treatments that are likely to be most effective for them may improve clinical outcomes and makes better use of health care resources. METHODS: We conducted an individual participant data meta-analysis of randomised controlled trials of three types of therapist delivered interventions for low back pain (active physical, passive physical and psychological treatments). We applied two statistical methods (recursive partitioning and adaptive risk group refinement) to identify potential subgroups who might gain greater benefits from different treatments from our individual participant data meta-analysis. RESULTS: We pooled data from 19 randomised controlled trials, totalling 9328 participants. There were 5349 (57%) females with similar ratios of females in control and intervention arms. The average age was 49 years (standard deviation, SD, 14). Participants with greater psychological distress and physical disability gained most benefit in improving on the mental component scale (MCS) of SF-12/36 from passive physical treatment than non-active usual care (treatment effects, 4.3; 95% confidence interval, CI, 3.39 to 5.15). Recursive partitioning method found that participants with worse disability at baseline gained most benefit in improving the disability (Roland Morris Disability Questionnaire) outcome from psychological treatment than non-active usual care (treatment effects, 1.7; 95% CI, 1.1 to 2.31). Adaptive risk group refinement did not find any subgroup that would gain much treatment effect between psychological and non-active usual care. Neither statistical method identified any subgroups who would gain an additional benefit from active physical treatment compared to non-active usual care. CONCLUSIONS: Our methodological approaches worked well and may have applicability in other clinical areas. Passive physical treatments were most likely to help people who were younger with higher levels of disability and low levels of psychological distress. Psychological treatments were more likely to help those with severe disability. Despite this, the clinical importance of identifying these subgroups is limited. The sizes of sub-groups more likely to benefit and the additional effect sizes observed are small. Our analyses provide no evidence to support the use of sub-grouping for people with low back pain.

Point and interval estimation in two-stage adaptive designs with time to event data and biomarker-driven subpopulation selection.

In personalized medicine, it is often desired to determine if all patients or only a subset of them benefit from a treatment. We consider estimation in two-stage adaptive designs that in stage 1 recruit patients from the full population. In stage 2, patient recruitment is restricted to the part of the population, which, based on stage 1 data, benefits from the experimental treatment. Existing estimators, which adjust for using stage 1 data for selecting the part of the population from which stage 2 patients are recruited, as well as for the confirmatory analysis after stage 2, do not consider time to event patient outcomes. In this work, for time to event data, we have derived a new asymptotically unbiased estimator for the log hazard ratio and a new interval estimator with good coverage probabilities and probabilities that the upper bounds are below the true values. The estimators are appropriate for several selection rules that are based on a single or multiple biomarkers, which can be categorical or continuous.

Comparison of Bayesian and frequentist group-sequential clinical trial designs.

BACKGROUND: There is a growing interest in the use of Bayesian adaptive designs in late-phase clinical trials. This includes the use of stopping rules based on Bayesian analyses in which the frequentist type I error rate is controlled as in frequentist group-sequential designs. METHODS: This paper presents a practical comparison of Bayesian and frequentist group-sequential tests. Focussing on the setting in which data can be summarised by normally distributed test statistics, we evaluate and compare boundary values and operating characteristics. RESULTS: Although Bayesian and frequentist group-sequential approaches are based on fundamentally different paradigms, in a single arm trial or two-arm comparative trial with a prior distribution specified for the treatment difference, Bayesian and frequentist group-sequential tests can have identical stopping rules if particular critical values with which the posterior probability is compared or particular spending function values are chosen. If the Bayesian critical values at different looks are restricted to be equal, O'Brien and Fleming's design corresponds to a Bayesian design with an exceptionally informative negative prior, Pocock's design to a Bayesian design with a non-informative prior and frequentist designs with a linear alpha spending function are very similar to Bayesian designs with slightly informative priors.This contrasts with the setting of a comparative trial with independent prior distributions specified for treatment effects in different groups. In this case Bayesian and frequentist group-sequential tests cannot have the same stopping rule as the Bayesian stopping rule depends on the observed means in the two groups and not just on their difference. In this setting the Bayesian test can only be guaranteed to control the type I error for a specified range of values of the control group treatment effect. CONCLUSIONS: Comparison of frequentist and Bayesian designs can encourage careful thought about design parameters and help to ensure appropriate design choices are made.

Adaptive seamless clinical trials using early outcomes for treatment or subgroup selection: Methods, simulation model and their implementation in R.

Adaptive seamless designs combine confirmatory testing, a domain of phase III trials, with features such as treatment or subgroup selection, typically associated with phase II trials. They promise to increase the efficiency of development programmes of new drugs, for example, in terms of sample size and/or development time. It is well acknowledged that adaptive designs are more involved from a logistical perspective and require more upfront planning, often in the form of extensive simulation studies, than conventional approaches. Here, we present a framework for adaptive treatment and subgroup selection using the same notation, which links the somewhat disparate literature on treatment selection on one side and on subgroup selection on the other. Furthermore, we introduce a flexible and efficient simulation model that serves both designs. As primary endpoints often take a long time to observe, interim analyses are frequently informed by early outcomes. Therefore, all methods presented accommodate interim analyses informed by either the primary outcome or an early outcome. The R package asd, previously developed to simulate designs with treatment selection, was extended to include subgroup selection (so-called adaptive enrichment designs). Here, we describe the functionality of the R package asd and use it to present some worked-up examples motivated by clinical trials in chronic obstructive pulmonary disease and oncology. The examples both illustrate various features of the R package and provide insights into the operating characteristics of adaptive seamless studies.

HIV self-testing alone or with additional interventions, including financial incentives, and linkage to care or prevention among male partners of antenatal care clinic attendees in Malawi: An adaptive multi-arm, multi-stage cluster randomised trial.

BACKGROUND: Conventional HIV testing services have been less comprehensive in reaching men than in reaching women globally, but HIV self-testing (HIVST) appears to be an acceptable alternative. Measurement of linkage to post-test services following HIVST remains the biggest challenge, yet is the biggest driver of cost-effectiveness. We investigated the impact of HIVST alone or with additional interventions on the uptake of testing and linkage to care or prevention among male partners of antenatal care clinic attendees in a novel adaptive trial. METHODS AND FINDINGS: An adaptive multi-arm, 2-stage cluster randomised trial was conducted between 8 August 2016 and 30 June 2017, with antenatal care clinic (ANC) days (i.e., clusters of women attending on a single day) as the unit of randomisation. Recruitment was from Ndirande, Bangwe, and Zingwangwa primary health clinics in urban Blantyre, Malawi. Women attending an ANC for the first time for their current pregnancy (regardless of trimester), 18 years and older, with a primary male partner not known to be on ART were enrolled in the trial after giving consent. Randomisation was to either the standard of care (SOC; with a clinic invitation letter to the male partner) or 1 of 5 intervention arms: the first arm provided women with 2 HIVST kits for their partners; the second and third arms provided 2 HIVST kits along with a conditional fixed financial incentive of $3 or $10; the fourth arm provided 2 HIVST kits and a 10% chance of receiving $30 in a lottery; and the fifth arm provided 2 HIVST kits and a phone call reminder for the women's partners. The primary outcome was the proportion of male partners who were reported to have tested for HIV and linked into care or prevention within 28 days, with referral for antiretroviral therapy (ART) or circumcision accordingly. Women were interviewed at 28 days about partner testing and adverse events. Cluster-level summaries compared each intervention versus SOC using eligible women as the denominator (intention-to-treat). Risk ratios were adjusted for male partner testing history and recruitment clinic. A total of 2,349/3,137 (74.9%) women participated (71 ANC days), with a mean age of 24.8 years (SD: 5.4). The majority (2,201/2,233; 98.6%) of women were married, 254/2,107 (12.3%) were unable to read and write, and 1,505/2,247 (67.0%) were not employed. The mean age for male partners was 29.6 years (SD: 7.5), only 88/2,200 (4.0%) were unemployed, and 966/2,210 (43.7%) had never tested for HIV before. Women in the SOC arm reported that 17.4% (71/408) of their partners tested for HIV, whereas a much higher proportion of partners were reported to have tested for HIV in all intervention arms (87.0%-95.4%, p < 0.001 in all 5 intervention arms). As compared with those who tested in the SOC arm (geometric mean 13.0%), higher proportions of partners met the primary endpoint in the HIVST + $3 (geometric mean 40.9%, adjusted risk ratio [aRR] 3.01 [95% CI 1.63-5.57], p < 0.001), HIVST + $10 (51.7%, aRR 3.72 [95% CI 1.85-7.48], p < 0.001), and phone reminder (22.3%, aRR 1.58 [95% CI 1.07-2.33], p = 0.021) arms. In contrast, there was no significant increase in partners meeting the primary endpoint in the HIVST alone (geometric mean 17.5%, aRR 1.45 [95% CI 0.99-2.13], p = 0.130) or lottery (18.6%, aRR 1.43 [95% CI 0.96-2.13], p = 0.211) arms. The lottery arm was dropped at interim analysis. Overall, 46 male partners were confirmed to be HIV positive, 42 (91.3%) of whom initiated ART within 28 days; 222 tested HIV negative and were not already circumcised, of whom 135 (60.8%) were circumcised as part of the trial. No serious adverse events were reported. Costs per male partner who attended the clinic with a confirmed HIV test result were $23.73 and $28.08 for the HIVST + $3 and HIVST + $10 arms, respectively. Notable limitations of the trial included the relatively small number of clusters randomised to each arm, proxy reporting of the male partner testing outcome, and being unable to evaluate retention in care. CONCLUSIONS: In this study, the odds of men's linkage to care or prevention increased substantially using conditional fixed financial incentives plus partner-delivered HIVST; combinations were potentially affordable. TRIAL REGISTRATION: ISRCTN 18421340.

To add or not to add a new treatment arm to a multiarm study: A decision-theoretic framework.

Multiarm clinical trials, which compare several experimental treatments against control, are frequently recommended due to their efficiency gain. In practise, all potential treatments may not be ready to be tested in a phase II/III trial at the same time. It has become appealing to allow new treatment arms to be added into on-going clinical trials using a "platform" trial approach. To the best of our knowledge, many aspects of when to add arms to an existing trial have not been explored in the literature. Most works on adding arm(s) assume that a new arm is opened whenever a new treatment becomes available. This strategy may prolong the overall duration of a study or cause reduction in marginal power for each hypothesis if the adaptation is not well accommodated. Within a two-stage trial setting, we propose a decision-theoretic framework to investigate when to add or not to add a new treatment arm based on the observed stage one treatment responses. To account for different prospect of multiarm studies, we define utility in two different ways; one for a trial that aims to maximise the number of rejected hypotheses; the other for a trial that would declare a success when at least one hypothesis is rejected from the study. Our framework shows that it is not always optimal to add a new treatment arm to an existing trial. We illustrate a case study by considering a completed trial on knee osteoarthritis.

Using Bayesian adaptive designs to improve phase III trials: a respiratory care example.

BACKGROUND: Bayesian adaptive designs can improve the efficiency of trials, and lead to trials that can produce high quality evidence more quickly, with fewer patients and lower costs than traditional methods. The aim of this work was to determine how Bayesian adaptive designs can be constructed for phase III clinical trials in critical care, and to assess the influence that Bayesian designs would have on trial efficiency and study results. METHODS: We re-designed the High Frequency OSCillation in Acute Respiratory distress syndrome (OSCAR) trial using Bayesian adaptive design methods, to allow for the possibility of early stopping for success or futility. We constructed several alternative designs and studied their operating characteristics via simulation. We then performed virtual re-executions by applying the Bayesian adaptive designs using the OSCAR data to demonstrate the practical applicability of the designs. RESULTS: We constructed five alternative Bayesian adaptive designs and identified a preferred design based on the simulated operating characteristics, which had similar power to the original design but recruited fewer patients on average. The virtual re-executions showed the Bayesian sequential approach and original OSCAR trial yielded similar trial conclusions. However, using a Bayesian sequential design could have led to a reduced sample size and earlier completion of the trial. CONCLUSIONS: Using the OSCAR trial as an example, this case study found that Bayesian adaptive designs can be constructed for phase III critical care trials. If the OSCAR trial had been run using one of the proposed Bayesian adaptive designs, it would have terminated at a smaller sample size with fewer deaths in the trial, whilst reaching the same conclusions. We recommend the wider use of Bayesian adaptive approaches in phase III clinical trials. TRIAL REGISTRATION: OSCAR Trial registration ISRCTN, ISRCTN10416500 . Retrospectively registered 13 June 2007.

An alternative method to analyse the biomarker-strategy design.

Recent developments in genomics and proteomics enable the discovery of biomarkers that allow identification of subgroups of patients responding well to a treatment. One currently used clinical trial design incorporating a predictive biomarker is the so-called biomarker strategy design (or marker-based strategy design). Conventionally, the results from this design are analysed by comparing the mean of the biomarker-led arm with the mean of the randomised arm. Several problems regarding the analysis of the data obtained from this design have been identified in the literature. In this paper, we show how these problems can be resolved if the sample sizes in the subgroups fulfil the specified orthogonality condition. We also propose a different analysis strategy that allows definition of test statistics for the biomarker-by-treatment interaction effect as well as for the classical treatment effect and the biomarker effect. We derive equations for the sample size calculation for the case of perfect and imperfect biomarker assays. We also show that the often used 1:1 randomisation does not necessarily lead to the smallest sample size. In addition, we provide point estimators and confidence intervals for the treatment effects in the subgroups. Application of our method is illustrated using a real data example.

A recursive partitioning approach for subgroup identification in individual patient data meta-analysis.

BACKGROUND: Motivated by the setting of clinical trials in low back pain, this work investigated statistical methods to identify patient subgroups for which there is a large treatment effect (treatment by subgroup interaction). Statistical tests for interaction are often underpowered. Individual patient data (IPD) meta-analyses provide a framework with improved statistical power to investigate subgroups. However, conventional approaches to subgroup analyses applied in both a single trial setting and an IPD setting have a number of issues, one of them being that factors used to define subgroups are investigated one at a time. As individuals have multiple characteristics that may be related to response to treatment, alternative exploratory statistical methods are required. METHODS: Tree-based methods are a promising alternative that systematically searches the covariate space to identify subgroups defined by multiple characteristics. A tree method in particular, SIDES, is described and extended for application in an IPD meta-analyses setting by incorporating fixed-effects and random-effects models to account for between-trial variation. The performance of the proposed extension was assessed using simulation studies. The proposed method was then applied to an IPD low back pain dataset. RESULTS: The simulation studies found that the extended IPD-SIDES method performed well in detecting subgroups especially in the presence of large between-trial variation. The IPD-SIDES method identified subgroups with enhanced treatment effect when applied to the low back pain data. CONCLUSIONS: This work proposes an exploratory statistical approach for subgroup analyses applicable in any research discipline where subgroup analyses in an IPD meta-analysis setting are of interest.