Adaptive Trial Research Articles

Optimizing decision-making in uncertain environments through analysis of stochastic stationary Multi-Armed Bandit algorithms

Reinforcement learning traditionally plays a pivotal role in artificial intelligence and various practical applications, focusing on the interaction between an agent and its environment. Within this broad field, the multi-armed bandit (MAB) problem represents a specific subset, characterized by a sequential interaction between a learner and an environment where the agents actions do not alter the environment or reward distributions. MABs are prevalent in recommendation systems and advertising and are increasingly applied in sectors like agriculture and adaptive clinical trials. The stochastic stationary bandit problem, a fundamental category of MAB, is the primary focus of this article. Here, we delve into the implementation and analytical comparison of several key bandit algorithmsincluding Explore-then-Commit (ETC), Upper Confidence Bound (UCB), Thompson Sampling (TS), Epsilon-Greedy (-Greedy), SoftMax, and Conservative Lower Confidence Bound (CON-LCB)across various datasets. These datasets vary in the number of options (arms), reward distributions, and specific parameters, offering a broad testing ground. Additionally, this work provides an overview of the diverse applications of bandit problems across different fields, highlighting their versatility and broad impact.

Two-stage randomized clinical trials with a right-censored endpoint: Comparison of frequentist and Bayesian adaptive designs.

Adaptive randomized clinical trials are of major interest when dealing with a time-to-event outcome in a prolonged observation window. No consensus exists either to define stopping boundaries or to combine values or test statistics in the terminal analysis in the case of a frequentist design and sample size adaptation. In a one-sided setting, we compared three frequentist approaches using stopping boundaries relying on -spending functions and a Bayesian monitoring setting with boundaries based on the posterior distribution of the log-hazard ratio. All designs comprised a single interim analysis with an efficacy stopping rule and the possibility of sample size adaptation at this interim step. Three frequentist approaches were defined based on the terminal analysis: combination of stagewise statistics (Wassmer) or of values (Desseaux), or on patientwise splitting (Jörgens), and we compared the results with those of the Bayesian monitoring approach (Freedman). These different approaches were evaluated in a simulation study and then illustrated on a real dataset from a randomized clinical trial conducted in elderly patients with chronic lymphocytic leukemia. All approaches controlled for the type I error rate, except for the Bayesian monitoring approach, and yielded satisfactory power. It appears that the frequentist approaches are the best in underpowered trials. The power of all the approaches was affected by the violation of the proportional hazards (PH) assumption. For adaptive designs with a survival endpoint and a one-sided alternative hypothesis, the Wassmer and Jörgens approaches after sample size adaptation should be preferred, unless violation of PH is suspected.

Adaptive designs were primarily used but inadequately reported in early phase drug trials

BackgroundFaced with the high cost and limited efficiency of classical randomized controlled trials, researchers are increasingly applying adaptive designs to speed up the development of new drugs. However, the application of adaptive design to drug randomized controlled trials (RCTs) and whether the reporting is adequate are unclear. Thus, this study aimed to summarize the epidemiological characteristics of the relevant trials and assess their reporting quality by the Adaptive designs CONSORT Extension (ACE) checklist.MethodsWe searched MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL) and ClinicalTrials.gov from inception to January 2020. We included drug RCTs that explicitly claimed to be adaptive trials or used any type of adaptative design. We extracted the epidemiological characteristics of included studies to summarize their adaptive design application. We assessed the reporting quality of the trials by Adaptive designs CONSORT Extension (ACE) checklist. Univariable and multivariable linear regression models were used to the association of four prespecified factors with the quality of reporting.ResultsOur survey included 108 adaptive trials. We found that adaptive design has been increasingly applied over the years, and was commonly used in phase II trials (n = 45, 41.7%). The primary reasons for using adaptive design were to speed the trial and facilitate decision-making (n = 24, 22.2%), maximize the benefit of participants (n = 21, 19.4%), and reduce the total sample size (n = 15, 13.9%). Group sequential design (n = 63, 58.3%) was the most frequently applied method, followed by adaptive randomization design (n = 26, 24.1%), and adaptive dose-finding design (n = 24, 22.2%). The proportion of adherence to the ACE checklist of 26 topics ranged from 7.4 to 99.1%, with eight topics being adequately reported (i.e., level of adherence ≥ 80%), and eight others being poorly reported (i.e., level of adherence ≤ 30%). In addition, among the seven items specific for adaptive trials, three were poorly reported: accessibility to statistical analysis plan (n = 8, 7.4%), measures for confidentiality (n = 14, 13.0%), and assessments of similarity between interim stages (n = 25, 23.1%). The mean score of the ACE checklist was 13.9 (standard deviation [SD], 3.5) out of 26. According to our multivariable regression analysis, later published trials (estimated β = 0.14, p < 0.01) and the multicenter trials (estimated β = 2.22, p < 0.01) were associated with better reporting.ConclusionAdaptive design has shown an increasing use over the years, and was primarily applied to early phase drug trials. However, the reporting quality of adaptive trials is suboptimal, and substantial efforts are needed to improve the reporting.

Graves-PCD: protocol for a randomised, dose-finding, adaptive trial of the plasma cell-depleting agent daratumumab in severe Graves’ disease

IntroductionSevere Graves’ disease is a life-changing condition with poor outcomes from currently available treatments. It is caused by directly pathogenic thyroid-stimulating hormone receptor-stimulating antibodies (TRAb), which are secreted from plasma...

CYCLONE 2: A phase 3 study of abemaciclib with abiraterone in patients with metastatic castration-resistant prostate cancer.

5001 Background: Oncogenic addiction to androgen receptor (AR) signaling drives mCRPC progression, highlighting the unmet need for novel treatment strategies to maximize AR-directed therapy. Preclinical evidence suggests a key role for CDK4/6 in sustained AR signaling, uncontrolled proliferation, and hormonal resistance in prostate cancer. Abemaciclib (ABEMA) is a potent CDK4/6 oral inhibitor that significantly augments the efficacy of endocrine therapy in hormonally driven (ER+) high-risk early-stage and metastatic breast cancer. ABEMA also showed single-agent activity in heavily pretreated mCRPC. Here, we report the primary results of CYCLONE 2, a Phase 3 study of ABEMA plus abiraterone (ABI) in pts with 1L mCRPC. Methods: CYCLONE 2 was a seamless Phase 2/3 adaptive trial with a dose-finding safety lead-in. Randomization to the ABEMA or placebo (PBO) plus ABI and predniso(lo)ne was stratified by prior docetaxel receipt for mHSPC, measurable disease, and radiographic progression at study entry. Primary endpoint was investigator-assessed radiographic progression-free survival (rPFS) per RECIST v1.1 and PCWG3. The study was powered at ~90%, assuming a HR of 0.55 for rPFS, at a cumulative 2-sided alpha level of 0.05. Results: Between Nov 2018 and Jul 2022, 393 pts were randomized. Baseline characteristics were balanced across arms. Primary endpoint of rPFS was not met (HR 0.829; 95% CI, 0.619–1.111; p=0.2123), medians were 21.96 months for the ABEMA plus ABI group vs 20.28 months for the PBO plus ABI group. rPFS by blinded independent central review was consistent with investigator assessment (HR 0.842; 95% CI, 0.611–1.160). OS was a gated secondary endpoint and not inferentially tested (HR 0.927; 95% CI, 0.669–1.285; 38.9% maturity). Other secondary endpoints included time to PSA progression (HR 0.637; 95% CI, 0.474–0.856), time to symptomatic progression (HR 0.768; 95% CI, 0.522–1.131), and time to worst pain progression (HR 0.935; 95% CI 0.665–1.314). The most common grade ≥3 adverse events (AEs) reported in the ABEMA plus ABI group were anemia (13.6% vs 4.3% in the PBO plus ABI group), neutropenia (12.6% vs 0.5%) and ALT increased (8.7% vs 6.5%). Discontinuations of all study treatments due to AEs were 13.1% vs 4.3% in ABEMA plus ABI vs PBO plus ABI groups, while discontinuations of ABEMA or PBO alone due to AEs were 5.8% vs 1.6%, respectively. Conclusions: In patients with mCRPC, adding abemaciclib to abiraterone did not significantly increase rPFS. While no OS detriment was observed, secondary endpoints were not meaningfully improved. Overall, the combination was well tolerated, and safety was consistent with the known profiles of the individual medicines. Clinical trial information: NCT03706365 Clinical trial information: NCT03706365 .

P090 Mission (almost) impossible: setting up a complex investigator led international adaptive platform trial – Finding the Optimal Regimen for Mycobacterium abscessus Treatment (FORMaT)

P090 Mission (almost) impossible: setting up a complex investigator led international adaptive platform trial – Finding the Optimal Regimen for Mycobacterium abscessus Treatment (FORMaT)

Evaluation of VAL-083 in GBM AGILE, a phase 3 registration platform trial for newly diagnosed and recurrent glioblastoma.

2005 Background: GBM AGILE (NCT03970447) is a phase 2/3 Bayesian adaptive registration platform trial testing multiple therapies efficiently against a common control (C) with a primary endpoint of overall survival (OS). VAL-083 (VAL) is a DNA targeting agent that, independent of O6-methylguanine DNA methyltransferase promoter methylation status, targets the N7 position of guanine residues and facilitates inter-strand DNA crosslinks, leading to DNA double-strand breaks and cell death. It entered the trial in January 2021, and it is the 2nd arm (of 6) to complete its evaluation. Methods: Patient subtypes considered in GBM AGILE are newly diagnosed methylated (NDM), ND unmethylated (NDU), &amp; recurrent disease (RD). C is temozolomide (in ND) &amp; lomustine (in RD). Arms open to all 3 subtypes are evaluated in 5 prospectively defined signatures (sig): NDU, NDM, RD, all ND and All. Randomization to C is 20% in each subtype. Exp arms in GBM AGILE have 1 or 2 stages. Efficacy is based on OS hazard ratio (HR) of arm/C. Efficacy goal is a final Bayesian probability ≥ 98% for HR &lt;1.00 in combined Stages 1 &amp; 2. Arms stop accruing in Stage 1 if they reach max sample size (N) or drop for futility or safety. Exp arms in Stage 1 are adaptively randomized with allocation being proportional to an arm’s current probability of having ≥ 30% benefit in OS, P(HR &lt;0.70). In stage 1, exp arms are evaluated monthly, and arms showing Bayesian predictive power (PP) ≥ 0.8, graduate into Stage 2 with fixed randomization in one sig. For all exp arms, follow up continues for 12 mos after accrual stops (clinical cutoff). Arms are declared futile at any monthly analysis when PP is &lt;0.25 for all sigs. Open to all 3 subtypes, VAL entered as the 1st arm in NDM and was randomized 1:1 to C in this subtype until additional arms entered. The target max N for VAL in its Stages 1 &amp; 2 were 150 and 50, resp. Results: At the interim after VAL reached max sample size in Stage 1, the PP for all signatures was &lt;0.8 and &gt;0.25 for at least one sig. Thus, VAL did not graduate nor drop for futility, but accrual stopped for maximum N in Stage 1 (see table). Final results will be presented at the meeting. Columns 2-5 show results at the interim after which VAL stopped for max N. Columns 6-8 show near final results. Conclusions: GBM AGILE is an efficient &amp; effective model for phase 3 drug development. VAL did not increase OS compared to C in any glioblastoma subtype. GBM AGILE evaluated this agent in less time, at lower cost, &amp; with fewer patients than typical registration trials &amp; is currently evaluating several other arms. Clinical trial information: NCT03970447 . [Table: see text]

P089 FORMaT: Finding the Optimal Regimen for Mycobacterium abscessus Treatment. A randomised, multi-arm, adaptive platform trial

P089 FORMaT: Finding the Optimal Regimen for Mycobacterium abscessus Treatment. A randomised, multi-arm, adaptive platform trial

Phase 1 Open-Label Study of Omigapil in Patients With LAMA2- or COL6-Related Dystrophy.

Omigapil is a small molecule which inhibits the GAPDH-Siah1-mediated apoptosis pathway. Apoptosis is a pathomechanism underlying the congenital muscular dystrophy subtypes LAMA2-related dystrophy (LAMA2-RD) and COL6-related dystrophy (COL6-RD). Studies of omigapil in the (dyw/dyw) LAMA2-RD mouse model demonstrated improved survival, and studies in the (dy2J/dy2J) LAMA2-RD mouse model and the (Col6a1-/-) COL6-RD mouse model demonstrated decreased apoptosis. A phase 1 open-label, sequential group, ascending oral dose, cohort study of omigapil in patients with LAMA2-RD or COL6-RD ages 5-16 years was performed (1) to establish the pharmacokinetic (PK) profile of omigapil at a range of doses, (2) to evaluate the safety and tolerability of omigapil at a range of doses, and (3) to establish the feasibility of conducting disease-relevant clinical assessments. Patients were enrolled in cohorts of size 4, with each patient receiving 4 weeks of vehicle run-in and 12 weeks of study drug (at daily doses ranging from 0.02 to 0.08 mg/kg). PK data from each cohort were analyzed before each subsequent dosing cohort was enrolled. A novel, adaptive dose-finding method (stochastic approximation with virtual observation recursion) was used to allow for dose escalation/reduction between cohorts based on PK data. Twenty patients were enrolled at the NIH (LAMA2-RD: N = 10; COL6-RD: N = 10). Slightly greater than dose-proportional increases in systemic exposure to omigapil were seen at doses 0.02-0.08 mg/kg/d. The dose which achieved patient exposure within the pre-established target area under the plasma concentration-vs-time curve (AUC0-24h) range was 0.06 mg/kg/d. In general, omigapil was safe and well tolerated. No consistent changes were seen in the disease-relevant clinical assessments during the duration of the study. This study represents the thus far only clinical trial of a therapeutic small molecule for LAMA2-RD and COL6-RD, completed with an adaptive trial design to arrive at dose adjustments. The trial met its primary end point and established that the PK profile of omigapil is suitable for further development in pediatric patients with LAMA2-RD or COL6-RD, the most common forms of congenital muscular dystrophy. While within the short duration of the study disease-relevant clinical assessments did not demonstrate significant changes, this study establishes the feasibility of performing interventional clinical trials in these rare disease patient populations. This study provides Class IV evidence of omigapil in a dose-finding phase 1 study. Clinical Trials NCT01805024.

Impact of an Adaptive Worksite Diabetes Prevention Trial on Health Action Process Approach Outcomes Regarding Dietary Intake and Physical Activity.

Weight loss, through a reduction in energy intake and increase in energy expenditure, can reduce diabetes risk in people with prediabetes. However, lifestyle change can be challenging even with positive intentions. The Health Action Process Approach (HAPA) theoretical framework bridges the intention-behavior gap by targeting planning behaviors and strengthening efficacious beliefs for behavioral change. In the current trial, an adaptive design was employed to examine differences in HAPA measures (i.e., planning and self-efficacy) regarding the target behaviors of dietary intake and physical activity (PA). Adults ≥ 21 years old with overweight or obesity and prediabetes (n = 185) received the standard Group Lifestyle Balance (GLB) intervention during the first month of treatment. Weight loss responders (lost > 2.5% of weight) at week five remained in GLB during weeks 5-16; slow responders (lost ≤ 2.5%) were stratified to the adaptive GLB Plus (GLB+) intervention during weeks 5-16. GLB+ augmented self-regulatory skills and practices consistent with HAPA. We conducted mixed model analyses with a group-by-time interaction for fixed effects at four months. GLB experienced greater improvement in behavioral intention for both diet and PA, planning behaviors (action and coping planning) for diet, and self-efficacy beliefs (action and maintenance self-efficacy) for PA compared to GLB+ (all ps < .0125). However, GLB+ also experienced statistically significant improvement in planning and self-efficacy and in energy intake and food group servings (all ps < .01). Whereas an adaptive intervention can be advantageous in improving HAPA measures and food choices, greater focus on increasing PA is needed. Additional research may help to determine effective PA strategies.

Adaptive clinical trials in surgery: A scoping review of methodological and reporting quality.

Adaptive surgical trials are scarce, but adopting these methods may help elevate the quality of surgical research when large-scale RCTs are impractical. Randomized-controlled trials (RCTs) are the gold standard for evidence-based healthcare. Despite an increase in the number of RCTs, the number of surgical trials remains unchanged. Adaptive clinical trials can streamline trial design and time to trial reporting. The advantages identified for ACTs may help to improve the quality of future surgical trials. We present a scoping review of the methodological and reporting quality of adaptive surgical trials. We performed a search of Ovid, Web of Science, and Cochrane Collaboration for all adaptive surgical RCTs performed from database inception to October 12, 2023. We included any published trials that had at least one surgical arm. All review and abstraction were performed in duplicate. Risk of bias (RoB) was assessed using the RoB 2.0 instrument and reporting quality was evaluated using CONSORT ACE 2020. All results were analyzed using descriptive methods. Of the 1338 studies identified, six trials met inclusion criteria. Trials were performed in cardiothoracic, oral, orthopedic, and urological surgery. The most common type of adaptive trial was group sequential design with pre-specified interim analyses planned for efficacy, futility, and/or sample size re-estimation. Two trials did use statistical simulations. Our risk of bias evaluation identified a high risk of bias in 50% of included trials. Reporting quality was heterogeneous regarding trial design and outcome assessment and details in relation to randomization and blinding concealment. Surgical trialists should consider implementing adaptive components to help improve patient recruitment and reduce trial duration. Reporting of future adaptive trials must adhere to existing CONSORT ACE 2020 guidelines. Future research is needed to optimize standardization of adaptive methods across medicine and surgery.

Future for cardiogenic shock research.

To discuss future research themes and study design in cardiogenic shock. Cardiogenic shock research faces multiple challenges, hindering progress in understanding and treating this life-threatening condition. Cardiogenic shock's heterogeneous nature poses challenges in patient selection for clinical trials, potentially leading to variability in treatment responses and outcomes. Ethical considerations arise due to the acuity and severity of the condition, posing challenges in obtaining informed consent and conducting randomized controlled trials where time to treatment is pivotal. This review discusses research in this area focusing on the importance of phenotyping patients with cardiogenic shock, based on artificial intelligence, machine learning, and unravel new molecular mechanisms using proteomics and metabolomics. Further, the future research focus in mechanical circulatory support and targeting inflammation is reviewed. Finally, newer trial designs including adaptive platform trials are discussed.

Does Adjunctive Clindamycin Have a Role in Staphylococcus aureus Bacteremia? A Protocol for the Adjunctive Treatment Domain of the Staphylococcus aureus Network Adaptive Platform (SNAP) Randomized Controlled Trial.

The use of adjunctive antibiotics directed against exotoxin production in Staphylococcus aureus bacteremia (SAB) is widespread, and it is recommended in many guidelines, but this is based on limited evidence. Existing guidelines are based on the theoretical premise of toxin suppression, as many strains of S. aureus produce toxins such as leukocidins (eg, Panton-Valentine leukocidin, toxic shock syndrome toxin 1, exfoliative toxins, and various enterotoxins). Many clinicians therefore believe that limiting exotoxin production release by S. aureus could reduce its virulence and improve clinical outcomes. Clindamycin, a protein synthesis inhibitor antibiotic, is commonly used for this purpose. We report the domain-specific protocol, embedded in a large adaptive, platform trial, seeking to definitively answer this question. The Staphylococcus aureus Network Adaptive Platform (SNAP) trial is a pragmatic, randomized, multicenter adaptive platform trial that aims to compare different SAB therapies, simultaneously, for 90-day mortality rates. The adjunctive treatment domain aims to test the effectiveness of adjunctive antibiotics, initially comparing clindamycin to no adjunctive antibiotic, but future adaptations may include other agents. Individuals will be randomized to receive either 5 days of adjunctive clindamycin (or lincomycin) or no adjunctive antibiotic therapy alongside standard-of-care antibiotics. Most participants with SAB (within 72 hours of index blood culture and with no contraindications) will be eligible to participate in this domain. Prespecified analyses are defined in the statistical appendix to the core protocol, and domain-specific secondary analyses will be adjusted for resistance to clindamycin, disease phenotype (complicated or uncomplicated SAB) and Panton-Valentine leukocidin-positive isolate.

An Adaptive Pragmatic Randomized Controlled Trial of Emergency Department Acupuncture for Acute Musculoskeletal Pain Management

Acute musculoskeletal pain in emergency department (ED) patients is frequently severe and challenging to treat with medications alone. The purpose of this study was to determine the feasibility, acceptability, and effectiveness of adding ED acupuncture to treat acute episodes of musculoskeletal pain in the neck, back, and extremities. In this pragmatic 2-stage adaptive open-label randomized clinical trial, Stage 1 identified whether auricular acupuncture (AA; based on the battlefield acupuncture protocol) or peripheral acupuncture (PA; needles in head, neck, and extremities only), when added to usual care was more feasible, acceptable, and efficacious in the ED. Stage 2 assessed effectiveness of the selected acupuncture intervention(s) on pain reduction compared to usual care only (UC). Licensed acupuncturists delivered AA and PA. They saw and evaluated but did not deliver acupuncture to the UC group as an attention control. All participants received UC from blinded ED providers. Primary outcome was 1-hour change in 11-point pain numeric rating scale. Stage 1 interim analysis found both acupuncture styles similar, so Stage 2 continued all 3 treatment arms. Among 236 participants randomized, demographics and baseline pain were comparable across groups. When compared to UC alone, reduction in pain was 1.6 (95% confidence interval [CI]: 0.7 to 2.6) points greater for AA+UC and 1.2 (95% CI: 0.3 to 2.1) points greater for PA+UC patients. Participants in both treatment arms reported high satisfaction with acupuncture. ED acupuncture is feasible and acceptable and can reduce acute musculoskeletal pain better than UC alone.

A comparison of alternative ranking methods in two-stage clinical trials with multiple interventions: An application to the anxiolysis for laceration repair in children trial.

Multi-arm, multi-stage trials frequently include a standard care to which all interventions are compared. This may increase costs and hinders comparisons among the experimental arms. Furthermore, the standard care may not be evident, particularly when there is a large variation in standard practice. Thus, we aimed to develop an adaptive clinical trial that drops ineffective interventions following an interim analysis before selecting the best intervention at the final stage without requiring a standard care. We used Bayesian methods to develop a multi-arm, two-stage adaptive trial and evaluated two different methods for ranking interventions, the probability that each intervention was optimal (Pbest) and using the surface under the cumulative ranking curve (SUCRA), at both the interim and final analysis. The proposed trial design determines the maximum sample size for each intervention using the Average Length Criteria. The interim analysis takes place at approximately half the pre-specified maximum sample size and aims to drop interventions for futility if either Pbest or the SUCRA is below a pre-specified threshold. The final analysis compares all remaining interventions at the maximum sample size to conclude superiority based on either Pbest or the SUCRA. The two ranking methods were compared across 12 scenarios that vary the number of interventions and the assumed differences between the interventions. The thresholds for futility and superiority were chosen to control type 1 error, and then the predictive power and expected sample size were evaluated across scenarios. A trial comparing three interventions that aim to reduce anxiety for children undergoing a laceration repair in the emergency department was then designed, known as the Anxiolysis for Laceration Repair in Children Trial (ALICE) trial. As the number of interventions increases, the SUCRA results in a higher predictive power compared with Pbest. Using Pbest results in a lower expected sample size when there is an effective intervention. Using the Average Length Criterion, the ALICE trial has a maximum sample size for each arm of 100 patients. This sample size results in a 86% and 85% predictive power using Pbest and the SUCRA, respectively. Thus, we chose Pbest as the ranking method for the ALICE trial. Bayesian ranking methods can be used in multi-arm, multi-stage trials with no clear control intervention. When more interventions are included, the SUCRA results in a higher power than Pbest. Future work should consider whether other ranking methods may also be relevant for clinical trial design.

Empirical meropenem versus piperacillin/tazobactam for adult patients with sepsis (EMPRESS) trial: Protocol.

Piperacillin/tazobactam may be associated with less favourable outcomes than carbapenems in patients with severe bacterial infections, but the certainty of evidence is low. The Empirical Meropenem versus Piperacillin/Tazobactam for Adult Patients with Sepsis (EMPRESS) trial is an investigator-initiated, international, parallel-group, randomised, open-label, adaptive clinical trial with an integrated feasibility phase. We will randomise adult, critically ill patients with sepsis to empirical treatment with meropenem or piperacillin/tazobactam for up to 30 days. The primary outcome is 30-day all-cause mortality. The secondary outcomes are serious adverse reactions within 30 days; isolation precautions due to resistant bacteria within 30 days; days alive without life support and days alive and out of hospital within 30 and 90 days; 90- and 180-day all-cause mortality and 180-day health-related quality of life. EMPRESS will use Bayesian statistical models with weak to somewhat sceptical neutral priors. Adaptive analyses will be conducted after follow-up of the primary outcome for the first 400 participants concludes and after every 300 subsequent participants, with adaptive stopping for superiority/inferiority and practical equivalence (absolute risk difference <2.5%-points) and response-adaptive randomisation. The expected sample sizes in scenarios with no, small or large differences are 5189, 5859 and 2570 participants, with maximum 14,000 participants and ≥99% probability of conclusiveness across all scenarios. EMPRESS will compare the effects of empirical meropenem against piperacillin/tazobactam in adult, critically ill patients with sepsis. Due to the pragmatic, adaptive design with high probability of conclusiveness, the trial results are expected to directly inform clinical practice.

A comparison of computational algorithms for the Bayesian analysis of clinical trials.

Clinical trials are increasingly using Bayesian methods for their design and analysis. Inference in Bayesian trials typically uses simulation-based approaches such as Markov Chain Monte Carlo methods. Markov Chain Monte Carlo has high computational cost and can be complex to implement. The Integrated Nested Laplace Approximations algorithm provides approximate Bayesian inference without the need for computationally complex simulations, making it more efficient than Markov Chain Monte Carlo. The practical properties of Integrated Nested Laplace Approximations compared to Markov Chain Monte Carlo have not been considered for clinical trials. Using data from a published clinical trial, we aim to investigate whether Integrated Nested Laplace Approximations is a feasible and accurate alternative to Markov Chain Monte Carlo and provide practical guidance for trialists interested in Bayesian trial design. Data from an international Bayesian multi-platform adaptive trial that compared therapeutic-dose anticoagulation with heparin to usual care in non-critically ill patients hospitalized for COVID-19 were used to fit Bayesian hierarchical generalized mixed models. Integrated Nested Laplace Approximations was compared to two Markov Chain Monte Carlo algorithms, implemented in the software JAGS and stan, using packages available in the statistical software R. Seven outcomes were analysed: organ-support free days (an ordinal outcome), five binary outcomes related to survival and length of hospital stay, and a time-to-event outcome. The posterior distributions for the treatment and sex effects and the variances for the hierarchical effects of age, site and time period were obtained. We summarized these posteriors by calculating the mean, standard deviations and the 95% equitailed credible intervals and presenting the results graphically. The computation time for each algorithm was recorded. The average overlap of the 95% credible interval for the treatment and sex effects estimated using Integrated Nested Laplace Approximations was 96% and 97.6% compared with stan, respectively. The graphical posterior densities for these effects overlapped for all three algorithms. The posterior mean for the variance of the hierarchical effects of age, site and time estimated using Integrated Nested Laplace Approximations are within the 95% credible interval estimated using Markov Chain Monte Carlo but the average overlap of the credible interval is lower, 77%, 85.6% and 91.3%, respectively, for Integrated Nested Laplace Approximations compared to stan. Integrated Nested Laplace Approximations and stan were easily implemented in clear, well-established packages in R, while JAGS required the direct specification of the model. Integrated Nested Laplace Approximations was between 85 and 269 times faster than stan and 26 and 1852 times faster than JAGS. Integrated Nested Laplace Approximations could reduce the computational complexity of Bayesian analysis in clinical trials as it is easy to implement in R, substantially faster than Markov Chain Monte Carlo methods implemented in JAGS and stan, and provides near identical approximations to the posterior distributions for the treatment effect. Integrated Nested Laplace Approximations was less accurate when estimating the posterior distribution for the variance of hierarchical effects, particularly for the proportional odds model, and future work should determine if the Integrated Nested Laplace Approximations algorithm can be adjusted to improve this estimation.

Heterogeneity, Bayesian thinking, and phenotyping in critical care: A primer.

To familiarize clinicians with the emerging concepts in critical care research of Bayesian thinking and personalized medicine through phenotyping and explain their clinical relevance by highlighting how they address the issues of frequent negative trials and heterogeneity of treatment effect. The past decades have seen many negative (effect-neutral) critical care trials of promising interventions, culminating in calls to improve the field's research through adopting Bayesian thinking and increasing personalization of critical care medicine through phenotyping. Bayesian analyses add interpretive power for clinicians as they summarize treatment effects based on probabilities of benefit or harm, contrasting with conventional frequentist statistics that either affirm or reject a null hypothesis. Critical care trials are beginning to include prospective Bayesian analyses, and many trials have undergone reanalysis with Bayesian methods. Phenotyping seeks to identify treatable traits to target interventions to patients expected to derive benefit. Phenotyping and subphenotyping have gained prominence in the most syndromic and heterogenous critical care disease states, acute respiratory distress syndrome and sepsis. Grouping of patients has been informative across a spectrum of clinically observable physiological parameters, biomarkers, and genomic data. Bayesian thinking and phenotyping are emerging as elements of adaptive clinical trials and predictive enrichment, paving the way for a new era of high-quality evidence. These concepts share a common goal, sifting through the noise of heterogeneity in critical care to increase the value of existing and future research. The future of critical care medicine will inevitably involve modification of statistical methods through Bayesian analyses and targeted therapeutics via phenotyping. Clinicians must be familiar with these systems that support recommendations to improve decision-making in the gray areas of critical care practice.

Adaptive designs in public health: Vaccine and cluster randomized trials go Bayesian.

Clinical trials in public health-particularly those conducted in low- and middle-income countries-often involve communicable and non-communicable diseases with high disease burden and unmet needs. Trials conducted in these regions often are faced with resource limitations, so improving the efficiencies of these trials is critical. Adaptive trial designs have the potential to save trial time and resources and reduce the number of patients receiving ineffective interventions. In this paper, we provide a detailed account of the implementation of vaccine and cluster randomized trials within the framework of Bayesian adaptive trials, with emphasis on computational efficiency and flexibility with regard to stopping rules and allocation ratios. We offer an educated approach to selecting prior distributions and a data-driven empirical Bayes method for plug-in estimates for nuisance parameters.

Designing a Bayesian adaptive clinical trial to evaluate novel mechanical ventilation strategies in acute respiratory failure using integrated nested Laplace approximations

BackgroundAdaptive trials usually require simulations to determine values for design parameters, demonstrate error rates, and establish the sample size. We designed a Bayesian adaptive trial comparing ventilation strategies for patients with acute hypoxemic respiratory failure using simulations. The complexity of the analysis would usually require computationally expensive Markov Chain Monte Carlo methods but this barrier to simulation was overcome using the Integrated Nested Laplace Approximations (INLA) algorithm to provide fast, approximate Bayesian inference. MethodsWe simulated two-arm Bayesian adaptive trials with equal randomization that stratified participants into two disease severity states. The analysis used a proportional odds model, fit using INLA. Trials were stopped based on pre-specified posterior probability thresholds for superiority or futility, separately for each state. We calculated the type I error and power across 64 scenarios that varied the probability thresholds and the initial minimum sample size before commencing adaptive analyses. Two designs that maintained a type I error below 5%, a power above 80%, and a feasible mean sample size were evaluated further to determine the optimal design. ResultsPower generally increased as the initial sample size and the futility threshold increased. The chosen design had an initial recruitment of 500 and a superiority threshold of 0.9925, and futility threshold of 0.95. It maintained high power and was likely to reach a conclusion before exceeding a feasible sample size. ConclusionsWe designed a Bayesian adaptive trial to evaluate novel strategies for ventilation using the INLA algorithm to efficiently evaluate a wide range of designs through simulation.