Value of Information for Clinical Trial Design: The Importance of Considering All Relevant Comparators.

Value of Information (VOI) analyses calculate the economic value that could be generated by obtaining further information to reduce uncertainty in a health economic decision model. VOI has been suggested as a tool for research prioritisation and trial design as it can highlight economically valuable avenues for future research. Recent methodological advances have made it increasingly feasible to use VOI in practice for research; however, there are critical differences between the VOI approach and the standard methods used to design research studies such as clinical trials. We aimed to highlight key differences between the research design approach based on VOI and standard clinical trial design methods, in particular the importance of considering the full decision context. We present two hypothetical examples to demonstrate that VOI methods are only accurate when (1) all feasible comparators are included in the decision model when designing research, and (2) all comparators are retained in the decision model once the data have been collected and a final treatment recommendation is made. Omitting comparators from either the design or analysis phase of research when using VOI methods can lead to incorrect trial designs and/or treatment recommendations. Overall, we conclude that incorrectly specifying the health economic model by ignoring potential comparators can lead to misleading VOI results and potentially waste scarce research resources.

Barriers and Facilitators of Using R for Decision Analytic Modeling in Health Technology Assessment: Focus Group Results.

BACKGROUND AND OBJECTIVE: Decision models for health technology assessment (HTA) are largely submitted to HTA agencies using commercial software, which has known limitations. The use of the open-source programming language R has been suggested because of its efficiency, transparency, reproducibility, and ability to consider complex analyses. However, its use in HTA remains limited. This qualitative study aimed to explore the main reasons for this slow uptake of R in HTA and identify tangible facilitators. METHODS: We undertook two semi-structured focus group discussions with 24 key stakeholders from government agencies, consultancy, pharmaceutical companies, and academia. Two 1.5-hour discussions reflected on barriers identified in a previous study and highlighted additional barriers. Discussions were recorded and semi-transcribed, and data were organized and summarized into key themes. RESULTS: Human resources constraints were identified as a key barrier, including a lack of training, prioritization and collaboration, and resistance to change. Another key barrier was the lack of acceptance, or clear guidance, around submissions in R by HTA agencies. Participants also highlighted a lack of communication around accepted packages and decision model structures, and between HTA agencies on standard decision modeling structures. CONCLUSIONS: There is a need for standardization, which can facilitate decision model sharing, coding homogeneity, and improved country adaptations. The creation of training materials and tailored workshops was identified as a key short-term facilitator. Increased communication and engagement of stakeholders could also facilitate the use of R by identifying needs and opportunities, encouraging HTA agencies to address structural barriers, and increasing incentives to use R.

Model-based standardization using multiple imputation.

BACKGROUND: When studying the association between treatment and a clinical outcome, a parametric multivariable model of the conditional outcome expectation is often used to adjust for covariates. The treatment coefficient of the outcome model targets a conditional treatment effect. Model-based standardization is typically applied to average the model predictions over the target covariate distribution, and generate a covariate-adjusted estimate of the marginal treatment effect. METHODS: The standard approach to model-based standardization involves maximum-likelihood estimation and use of the non-parametric bootstrap. We introduce a novel, general-purpose, model-based standardization method based on multiple imputation that is easily applicable when the outcome model is a generalized linear model. We term our proposed approach multiple imputation marginalization (MIM). MIM consists of two main stages: the generation of synthetic datasets and their analysis. MIM accommodates a Bayesian statistical framework, which naturally allows for the principled propagation of uncertainty, integrates the analysis into a probabilistic framework, and allows for the incorporation of prior evidence. RESULTS: We conduct a simulation study to benchmark the finite-sample performance of MIM in conjunction with a parametric outcome model. The simulations provide proof-of-principle in scenarios with binary outcomes, continuous-valued covariates, a logistic outcome model and the marginal log odds ratio as the target effect measure. When parametric modeling assumptions hold, MIM yields unbiased estimation in the target covariate distribution, valid coverage rates, and similar precision and efficiency than the standard approach to model-based standardization. CONCLUSION: We demonstrate that multiple imputation can be used to marginalize over a target covariate distribution, providing appropriate inference with a correctly specified parametric outcome model and offering statistical performance comparable to that of the standard approach to model-based standardization.

Pharmacologic Procedural Distress Management During Laceration Repair in Children: A Systematic Review.

OBJECTIVES: To systematically appraise the literature on the relative effectiveness of pharmacologic procedural distress management agents for children undergoing laceration repair. METHODS: Six databases were searched in August 2021, and the search was updated in January 2023. We included completed randomized or quasi-randomized trials involving ( a ) children younger than 15 years undergoing laceration repair in the emergency department; ( b ) randomization to at least one anxiolytic, sedative, and/or analgesic agent versus any comparator agent or placebo; ( c ) efficacy of procedural distress management measured on any scale. Secondary outcomes were pain during the procedure, administration acceptance, sedation duration, additional sedation, length of stay, and stakeholder satisfaction. Cochrane Collaboration's risk-of-bias tool assessed individual studies. Ranges and proportions summarized results where applicable. RESULTS: Among 21 trials (n = 1621 participants), the most commonly studied anxiolytic agents were midazolam, ketamine, and N 2 O. Oral midazolam, oral ketamine, and N 2 O were found to reduce procedural distress more effectively than their comparators in 4, 3, and 2 studies, respectively. Eight studies comparing routes, doses, or volumes of administration of the same agent led to indeterminate results. Meta-analysis was not performed because of heterogeneity in comparators, routes, and outcome measures across studies. CONCLUSIONS: Based on procedural distress reduction, this study favors oral midazolam and oral ketamine. However, this finding should be interpreted with caution because of heterogeneous comparators across studies and minor conflicting results. An optimal agent for procedural distress management cannot be recommended based on the limited evidence. Future research should seek to identify the minimal, essential measures of patient distress during pharmacologic anxiolysis and/or sedation in laceration repair to guide future trials and reviews.

Effects of inflammation on thrombosis and outcomes in COVID-19: secondary analysis of the ATTACC/ACTIV-4a trial.

Background: Patients hospitalized for COVID-19 are at high risk of thrombotic complications and organ failure, and often exhibit severe inflammation, which may contribute to hypercoagulability. Objectives: To determine whether patients hospitalized for COVID-19 experience differing frequencies of thrombotic and organ failure complications and derive variable benefits from therapeutic-dose heparin dependent on the extent of systemic inflammation and whether observed benefit from therapeutic-dose anticoagulation varies depending on the degree of systemic inflammation. Methods: We analyzed data from 1346 patients hospitalized for COVID-19 enrolled in the ATTACC and ACTIV-4a platforms who were randomized to therapeutic-dose heparin or usual care for whom levels of C-reactive protein (CRP) were reported at baseline. Results: Increased CRP was associated with worse patient outcomes, including a >98% posterior probability of increased organ support requirement, hospital length of stay, risk of 28-day mortality, and incidence of major thrombotic events or death (patients with CRP 40-100 mg/L or ≥100 mg/L compared to patients with CRP <40 mg/L). Patients with CRP 40 to 100 mg/L experienced the greatest degree of benefit from treatment with therapeutic doses of unfractionated or low molecular weight heparin compared with usual-care prophylactic doses. This was most significant for an increase in organ support-free days (odds ratio: 1.63; 95% confidence interval, 1.09-2.40; 97.9% posterior probability of beneficial effect), with trends toward benefit for other evaluated outcomes. Conclusion: Moderately ill patients hospitalized for COVID-19 with CRP between 40 mg/L and 100 mg/L derived the greatest benefit from treatment with therapeutic-dose heparin.

Effect of therapeutic-dose heparin on severe acute kidney injury and death in noncritically ill patients hospitalized for COVID-19: a prespecified secondary analysis of the ACTIV4a and ATTACC randomized trial.

Background: Acute kidney injury (AKI) in patients with COVID-19 is partly mediated by thromboinflammation. In noncritically ill patients with COVID-19, therapeutic-dose anticoagulation with heparin increased the probability of survival to hospital discharge with reduced use of cardiovascular or respiratory organ support. Objectives: We investigated whether therapeutic-dose heparin reduces the incidence of AKI or death in noncritically ill patients hospitalized for COVID-19. Methods: We report a prespecified secondary analysis of the ACTIV4a and ATTACC open-label, multiplatform randomized trial of therapeutic-dose heparin vs usual-care pharmacologic thromboprophylaxis on the incidence of severe AKI (≥2-fold increase in serum creatinine or initiation of kidney replacement therapy (KDIGO stage 2 or 3) or all-cause mortality in noncritically ill patients hospitalized for COVID-19. Bayesian statistical models were adjusted for age, sex, D-dimer, enrollment period, country, site, and platform. Results: Among 1922 enrolled, 23 were excluded due to pre-existing end stage kidney disease and 205 were missing baseline or follow-up creatinine measurements. Severe AKI or death occurred in 4.4% participants assigned to therapeutic-dose heparin and 5.5% assigned to thromboprophylaxis (adjusted relative risk [aRR]: 0.72; 95% credible interval (CrI): 0.47, 1.10); the posterior probability of superiority for therapeutic-dose heparin (relative risk < 1.0) was 93.6%. Therapeutic-dose heparin was associated with a 97.7% probability of superiority to reduce the composite of stage 3 AKI or death (3.1% vs 4.6%; aRR: 0.64; 95% CrI: 0.40, 0.99) compared to thromboprophylaxis. Conclusion: Therapeutic-dose heparin was associated with a high probability of superiority to reduce the incidence of in-hospital severe AKI or death in patients hospitalized for COVID-19.

Consolidated Health Economic Evaluation Reporting Standards - Value of Information (CHEERS-VOI): Explanation and Elaboration.

OBJECTIVES: Although the ISPOR Value of Information (VOI) Task Force's reports outline VOI concepts and provide good-practice recommendations, there is no guidance for reporting VOI analyses. VOI analyses are usually performed alongside economic evaluations for which the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) 2022 Statement provides reporting guidelines. Thus, we developed the CHEERS-VOI checklist to provide reporting guidance and checklist to support the transparent, reproducible, and high-quality reporting of VOI analyses. METHODS: A comprehensive literature review generated a list of 26 candidate reporting items. These candidate items underwent a Delphi procedure with Delphi participants through 3 survey rounds. Participants rated each item on a 9-point Likert scale to indicate its relevance when reporting the minimal, essential information about VOI methods and provided comments. The Delphi results were reviewed at 2-day consensus meetings and the checklist was finalized using anonymous voting. RESULTS: We had 30, 25, and 24 Delphi respondents in rounds 1, 2, and 3, respectively. After incorporating revisions recommended by the Delphi participants, all 26 candidate items proceeded to the 2-day consensus meetings. The final CHEERS-VOI checklist includes all CHEERS items, but 7 items require elaboration when reporting VOI. Further, 6 new items were added to report information relevant only to VOI (eg, VOI methods applied). CONCLUSIONS: The CHEERS-VOI checklist should be used when a VOI analysis is performed alongside economic evaluations. The CHEERS-VOI checklist will help decision makers, analysts and peer reviewers in the assessment and interpretation of VOI analyses and thereby increase transparency and rigor in decision making.

Methodological considerations for novel approaches to covariate-adjusted indirect treatment comparisons.

We examine four important considerations in the development of covariate adjustment methodologies for indirect treatment comparisons. First, we consider potential advantages of weighting versus outcome modeling, placing focus on bias-robustness. Second, we outline why model-based extrapolation may be required and useful, in the specific context of indirect treatment comparisons with limited overlap. Third, we describe challenges for covariate adjustment based on data-adaptive outcome modeling. Finally, we offer further perspectives on the promise of doubly robust covariate adjustment frameworks.

Effectiveness of Structured Care Coordination for Children With Medical Complexity: The Complex Care for Kids Ontario (CCKO) Randomized Clinical Trial.

Importance: Children with medical complexity (CMC) have chronic conditions and high health needs and may experience fragmented care. Objective: To compare the effectiveness of a structured complex care program, Complex Care for Kids Ontario (CCKO), with usual care. Design, Setting, and Participants: This randomized clinical trial used a waitlist variation for randomizing patients from 12 complex care clinics in Ontario, Canada, over 2 years. The study was conducted from December 2016 to June 2021. Participants were identified based on complex care clinic referral and randomly allocated into an intervention group, seen at the next available clinic appointment, or a control group that was placed on a waitlist to receive the intervention after 12 months. Intervention: Assignment of a nurse practitioner-pediatrician dyad partnering with families in a structured complex care clinic to provide intensive care coordination and comprehensive plans of care. Main Outcomes and Measures: Co-primary outcomes, assessed at baseline and at 6, 12, and 24 months postrandomization, were service delivery indicators from the Family Experiences With Coordination of Care that scored (1) coordination of care among health care professionals, (2) coordination of care between health care professionals and families, and (3) utility of care planning tools. Secondary outcomes included child and parent health outcomes and child health care system utilization and cost. Results: Of 144 participants randomized, 141 had complete health administrative data, and 139 had complete baseline surveys. The median (IQR) age of the participants was 29 months (9-102); 83 (60%) were male. At 12 months, scores for utility of care planning tools improved in the intervention group compared with the waitlist group (adjusted odds ratio, 9.3; 95% CI, 3.9-21.9; P < .001), with no difference between groups for the other 2 co-primary outcomes. There were no group differences for secondary outcomes of child outcomes, parent outcomes, and health care system utilization and cost. At 24 months, when both groups were receiving the intervention, no primary outcome differences were observed. Total health care costs in the second year were lower for the intervention group (median, CAD$17 891; IQR, 6098-61 346; vs CAD$37 524; IQR, 9338-119 547 [US $13 415; IQR, 4572-45 998; vs US $28 136; IQR, 7002-89 637]; P = .01). Conclusions and Relevance: The CCKO program improved the perceived utility of care planning tools but not other outcomes at 1 year. Extended evaluation periods may be helpful in assessing pediatric complex care interventions. Trial Registration: ClinicalTrials.gov Identifier: NCT02928757.

General-Purpose Methods for Simulating Survival Data for Expected Value of Sample Information Calculations.

BACKGROUND: Expected value of sample information (EVSI) quantifies the expected value to a decision maker of reducing uncertainty by collecting additional data. EVSI calculations require simulating plausible data sets, typically achieved by evaluating quantile functions at random uniform numbers using standard inverse transform sampling (ITS). This is straightforward when closed-form expressions for the quantile function are available, such as for standard parametric survival models, but these are often unavailable when assuming treatment effect waning and for flexible survival models. In these circumstances, the standard ITS method could be implemented by numerically evaluating the quantile functions at each iteration in a probabilistic analysis, but this greatly increases the computational burden. Thus, our study aims to develop general-purpose methods that standardize and reduce the computational burden of the EVSI data-simulation step for survival data. METHODS: We developed a discrete sampling method and an interpolated ITS method for simulating survival data from a probabilistic sample of survival probabilities over discrete time units. We compared the general-purpose and standard ITS methods using an illustrative partitioned survival model with and without adjustment for treatment effect waning. RESULTS: The discrete sampling and interpolated ITS methods agree closely with the standard ITS method, with the added benefit of a greatly reduced computational cost in the scenario with adjustment for treatment effect waning. CONCLUSIONS: We present general-purpose methods for simulating survival data from a probabilistic sample of survival probabilities that greatly reduce the computational burden of the EVSI data-simulation step when we assume treatment effect waning or use flexible survival models. The implementation of our data-simulation methods is identical across all possible survival models and can easily be automated from standard probabilistic decision analyses. HIGHLIGHTS: Expected value of sample information (EVSI) quantifies the expected value to a decision maker of reducing uncertainty through a given data collection exercise, such as a randomized clinical trial. In this article, we address the problem of computing EVSI when we assume treatment effect waning or use flexible survival models, by developing general-purpose methods that standardize and reduce the computational burden of the EVSI data-generation step for survival data.We developed 2 methods for simulating survival data from a probabilistic sample of survival probabilities over discrete time units, a discrete sampling method and an interpolated inverse transform sampling method, which can be combined with a recently proposed nonparametric EVSI method to accurately estimate EVSI for collecting survival data.Our general-purpose data-simulation methods greatly reduce the computational burden of the EVSI data-simulation step when we assume treatment effect waning or use flexible survival models. The implementation of our data-simulation methods is identical across all possible survival models and can therefore easily be automated from standard probabilistic decision analyses.

Optimal Dose of Intranasal Dexmedetomidine for Laceration Repair in Children: A Phase II Dose-Ranging Study.

STUDY OBJECTIVE: To determine the optimal sedative dose of intranasal dexmedetomidine for children undergoing laceration repair. METHODS: This dose-ranging study employing the Bayesian Continual Reassessment Method enrolled children aged 0 to 10 years with a single laceration (<5 cm), requiring single-layer closure, who received topical anesthetic. Children were administered 1, 2, 3, or 4 mcg/kg intranasal dexmedetomidine. The primary outcome was the proportion with adequate sedation (Pediatric Sedation State Scale score of 2 or 3 for ≥90% of the time from sterile preparation to tying of the last suture). Secondary outcomes included the Observational Scale of Behavior Distress-Revised (range: 0 [no distress] to 23.5 [maximal distress]), postprocedure length of stay, and adverse events. RESULTS: We enrolled 55 children (35/55 [64%] males; median [interquartile range {IQR}] age 4 [2, 6] years). At 1, 2, 3, and 4 mcg/kg intranasal dexmedetomidine, respectively, the proportion of participants "adequately" sedated was 1/3 (33%), 2/9 (22%), 13/21 (62%), and 12/21 (57%); the posterior mean (95% equitailed credible intervals) for the probability of adequate sedation was 0.38 (0.04, 0.82), 0.25 (0.05, 0.54), 0.61 (0.41, 0.80), and 0.57 (0.36, 0.76); the median (IQR) Observational Scale of Behavior Distress-Revised scores during suturing was 2.7 (0.3, 3), 0 (0, 3.8), 0.6 (0, 5), and 0 (0, 3.7); the median (IQR) postprocedure length of stay was 67 (60, 78), 76 (60, 100), 89 (76, 109), and 113 (76, 150) minutes. There was 1 adverse event, a decrease in oxygen saturation at 4 mcg/kg, which resolved with head repositioning. CONCLUSION: Despite limitations, such as our limited sample size and subjectivity in Pediatric Sedation State Scale scoring, sedation efficacy for 3 and 4 mcg/kg were similarly based on equitailed credible intervals suggesting either could be considered optimal.

Methodology and design of platform trials: a meta-epidemiological study.

OBJECTIVES: Adaptive platforms allow for the evaluation of multiple interventions at a lower cost and have been growing in popularity, especially during the COVID-19 pandemic. The objective of this review is to summarize published platform trials, examine specific methodological design features among these studies, and hopefully aid readers in the evaluation and interpretation of platform trial results. METHODS: We performed a systematic review of EMBASE, MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL), and clinicaltrials.gov from January 2015 to January 2022 for protocols or results of platform trials. Pairs of reviewers, working independently and in duplicate, collected data on trial characteristics of trial registrations, protocols, and publications of platform trials. We reported our results using total numbers and percentages, as well as medians with interquartile range (IQR) when appropriate. RESULTS: We identified 15,277 unique search records and screened 14,403 titles and abstracts after duplicates were removed. We identified 98 unique randomized platform trials. Sixteen platform trials were sourced from a systematic review completed in 2019, which included platform trials reported prior to 2015. Most platform trials (n = 67, 68.3%) were registered between 2020 and 2022, coinciding with the COVID-19 pandemic. The included platform trials primarily recruited or plan to recruit patients from North America or Europe, with most subjects being recruited from the United States (n = 39, 39.7%) and the United Kingdom (n = 31, 31.6%). Bayesian methods were used in 28.6% (n = 28) of platform RCTs and frequentist methods in 66.3% (n = 65) of trials, including 1 (1%) that used methods from both paradigms. Out of the twenty-five trials with peer-reviewed publication of results, seven trials used Bayesian methods (28%), and of those, two (8%) used a predefined sample size calculation while the remainder used pre-specified probabilities of futility, harm, or benefit calculated at (pre-specified) intervals to inform decisions about stopping interventions or the entire trial. Seventeen (68%) peer-reviewed publications used frequentist methods. Out of the seven published Bayesian trials, seven (100%) reported thresholds for benefit. The threshold for benefit ranged from 80% to >99%. CONCLUSION: We identified and summarized key components of platform trials, including the basics of the methodological and statistical considerations. Ultimately, improving standardization and reporting in platform trials require an understanding of the current landscape. We provide the most updated and rigorous review of platform trials to date.

HEROIC Trials to Answer Pragmatic Questions for Hospitalized Children.

Although the number of randomized controlled trials (RCTs) published each year involving adult populations is steadily rising, the annual number of RCTs published involving pediatric populations has not changed since 2005. Barriers to the broader utilization of RCTs in pediatrics include a lower prevalence of disease, less available funding, and more complicated regulatory requirements. Although child health researchers have been successful in overcoming these barriers for isolated diseases such as pediatric cancer, common pediatric diseases are underrepresented in RCTs relative to their burden. This article proposes a strategy called High-Efficiency RandOmIzed Controlled (HEROIC) trials to increase RCTs focused on common diseases among hospitalized children. HEROIC trials are multicenter RCTs that pursue the rapid, low-cost accumulation of study participants with minimal burden for individual sites. Five key strategies distinguish HEROIC trials: (1) dispersed low-volume recruitment, in which a large number of sites (50-150 hospitals) enroll a small number of participants per site (2-10 participants per site), (2) incentivizing site leads with authorship, training, education credits, and modest financial support, (3) a focus on pragmatic questions that examine simple, widely used interventions, (4) the use of a single institutional review board, integrated consent, and other efficient solutions to regulatory requirements, and (5) scaling the HEROIC trial strategy to accomplish multiple trials simultaneously. HEROIC trials can boost RCT feasibility and volume to answer fundamental clinical questions and improve care for hospitalized children.

Economic Evaluation of Cost and Time Required for a Platform Trial vs Conventional Trials.

Importance: Platform trial design allows the introduction of new interventions after the trial is initiated and offers efficiencies to clinical research. However, limited guidance exists on the economic resources required to establish and maintain platform trials. Objective: To compare cost (US dollars) and time requirements of conducting a platform trial vs a series of conventional (nonplatform) trials using a real-life example. Design, Setting, and Participants: For this economic evaluation, an online survey was administered to a group of international experts (146 participants) with publication records of platform trials to elicit their opinions on cost and time to set up and conduct platform, multigroup, and 2-group trials. Using the reported entry dates of 10 interventions into Systemic Therapy in Advancing Metastatic Prostate Cancer: Evaluation of Drug Efficacy, the longest ongoing platform trial, 3 scenarios were designed involving a single platform trial (scenario 1), 1 multigroup followed by 5 2-group trials (scenario 2), and a series of 10 2-group trials (scenario 3). All scenarios started with 5 interventions, then 5 more interventions were either added to the platform or evaluated independently. Simulations with the survey results as inputs were used to compare the platform vs conventional trial designs. Data were analyzed from July to September 2021. Exposure: Platform trial design. Main Outcomes and Measures: Total trial setup and conduct cost and cumulative duration. Results: Although setup time and cost requirements of a single trial were highest for the platform trial, cumulative requirements of setting up a series of multiple trials in scenarios 2 and 3 were larger. Compared with the platform trial, there was a median (IQR) increase of 216.7% (202.2%-242.5%) in cumulative setup costs for scenario 2 and 391.1% (365.3%-437.9%) for scenario 3. In terms of total cost, there was a median (IQR) increase of 17.4% (12.1%-22.5%) for scenario 2 and 57.5% (43.1%-69.9%) for scenario 3. There was a median (IQR) increase in cumulative trial duration of 171.1% (158.3%-184.3%) for scenario 2 and 311.9% (282.0%-349.1%) for scenario 3. Cost and time reductions in the platform trial were observed in both the initial and subsequently evaluated interventions. Conclusions and Relevance: Although setting up platform trials can take longer and be costly, the findings of this study suggest that having a single infrastructure can improve efficiencies with respect to costs and efforts.

Parametric G-computation for compatible indirect treatment comparisons with limited individual patient data.

Population adjustment methods such as matching-adjusted indirect comparison (MAIC) are increasingly used to compare marginal treatment effects when there are cross-trial differences in effect modifiers and limited patient-level data. MAIC is based on propensity score weighting, which is sensitive to poor covariate overlap and cannot extrapolate beyond the observed covariate space. Current outcome regression-based alternatives can extrapolate but target a conditional treatment effect that is incompatible in the indirect comparison. When adjusting for covariates, one must integrate or average the conditional estimate over the relevant population to recover a compatible marginal treatment effect. We propose a marginalization method based on parametric G-computation that can be easily applied where the outcome regression is a generalized linear model or a Cox model. The approach views the covariate adjustment regression as a nuisance model and separates its estimation from the evaluation of the marginal treatment effect of interest. The method can accommodate a Bayesian statistical framework, which naturally integrates the analysis into a probabilistic framework. A simulation study provides proof-of-principle and benchmarks the method's performance against MAIC and the conventional outcome regression. Parametric G-computation achieves more precise and more accurate estimates than MAIC, particularly when covariate overlap is poor, and yields unbiased marginal treatment effect estimates under no failures of assumptions. Furthermore, the marginalized regression-adjusted estimates provide greater precision and accuracy than the conditional estimates produced by the conventional outcome regression, which are systematically biased because the measure of effect is non-collapsible.

Value of Information Analysis in Models to Inform Health Policy.

Value of information (VoI) is a decision-theoretic approach to estimating the expected benefits from collecting further information of different kinds, in scientific problems based on combining one or more sources of data. VoI methods can assess the sensitivity of models to different sources of uncertainty and help to set priorities for further data collection. They have been widely applied in healthcare policy making, but the ideas are general to a range of evidence synthesis and decision problems. This article gives a broad overview of VoI methods, explaining the principles behind them, the range of problems that can be tackled with them, and how they can be implemented, and discusses the ongoing challenges in the area.

Remote, real-time expert elicitation to determine the prior probability distribution for Bayesian sample size determination in international randomised controlled trials: Bronchiolitis in Infants Placebo Versus Epinephrine and Dexamethasone (BIPED) study.

BACKGROUND: Bayesian methods are increasing in popularity in clinical research. The design of Bayesian clinical trials requires a prior distribution, which can be elicited from experts. In diseases with international differences in management, the elicitation exercise should recruit internationally, making a face-to-face elicitation session expensive and more logistically challenging. Thus, we used a remote, real-time elicitation exercise to construct prior distributions. These elicited distributions were then used to determine the sample size of the Bronchiolitis in Infants with Placebo Versus Epinephrine and Dexamethasone (BIPED) study, an international randomised controlled trial in the Pediatric Emergency Research Network (PERN). The BIPED study aims to determine whether the combination of epinephrine and dexamethasone, compared to placebo, is effective in reducing hospital admission for infants presenting with bronchiolitis to the emergency department. METHODS: We developed a Web-based tool to support the elicitation of the probability of hospitalisation for infants with bronchiolitis. Experts participated in online workshops to specify their individual prior distributions, which were aggregated using the equal-weighted linear pooling method. Experts were then invited to provide their comments on the aggregated distribution. The average length criterion determined the BIPED sample size. RESULTS: Fifteen paediatric emergency medicine clinicians from Canada, the USA, Australia and New Zealand participated in three workshops to provide their elicited prior distributions. The mean elicited probability of admission for infants with bronchiolitis was slightly lower for those receiving epinephrine and dexamethasone compared to supportive care in the aggregate distribution. There were substantial differences in the individual beliefs but limited differences between North America and Australasia. From this aggregate distribution, a sample size of 410 patients per arm results in an average 95% credible interval length of less than 9% and a relative predictive power of 90%. CONCLUSION: Remote, real-time expert elicitation is a feasible, useful and practical tool to determine a prior distribution for international randomised controlled trials. Bayesian methods can then determine the trial sample size using these elicited prior distributions. The ease and low cost of remote expert elicitation mean that this approach is suitable for future international randomised controlled trials. TRIAL REGISTRATION: ClinicalTrials.gov NCT03567473.