Applying the Estimand Framework to Non­Inferiority Trials.

Most published applications of the estimand framework have focused on superiority trials. However, non-inferiority trials present specific challenges compared to superiority trials. The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use notes in their addendum on estimands and sensitivity analysis in clinical trials that there may be special considerations to the implementation of estimands in clinical trials with a non-inferiority objective yet provides little guidance. This paper discusses considerations that trial teams should make when defining estimands for a clinical trial with a non-inferiority objective. We discuss how the pre-addendum way of establishing non-inferiority can be embraced by the estimand framework including a discussion of the role of the Per Protocol analysis set. We examine what clinical questions of interest can be formulated in the context of non-inferiority trials and outline why we do not think it is sensible to describe an estimand as 'conservative'. The impact of the estimand framework on key considerations in non-inferiority trials such as whether trials should have more than one primary estimand, the choice of non-inferiority margin, assay sensitivity, switching from non-inferiority to superiority and estimation are discussed. We conclude by providing a list of recommendations, and important considerations for defining estimands for trials with a non-inferiority objective.

Handling intercurrent events and missing data in non-inferiority trials using the estimand framework: A tuberculosis case study.

INTRODUCTION: The ICH E9 addendum outlining the estimand framework for clinical trials was published in 2019 but provides limited guidance around how to handle intercurrent events for non-inferiority studies. Once an estimand is defined, it is also unclear how to deal with missing values using principled analyses for non-inferiority studies. METHODS: Using a tuberculosis clinical trial as a case study, we propose a primary estimand, and an additional estimand suitable for non-inferiority studies. For estimation, multiple imputation methods that align with the estimands for both primary and sensitivity analysis are proposed. We demonstrate estimation methods using the twofold fully conditional specification multiple imputation algorithm and then extend and use reference-based multiple imputation for a binary outcome to target the relevant estimands, proposing sensitivity analyses under each. We compare the results from using these multiple imputation methods with those from the original study. RESULTS: Consistent with the ICH E9 addendum, estimands can be constructed for a non-inferiority trial which improves on the per-protocol/intention-to-treat-type analysis population previously advocated, involving respectively a hypothetical or treatment policy strategy to handle relevant intercurrent events. Results from using the 'twofold' multiple imputation approach to estimate the primary hypothetical estimand, and using reference-based methods for an additional treatment policy estimand, including sensitivity analyses to handle the missing data, were consistent with the original study's reported per-protocol and intention-to-treat analysis in failing to demonstrate non-inferiority. CONCLUSIONS: Using carefully constructed estimands and appropriate primary and sensitivity estimators, using all the information available, results in a more principled and statistically rigorous approach to analysis. Doing so provides an accurate interpretation of the estimand.

Randomized Controlled Trial of Urokinase versus Placebo for Nondraining Malignant Pleural Effusion.

RATIONALE: Patients with malignant pleural effusion experience breathlessness, which is treated by drainage and pleurodesis. Incomplete drainage results in residual dyspnea and pleurodesis failure. Intrapleural fibrinolytics lyse septations within pleural fluid, improving drainage. OBJECTIVES: To assess the effects of intrapleural urokinase on dyspnea and pleurodesis success in patients with nondraining malignant effusion. METHODS: We conducted a prospective, double-blind, randomized trial. Patients with nondraining effusion were randomly allocated in a 1:1 ratio to intrapleural urokinase (100,000 IU, three doses, 12-hourly) or matched placebo. MEASUREMENTS AND MAIN RESULTS: Co-primary outcome measures were dyspnea (average daily 100-mm visual analog scale scores over 28 d) and time to pleurodesis failure to 12 months. Secondary outcomes were survival, hospital length of stay, and radiographic change. A total of 71 subjects were randomized (36 received urokinase, 35 placebo) from 12 U.K. centers. The baseline characteristics were similar between the groups. There was no difference in mean dyspnea between groups (mean difference, 3.8 mm; 95% confidence interval [CI], -12 to 4.4 mm; P = 0.36). Pleurodesis failure rates were similar (urokinase, 13 of 35 [37%]; placebo, 11 of 34 [32%]; adjusted hazard ratio, 1.2; P = 0.65). Urokinase was associated with decreased effusion size visualized by chest radiography (adjusted relative improvement, -19%; 95% CI, -28 to -11%; P < 0.001), reduced hospital stay (1.6 d; 95% CI, 1.0 to 2.6; P = 0.049), and improved survival (69 vs. 48 d; P = 0.026). CONCLUSIONS: Use of intrapleural urokinase does not reduce dyspnea or improve pleurodesis success compared with placebo and cannot be recommended as an adjunct to pleurodesis. Other palliative treatments should be used. Improvements in hospital stay, radiographic appearance, and survival associated with urokinase require further evaluation. Clinical trial registered with ISRCTN (12852177) and EudraCT (2008-000586-26).

Effect of Home Noninvasive Ventilation With Oxygen Therapy vs Oxygen Therapy Alone on Hospital Readmission or Death After an Acute COPD Exacerbation: A Randomized Clinical Trial.

Importance: Outcomes after exacerbations of chronic obstructive pulmonary disease (COPD) requiring acute noninvasive ventilation (NIV) are poor and there are few treatments to prevent hospital readmission and death. Objective: To investigate the effect of home NIV plus oxygen on time to readmission or death in patients with persistent hypercapnia after an acute COPD exacerbation. Design, Setting, and Participants: A randomized clinical trial of patients with persistent hypercapnia (Paco2 >53 mm Hg) 2 weeks to 4 weeks after resolution of respiratory acidemia, who were recruited from 13 UK centers between 2010 and 2015. Exclusion criteria included obesity (body mass index [BMI] >35), obstructive sleep apnea syndrome, or other causes of respiratory failure. Of 2021 patients screened, 124 were eligible. Interventions: There were 59 patients randomized to home oxygen alone (median oxygen flow rate, 1.0 L/min [interquartile range {IQR}, 0.5-2.0 L/min]) and 57 patients to home oxygen plus home NIV (median oxygen flow rate, 1.0 L/min [IQR, 0.5-1.5 L/min]). The median home ventilator settings were an inspiratory positive airway pressure of 24 (IQR, 22-26) cm H2O, an expiratory positive airway pressure of 4 (IQR, 4-5) cm H2O, and a backup rate of 14 (IQR, 14-16) breaths/minute. Main Outcomes and Measures: Time to readmission or death within 12 months adjusted for the number of previous COPD admissions, previous use of long-term oxygen, age, and BMI. Results: A total of 116 patients (mean [SD] age of 67 [10] years, 53% female, mean BMI of 21.6 [IQR, 18.2-26.1], mean [SD] forced expiratory volume in the first second of expiration of 0.6 L [0.2 L], and mean [SD] Paco2 while breathing room air of 59 [7] mm Hg) were randomized. Sixty-four patients (28 in home oxygen alone and 36 in home oxygen plus home NIV) completed the 12-month study period. The median time to readmission or death was 4.3 months (IQR, 1.3-13.8 months) in the home oxygen plus home NIV group vs 1.4 months (IQR, 0.5-3.9 months) in the home oxygen alone group, adjusted hazard ratio of 0.49 (95% CI, 0.31-0.77; P = .002). The 12-month risk of readmission or death was 63.4% in the home oxygen plus home NIV group vs 80.4% in the home oxygen alone group, absolute risk reduction of 17.0% (95% CI, 0.1%-34.0%). At 12 months, 16 patients had died in the home oxygen plus home NIV group vs 19 in the home oxygen alone group. Conclusions and Relevance: Among patients with persistent hypercapnia following an acute exacerbation of COPD, adding home noninvasive ventilation to home oxygen therapy prolonged the time to readmission or death within 12 months. Trial Registration: clinicaltrials.gov Identifier: NCT00990132.

A dose-ranging trial to optimize the dose of rifampin in the treatment of tuberculosis.

RATIONALE: Rifampin at a dose of 10 mg/kg was introduced in 1971 based on pharmacokinetic, toxicity, and cost considerations. Available data in mice and humans showed that an increase in dose may shorten the duration of tuberculosis treatment. OBJECTIVES: To evaluate the safety and tolerability, the pharmacokinetics, and the extended early bactericidal activity of increasing doses of rifampin. METHODS: Patients with drug-susceptible tuberculosis were enrolled into a control group of eight patients receiving the standard dose of 10 mg/kg rifampin, followed by consecutive experimental groups with 15 patients each receiving rifampin 20, 25, 30, and 35 mg/kg, respectively, for 14 days. In all patients isoniazid, pyrazinamide, and ethambutol were added in standard doses for the second 7 days of treatment. Safety, pharmacokinetics of rifampin, and fall in bacterial load were assessed. MEASUREMENTS AND MAIN RESULTS: Grade 1 and 2 adverse events were equally distributed between the five dose groups; there were five grade 3 events of which one was a possibly related hepatotoxicity. Areas under the time-concentration curves and peak serum concentrations of rifampin showed a more than proportional increase with dose. The daily fall in bacterial load over 14 days was 0.176, 0.168, 0.167, 0.265, and 0.261 log10 colony-forming units/ml sputum in the 10, 20, 25, 30, and 35 mg/kg groups, respectively. CONCLUSIONS: Two weeks of rifampin up to 35 mg/kg was safe and well tolerated. There was a nonlinear increase in exposure to rifampin without an apparent ceiling effect and a greater estimated fall in bacterial load in the higher dosing groups. Clinical trial registered with www.clinicaltrials.gov (NCT 01392911).

Effect of Opioids vs NSAIDs and Larger vs Smaller Chest Tube Size on Pain Control and Pleurodesis Efficacy Among Patients With Malignant Pleural Effusion: The TIME1 Randomized Clinical Trial.

IMPORTANCE: For treatment of malignant pleural effusion, nonsteroidal anti-inflammatory drugs (NSAIDs) are avoided because they may reduce pleurodesis efficacy. Smaller chest tubes may be less painful than larger tubes, but efficacy in pleurodesis has not been proven. OBJECTIVE: To assess the effect of chest tube size and analgesia (NSAIDs vs opiates) on pain and clinical efficacy related to pleurodesis in patients with malignant pleural effusion. DESIGN, SETTING, AND PARTICIPANTS: A 2×2 factorial phase 3 randomized clinical trial among 320 patients requiring pleurodesis in 16 UK hospitals from 2007 to 2013. INTERVENTIONS: Patients undergoing thoracoscopy (n = 206; clinical decision if biopsy was required) received a 24F chest tube and were randomized to receive opiates (n = 103) vs NSAIDs (n = 103), and those not undergoing thoracoscopy (n = 114) were randomized to 1 of 4 groups (24F chest tube and opioids [n = 28]; 24F chest tube and NSAIDs [n = 29]; 12F chest tube and opioids [n = 29]; or 12F chest tube and NSAIDs [n = 28]). MAIN OUTCOMES AND MEASURES: Pain while chest tube was in place (0- to 100-mm visual analog scale [VAS] 4 times/d; superiority comparison) and pleurodesis efficacy at 3 months (failure defined as need for further pleural intervention; noninferiority comparison; margin, 15%). RESULTS: Pain scores in the opiate group (n = 150) vs the NSAID group (n = 144) were not significantly different (mean VAS score, 23.8 mm vs 22.1 mm; adjusted difference, -1.5 mm; 95% CI, -5.0 to 2.0 mm; P = .40), but the NSAID group required more rescue analgesia (26.3% vs 38.1%; rate ratio, 2.1; 95% CI, 1.3-3.4; P = .003). Pleurodesis failure occurred in 30 patients (20%) in the opiate group and 33 (23%) in the NSAID group, meeting criteria for noninferiority (difference, -3%; 1-sided 95% CI, -10% to ∞; P = .004 for noninferiority). Pain scores were lower among patients in the 12F chest tube group (n = 54) vs the 24F group (n = 56) (mean VAS score, 22.0 mm vs 26.8 mm; adjusted difference, -6.0 mm; 95% CI, -11.7 to -0.2 mm; P = .04) and 12F chest tubes vs 24F chest tubes were associated with higher pleurodesis failure (30% vs 24%), failing to meet noninferiority criteria (difference, -6%; 1-sided 95% CI, -20% to ∞; P = .14 for noninferiority). Complications during chest tube insertion occurred more commonly with 12F tubes (14% vs 24%; odds ratio, 1.91; P = .20). CONCLUSIONS AND RELEVANCE: Use of NSAIDs vs opiates resulted in no significant difference in pain scores but was associated with more rescue medication. NSAID use resulted in noninferior rates of pleurodesis efficacy at 3 months. Placement of 12F chest tubes vs 24F chest tubes was associated with a statistically significant but clinically modest reduction in pain but failed to meet noninferiority criteria for pleurodesis efficacy. TRIAL REGISTRATION: isrctn.org Identifier: ISRCTN33288337.

A randomized, cross-over study comparing critical and overall errors, learning time, and preference of the ELLIPTA versus BREEZHALER dry powder inhalers in patients with asthma.

BACKGROUND: Many patients with asthma make errors using inhalers, affecting the amount of medication received. Previous evidence demonstrated that patients with asthma or chronic obstructive pulmonary disease make fewer critical errors with the ELLIPTA inhaler after reading the patient information leaflet (PIL) versus other dry powder inhalers. We assessed errors made by patients with asthma using placebo ELLIPTA or BREEZHALER inhalers. METHODS: This randomized, multicenter, open-label placebo inhaler-handling study (ClinicalTrials.gov: NCT04813354) with 2x2 complete block crossover design was conducted at three centers in the Netherlands and enrolled patients aged ≥18 years with mild-to-moderate asthma. Inclusion criteria were inhaler use for ≥12 weeks prior to enrollment and naivety to ELLIPTA and BREEZHALER inhalers. Patients were randomized to ELLIPTA or BREEZHALER inhaler first and were assessed for errors in use of both inhalers after 1) reading PIL instructions, 2) receiving further instruction from a healthcare professional (HCP) if they made an error. RESULTS: 114 patients with asthma (57% female; mean age of 55.3 years) were assessed. After reading the PIL, 6% of patients made ≥1 critical error with ELLIPTA versus 26% with BREEZHALER (odds ratio [OR]: 0.11 [95% confidence interval (CI): 0.01-0.40]; p < 0.001). With ELLIPTA, 27% of patients made ≥1 overall error after reading the PIL versus 41% with BREEZHALER (OR: 0.25 [95% CI: 0.03-0.74]; p = 0.005). Fewer patients required HCP instruction with ELLIPTA than BREEZHALER (25% versus 32%). CONCLUSIONS: Fewer patients made critical and overall errors using the ELLIPTA inhaler versus BREEZHALER after reading the PIL.

Evaluating how clear the questions being investigated in randomised trials are: systematic review of estimands.

OBJECTIVES: To evaluate how often the precise research question being addressed about an intervention (the estimand) is stated or can be determined from reported methods, and to identify what types of questions are being investigated in phase 2-4 randomised trials. DESIGN: Systematic review of the clarity of research questions being investigated in randomised trials in 2020 in six leading general medical journals. DATA SOURCE: PubMed search in February 2021. ELIGIBILITY CRITERIA FOR SELECTING STUDIES: Phase 2-4 randomised trials, with no restrictions on medical conditions or interventions. Cluster randomised, crossover, non-inferiority, and equivalence trials were excluded. MAIN OUTCOME MEASURES: Number of trials that stated the precise primary question being addressed about an intervention (ie, the primary estimand), or for which the primary estimand could be determined unambiguously from the reported methods using statistical knowledge. Strategies used to handle post-randomisation events that affect the interpretation or existence of patient outcomes, such as intervention discontinuations or uses of additional drug treatments (known as intercurrent events), and the corresponding types of questions being investigated. RESULTS: 255 eligible randomised trials were identified. No trials clearly stated all the attributes of the estimand. In 117 (46%) of 255 trials, the primary estimand could be determined from the reported methods. Intercurrent events were reported in 242 (95%) of 255 trials; but the handling of these could only be determined in 125 (49%) of 255 trials. Most trials that provided this information considered the occurrence of intercurrent events as irrelevant in the calculation of the treatment effect and assessed the effect of the intervention regardless (96/125, 77%)-that is, they used a treatment policy strategy. Four (4%) of 99 trials with treatment non-adherence owing to adverse events estimated the treatment effect in a hypothetical setting (ie, the effect as if participants continued treatment despite adverse events), and 19 (79%) of 24 trials where some patients died estimated the treatment effect in a hypothetical setting (ie, the effect as if participants did not die). CONCLUSIONS: The precise research question being investigated in most trials is unclear, mainly because of a lack of clarity on the approach to handling intercurrent events. Clear reporting of estimands is necessary in trial reports so that all stakeholders, including clinicians, patients and policy makers, can make fully informed decisions about medical interventions. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42021238053.

Most noninferiority trials were not designed to preserve active comparator treatment effects.

OBJECTIVES: To evaluate whether noninferiority trials are designed to adequately preserve the historical treatment effect of their active comparators. STUDY DESIGN AND SETTING: We reviewed 162 noninferiority trials published in high-impact medical journals. We assessed whether trials were designed to ensure that interventions could only be declared noninferior if they preserved at least 50% of the active comparator's historical treatment effect. RESULTS: Only 25 of 162 trials (15%) were designed so that interventions could only be declared noninferior if they preserved at least 50% of the active comparator's historical treatment effect. Most trials did not provide evidence that the active comparator was effective (n = 101), provided inadequate evidence (n = 18), or used a noninferiority margin that was too wide (n = 18). In a subset of 61 noninferiority trials which referenced a prior randomized trial or meta-analysis evaluating the active comparator, only 25 (41%) used a noninferiority margin small enough to preserve at least 50% of the active comparator's treatment effect. Overall, 14 of 162 noninferiority trials (9%) would have allowed the intervention to be declared noninferior even if it was worse than either placebo or another historical control. CONCLUSION: Most noninferiority trials published in major medical journals could allow erroneous declarations of noninferiority.

Gantenerumab reduces amyloid-ß plaques in patients with prodromal to moderate Alzheimer's disease: a PET substudy interim analysis.

BACKGROUND: We previously investigated low doses (105 or 225 mg) of gantenerumab, a fully human monoclonal antibody that binds and removes aggregated amyloid-ß by Fc receptor-mediated phagocytosis, in the SCarlet RoAD (SR) and Marguerite RoAD (MR) phase 3 trials. Several lines of evidence suggested that higher doses may be necessary to achieve clinical efficacy. We therefore designed a positron emission tomography (PET) substudy to evaluate the effect of gantenerumab uptitrated to 1200 mg every 4 weeks on amyloid-ß plaques as measured using florbetapir PET in patients with prodromal to moderate Alzheimer's disease (AD). METHODS: A subset of patients enrolled in the SR and MR studies who subsequently entered the open-label extensions (OLEs) were included in this substudy. Patients were aged 50 to 90 years with a clinical diagnosis of probable prodromal to moderate AD and were included based on a visual read of the original screening scan in the double-blind phase. Patients were assigned to 1 of 5 titration schedules (ranging from 2 to 10 months) with a target gantenerumab dose of 1200 mg every 4 weeks. The main endpoint of this substudy was change in amyloid-ß plaque burden from OLE baseline to week 52 and week 104, assessed using florbetapir PET. Florbetapir global cortical signal was calculated using a prespecified standard uptake value ratio method converted to the Centiloid scale. RESULTS: Sixty-seven of the 89 patients initially enrolled had ≥ 1 follow-up scan by August 15, 2018. Mean amyloid levels were reduced by 39 Centiloids by the first year and 59 Centiloids by year 2, a 3.5-times greater reduction than was seen after 2 years at 225 mg in SR. At years 1 and 2, 37% and 51% of patients, respectively, had amyloid-ß plaque levels below the amyloid-ß positivity threshold. CONCLUSION: Results from this exploratory interim analysis of the PET substudy suggest that gantenerumab doses up to 1200 mg resulted in robust amyloid-ß plaque removal at 2 years. PET amyloid levels were consistent with sparse-to-no neuritic amyloid-ß plaques in 51% of patients after 2 years of therapy. Amyloid reductions were similar to those observed in other placebo-controlled studies that have suggested potential clinical benefit. TRIAL REGISTRATION: ClinicalTrials.gov, NCT01224106 (SCarlet RoAD) and NCT02051608 (Marguerite RoAD).

High-dose rifampicin, moxifloxacin, and SQ109 for treating tuberculosis: a multi-arm, multi-stage randomised controlled trial.

BACKGROUND: Tuberculosis is the world's leading infectious disease killer. We aimed to identify shorter, safer drug regimens for the treatment of tuberculosis. METHODS: We did a randomised controlled, open-label trial with a multi-arm, multi-stage design. The trial was done in seven sites in South Africa and Tanzania, including hospitals, health centres, and clinical trial centres. Patients with newly diagnosed, rifampicin-sensitive, previously untreated pulmonary tuberculosis were randomly assigned in a 1:1:1:1:2 ratio to receive (all orally) either 35 mg/kg rifampicin per day with 15-20 mg/kg ethambutol, 20 mg/kg rifampicin per day with 400 mg moxifloxacin, 20 mg/kg rifampicin per day with 300 mg SQ109, 10 mg/kg rifampicin per day with 300 mg SQ109, or a daily standard control regimen (10 mg/kg rifampicin, 5 mg/kg isoniazid, 25 mg/kg pyrazinamide, and 15-20 mg/kg ethambutol). Experimental treatments were given with oral 5 mg/kg isoniazid and 25 mg/kg pyrazinamide per day for 12 weeks, followed by 14 weeks of 5 mg/kg isoniazid and 10 mg/kg rifampicin per day. Because of the orange discoloration of body fluids with higher doses of rifampicin it was not possible to mask patients and clinicians to treatment allocation. The primary endpoint was time to culture conversion in liquid media within 12 weeks. Patients without evidence of rifampicin resistance on phenotypic test who took at least one dose of study treatment and had one positive culture on liquid or solid media before or within the first 2 weeks of treatment were included in the primary analysis (modified intention to treat). Time-to-event data were analysed using a Cox proportional-hazards regression model and adjusted for minimisation variables. The proportional hazard assumption was tested using Schoelfeld residuals, with threshold p<0·05 for non-proportionality. The trial is registered with ClinicalTrials.gov (NCT01785186). FINDINGS: Between May 7, 2013, and March 25, 2014, we enrolled and randomly assigned 365 patients to different treatment arms (63 to rifampicin 35 mg/kg, isoniazid, pyrazinamide, and ethambutol; 59 to rifampicin 10 mg/kg, isoniazid, pyrazinamide, SQ109; 57 to rifampicin 20 mg/kg, isoniazid, pyrazinamide, and SQ109; 63 to rifampicin 10 mg/kg, isoniazid, pyrazinamide, and moxifloxacin; and 123 to the control arm). Recruitment was stopped early in the arms containing SQ109 since prespecified efficacy thresholds were not met at the planned interim analysis. Time to stable culture conversion in liquid media was faster in the 35 mg/kg rifampicin group than in the control group (median 48 days vs 62 days, adjusted hazard ratio 1·78; 95% CI 1·22-2·58, p=0·003), but not in other experimental arms. There was no difference in any of the groups in time to culture conversion on solid media. 11 patients had treatment failure or recurrent disease during post-treatment follow-up: one in the 35 mg/kg rifampicin arm and none in the moxifloxacin arm. 45 (12%) of 365 patients reported grade 3-5 adverse events, with similar proportions in each arm. INTERPRETATION: A dose of 35 mg/kg rifampicin was safe, reduced the time to culture conversion in liquid media, and could be a promising component of future, shorter regimens. Our adaptive trial design was successfully implemented in a multi-centre, high tuberculosis burden setting, and could speed regimen development at reduced cost. FUNDING: The study was funded by the European and Developing Countries Clinical Trials partnership (EDCTP), the German Ministry for Education and Research (BmBF), and the Medical Research Council UK (MRC).

Non-inferiority trials: are they inferior? A systematic review of reporting in major medical journals.

OBJECTIVE: To assess the adequacy of reporting of non-inferiority trials alongside the consistency and utility of current recommended analyses and guidelines. DESIGN: Review of randomised clinical trials that used a non-inferiority design published between January 2010 and May 2015 in medical journals that had an impact factor >10 (JAMA Internal Medicine, Archives Internal Medicine, PLOS Medicine, Annals of Internal Medicine, BMJ, JAMA, Lancet and New England Journal of Medicine). DATA SOURCES: Ovid (MEDLINE). METHODS: We searched for non-inferiority trials and assessed the following: choice of non-inferiority margin and justification of margin; power and significance level for sample size; patient population used and how this was defined; any missing data methods used and assumptions declared and any sensitivity analyses used. RESULTS: A total of 168 trial publications were included. Most trials concluded non-inferiority (132; 79%). The non-inferiority margin was reported for 98% (164), but less than half reported any justification for the margin (77; 46%). While most chose two different analyses (91; 54%) the most common being intention-to-treat (ITT) or modified ITT and per-protocol, a large number of articles only chose to conduct and report one analysis (65; 39%), most commonly the ITT analysis. There was lack of clarity or inconsistency between the type I error rate and corresponding CIs for 73 (43%) articles. Missing data were rarely considered with (99; 59%) not declaring whether imputation techniques were used. CONCLUSIONS: Reporting and conduct of non-inferiority trials is inconsistent and does not follow the recommendations in available statistical guidelines, which are not wholly consistent themselves. Authors should clearly describe the methods used and provide clear descriptions of and justifications for their design and primary analysis. Failure to do this risks misleading conclusions being drawn, with consequent effects on clinical practice.

Blinded Outcome Assessment Was Infrequently Used and Poorly Reported in Open Trials.

OBJECTIVE: Unblinded outcome assessment can lead to biased estimates of treatment effect in randomised trials. We reviewed published trials to assess how often blinded assessment is used, and whether its use varies according to the type of outcome or assessor. DESIGN AND SETTING: A review of parallel group, individually randomised phase III trials published in four general medical journals (BMJ, Journal of the American Medical Association, The Lancet, and New England Journal of Medicine) in 2010. MAIN OUTCOME MEASURES: Whether assessment of the primary outcome was blinded, and whether this differed according to outcome or assessor type. RESULTS: We identified 258 eligible trials. Of these, 106 (41%) were reported as double-blind, and 152 (59%) as partially or fully open-label (that is, they included some groups who were unblinded, such as patients, those delivering the intervention, or those in charge of medical care). Of the 152 open trials, 125 required outcome assessment. Of these 125 trials, only 26% stated that outcome assessment was blinded; 51% gave no information on whether assessment was blinded or not. Furthermore, 18% of trials did not state who performed the assessment. The choice of outcome type (e.g. instrument measured, rated, or naturally occurring event) did not appear to influence whether blinded assessment was performed (range 24-32% for the most common outcome types). However, the choice of outcome assessor did influence blinding; independent assessors were blinded much more frequently (71%) than participant (5%) or physician (24%) assessors. Despite this, open trials did not use independent assessors any more frequently than double-blind trials (17% vs. 18% respectively). CONCLUSIONS: Blinding of outcome assessors is infrequently used and poorly reported. Increased use of independent assessors could increase the frequency of blinded assessment.

Risk of selection bias in randomised trials.

BACKGROUND: Selection bias occurs when recruiters selectively enrol patients into the trial based on what the next treatment allocation is likely to be. This can occur even if appropriate allocation concealment is used if recruiters can guess the next treatment assignment with some degree of accuracy. This typically occurs in unblinded trials when restricted randomisation is implemented to force the number of patients in each arm or within each centre to be the same. Several methods to reduce the risk of selection bias have been suggested; however, it is unclear how often these techniques are used in practice. METHODS: We performed a review of published trials which were not blinded to assess whether they utilised methods for reducing the risk of selection bias. We assessed the following techniques: (a) blinding of recruiters; (b) use of simple randomisation; (c) avoidance of stratification by site when restricted randomisation is used; (d) avoidance of permuted blocks if stratification by site is used; and (e) incorporation of prognostic covariates into the randomisation procedure when restricted randomisation is used. We included parallel group, individually randomised phase III trials published in four general medical journals (BMJ, Journal of the American Medical Association, The Lancet, and New England Journal of Medicine) in 2010. RESULTS: We identified 152 eligible trials. Most trials (98%) provided no information on whether recruiters were blind to previous treatment allocations. Only 3% of trials used simple randomisation; 63% used some form of restricted randomisation, and 35% did not state the method of randomisation. Overall, 44% of trials were stratified by site of recruitment; 27% were not, and 29% did not report this information. Most trials that did stratify by site of recruitment used permuted blocks (58%), and only 15% reported using random block sizes. Many trials that used restricted randomisation also included prognostic covariates in the randomisation procedure (56%). CONCLUSIONS: The risk of selection bias could not be ascertained for most trials due to poor reporting. Many trials which did provide details on the randomisation procedure were at risk of selection bias due to a poorly chosen randomisation methods. Techniques to reduce the risk of selection bias should be more widely implemented.

Reducing bias in open-label trials where blinded outcome assessment is not feasible: strategies from two randomised trials.

BACKGROUND: Blinded outcome assessment is recommended in open-label trials to reduce bias, however it is not always feasible. It is therefore important to find other means of reducing bias in these scenarios. METHODS: We describe two randomised trials where blinded outcome assessment was not possible, and discuss the strategies used to reduce the possibility of bias. RESULTS: TRIGGER was an open-label cluster randomised trial whose primary outcome was further bleeding. Because of the cluster randomisation, all researchers in a hospital were aware of treatment allocation and so could not perform a blinded assessment. A blinded adjudication committee was also not feasible as it was impossible to compile relevant information to send to the committee in a blinded manner. Therefore, the definition of further bleeding was modified to exclude subjective aspects (such as whether symptoms like vomiting blood were severe enough to indicate the outcome had been met), leaving only objective aspects (the presence versus absence of active bleeding in the upper gastrointestinal tract confirmed by an internal examination).TAPPS was an open-label trial whose primary outcome was whether the patient was referred for a pleural drainage procedure. Allowing a blinded assessor to decide whether to refer the patient for a procedure was not feasible as many clinicians may be reluctant to enrol patients into the trial if they cannot be involved in their care during follow-up. Assessment by an adjudication committee was not possible, as the outcome either occurred or did not. Therefore, the decision pathway for procedure referral was modified. If a chest x-ray indicated that more than a third of the pleural space filled with fluid, the patient could be referred for a procedure; otherwise, the unblinded clinician was required to reach a consensus on referral with a blinded assessor. This process allowed the unblinded clinician to be involved in the patient's care, while reducing the potential for bias. CONCLUSIONS: When blinded outcome assessment is not possible, it may be useful to modify the outcome definition or method of assessment to reduce the risk of bias. TRIAL REGISTRATION TRIGGER: ISRCTN85757829. Registered 26 July 2012.TAPPS: ISRCTN47845793. Registered 28 May 2012.

Acute variceal haemorrhage in the United Kingdom: patient characteristics, management and outcomes in a nationwide audit.

BACKGROUND: Despite advances in treatment, acute variceal haemorrhage remains life-threatening. AIM: To describe contemporary characteristics, management and outcomes of patients with cirrhosis and acute variceal haemorrhage and risk factors for rebleeding and mortality. METHODS: Multi-centre clinical audit conducted in 212 UK hospitals. RESULTS: In 526 cases of acute variceal haemorrhage, 66% underwent endoscopy within 24h with 64% (n=339) receiving endoscopic therapy. Prior to endoscopy, 57% (n=299) received proton pump inhibitors, 44% (n=232) vasopressors and 27% (n=144) antibiotics. 73% (n=386) received red cell transfusion, 35% (n=184) fresh frozen plasma and 14% (n=76) platelets, with widely varying transfusion thresholds. 26% (n=135) experienced further bleeding and 15% (n=80) died by day 30. The Model for End Stage Liver Disease score was the best predictor of mortality (area under the receiver operating curve=0.74, P<0.001). Neither the clinical nor full Rockall scores were useful predictors of outcome. Coagulopathy was strongly associated with rebleeding (odds ratio 2.23, 95% CI 1.22-4.07, P=0.01, up to day 30) and mortality (odds ratio 3.06, 95% CI 1.29-7.26, P=0.01). CONCLUSIONS: Although mortality has improved following acute variceal haemorrhage, rebleeding rates remain appreciably high. There are notable deficiencies in the use of vasopressors and endoscopic therapy. More work is needed to understand the optimum transfusion strategies. Better risk stratification tools are required to identify patients needing more intensive support.