Gemma Clayton

Empirical evidence of study design biases in randomized trials: Systematic review of meta-epidemiological studies

Objective To synthesise evidence on the average bias and heterogeneity associated with reported methodological features of randomized trials. Design Systematic review of meta-epidemiological studies. Methods We retrieved eligible studies included in a recent AHRQ-EPC review on this topic (latest search September 2012), and searched Ovid MEDLINE and Ovid EMBASE for studies indexed from Jan 2012-May 2015. Data were extracted by one author and verified by another. We combined estimates of average bias (e.g. ratio of odds ratios (ROR) or difference in standardised mean differences (dSMD)) in meta-analyses using the random-effects model. Analyses were stratified by type of outcome (“mortality” versus “other objective” versus “subjective”). Direction of effect was standardised so that ROR < 1 and dSMD < 0 denotes a larger intervention effect estimate in trials with an inadequate or unclear (versus adequate) characteristic. Results We included 24 studies. The available evidence suggests that intervention effect estimates may be exaggerated in trials with inadequate/unclear (versus adequate) sequence generation (ROR 0.93, 95% CI 0.86 to 0.99; 7 studies) and allocation concealment (ROR 0.90, 95% CI 0.84 to 0.97; 7 studies). For these characteristics, the average bias appeared to be larger in trials of subjective outcomes compared with other objective outcomes. Also, intervention effects for subjective outcomes appear to be exaggerated in trials with lack of/unclear blinding of participants (versus blinding) (dSMD -0.37, 95% CI -0.77 to 0.04; 2 studies), lack of/unclear blinding of outcome assessors (ROR 0.64, 95% CI 0.43 to 0.96; 1 study) and lack of/unclear double blinding (ROR 0.77, 95% CI 0.61 to 0.93; 1 study). The influence of other characteristics (e.g. unblinded trial personnel, attrition) is unclear. Conclusions Certain characteristics of randomized trials may exaggerate intervention effect estimates. The average bias appears to be greatest in trials of subjective outcomes. More research on several characteristics, particularly attrition and selective reporting, is needed.

The Effect of Patient-Specific Cerebral Oxygenation Monitoring on Postoperative Cognitive Function: A Multicenter Randomized Controlled Trial

Background Indices of global tissue oxygen delivery and utilization such as mixed venous oxygen saturation, serum lactate concentration, and arterial hematocrit are commonly used to determine the adequacy of tissue oxygenation during cardiopulmonary bypass (CPB). However, these global measures may not accurately reflect regional tissue oxygenation and ischemic organ injury remains a common and serious complication of CPB. Near-infrared spectroscopy (NIRS) is a noninvasive technology that measures regional tissue oxygenation. NIRS may be used alongside global measures to optimize regional perfusion and reduce organ injury. It may also be used as an indicator of the need for red blood cell transfusion in the presence of anemia and tissue hypoxia. However, the clinical benefits of using NIRS remain unclear and there is a lack of high-quality evidence demonstrating its efficacy and cost effectiveness. Objective The aim of the patient-specific cerebral oxygenation monitoring as part of an algorithm to reduce transfusion during heart valve surgery (PASPORT) trial is to determine whether the addition of NIRS to CPB management algorithms can prevent cognitive decline, postoperative organ injury, unnecessary transfusion, and reduce health care costs. Methods Adults aged 16 years or older undergoing valve or combined coronary artery bypass graft and valve surgery at one of three UK cardiac centers (Bristol, Hull, or Leicester) are randomly allocated in a 1:1 ratio to either a standard algorithm for optimizing tissue oxygenation during CPB that includes a fixed transfusion threshold, or a patient-specific algorithm that incorporates cerebral NIRS monitoring and a restrictive red blood cell transfusion threshold. Allocation concealment, Internet-based randomization stratified by operation type and recruiting center, and blinding of patients, ICU and ward care staff, and outcome assessors reduce the risk of bias. The primary outcomes are cognitive function 3 months after surgery and infectious complications during the first 3 months after surgery. Secondary outcomes include measures of inflammation, organ injury, and volumes of blood transfused. The cost effectiveness of the NIRS-based algorithm is described in terms of a cost-effectiveness acceptability curve. The trial tests the superiority of the patient-specific algorithm versus standard care. A sample size of 200 patients was chosen to detect a small to moderate target difference with 80% power and 5% significance (two tailed). Results Over 4 years, 208 patients have been successfully randomized and have been followed up for a 3-month period. Results are to be reported in 2015. Conclusions This study provides high-quality evidence, both valid and widely applicable, to determine whether the use of NIRS monitoring as part of a patient-specific management algorithm improves clinical outcomes and is cost effective. Trial Registration International Standard Randomized Controlled Trial Number (ISRCTN): 23557269; http://www.isrctn.com/ISRCTN23557269 (Archived by Webcite at http://www.webcitation.org/6buyrbj64)

The INVEST project: Investigating the use of evidence synthesis in the design and analysis of clinical trials

BackgroundWhen designing and analysing clinical trials, using previous relevant information, perhaps in the form of evidence syntheses, can reduce research waste. We conducted the INVEST (INVestigating the use of Evidence Synthesis in the design and analysis of clinical Trials) survey to summarise the current use of evidence synthesis in trial design and analysis, to capture opinions of trialists and methodologists on such use, and to understand any barriers.MethodsOur sampling frame was all delegates attending the International Clinical Trials Methodology Conference in November 2015. Respondents were asked to indicate (1) their views on the use of evidence synthesis in trial design and analysis, (2) their own use during the past 10 years and (3) the three greatest barriers to use in practice.ResultsOf approximately 638 attendees of the conference, 106 (17%) completed the survey, half of whom were statisticians. Support was generally high for using a description of previous evidence, a systematic review or a meta-analysis in trial design. Generally, respondents did not seem to be using evidence syntheses as often as they felt they should. For example, only 50% (42/84 relevant respondents) had used a meta-analysis to inform whether a trial is needed compared with 74% (62/84) indicating that this is desirable. Only 6% (5/81 relevant respondents) had used a value of information analysis to inform sample size calculations versus 22% (18/81) indicating support for this. Surprisingly large numbers of participants indicated support for, and previous use of, evidence syntheses in trial analysis. For example, 79% (79/100) of respondents indicated that external information about the treatment effect should be used to inform aspects of the analysis. The greatest perceived barrier to using evidence synthesis methods in trial design or analysis was time constraints, followed by a belief that the new trial was the first in the area.ConclusionsEvidence syntheses can be resource-intensive, but their use in informing the design, conduct and analysis of clinical trials is widely considered desirable. We advocate additional research, training and investment in resources dedicated to ways in which evidence syntheses can be undertaken more efficiently, offering the potential for cost savings in the long term.

Improving the efficiency of testing database functionality through statistical involvement

Background As a registered clinical trials unit we develop customised databases to collect and store study data and manage clinical trials; these databases need rigorous testing to ensure they function as intended, that data validation is implemented correctly and that study data extracts are complete and accurate. We describe how, with statistical involvement, the testing has been streamlined and the timelines reduced.

Normothermic versus hypothermic cardiopulmonary bypass in low-risk paediatric heart surgery: a randomised controlled trial

Objective To compare normothermic (35°C–36°C) versus hypothermic (28°C) cardiopulmonary bypass (CPB) in paediatric patients undergoing open heart surgery to test the hypothesis that normothermic CPB perfusion maintains the functional integrity of major organ systems leading to faster recovery. Methods Two single-centre, randomised controlled trials (known as Thermic-1 and Thermic-2, respectively) were carried out to compare the effectiveness and acceptability of normothermic versus hypothermic CPB in children with congenital heart disease undergoing open heart surgery. In both studies, the co-primary clinical outcomes were duration of inotropic support, intubation time and postoperative hospital stay. Results In total, 200 participants were recruited; 59 to the Thermic-1 study and 141 to the Thermic-2 study. 98 patients received normothermic CPB and 102 patients received hypothermic CPB. There were no significant differences between the treatment groups for any of the co-primary outcomes: inotrope duration HR=1.01, 95% CI (0.72 to 1.41); intubation time HR=1.14, 95% CI (0.86 to 1.51); postoperative hospital stay HR=1.06, 95% CI (0.80 to 1.40). Differences favouring normothermia were found in urea nitrogen at 2 days geometric mean ratio (GMR)=0.86 95% CI (0.77 to 0.97); serum creatinine at 3 days GMR=0.89, 95% CI (0.81 to 0.98); urinary albumin at 48 hours GMR=0.32, 95% CI (0.14 to 0.74) and neutrophil gelatinase-associated lipocalin at 4 hours GMR=0.47, 95% CI (0.22 to 1.02), but not at other postoperative time points. Conclusions Normothermic CPB is as safe and effective as hypothermic CPB and can be routinely adopted as a perfusion strategy in low-risk infants and children undergoing open heart surgery. Trial registration number ISRCTN93129502.

How Often Do Safety Signals Occur by Chance in First‐in‐Human Trials?

Clinicians working on first‐in‐human clinical studies need to be able to judge whether safety signals observed on an investigational drug were more likely to have occurred by chance or to have been caused by the drug. We retrospectively reviewed 84 Novartis studies including 1,234 healthy volunteers receiving placebo to determine the expected incidence of changes in commonly measured laboratory parameters and vital signs, in the absence of any active agent. We calculated the frequency of random incidence of safety signals, focusing on the liver, cardiovascular system, kidney, and pancreas. Using the liver enzyme alanine aminotransferase (ALT) as an example, we illustrate how a predictive model can be used to determine the probability of a given subject to experience an elevation of ALT above the upper limit of the normal range under placebo, conditional on the characteristics of this subject and the study.

A mixed-methods feasibility and external pilot study to inform a large pragmatic randomised controlled trial of the effects of surgical wound dressing strategies on surgical site infections (Bluebelle Phase B): study protocol for a randomised controlled trial

BackgroundSurgical site infections (SSIs) are common, occurring in up to 25% of > 4 million operations performed in England each year. Previous trials of the effect of wound dressings on the risk of developing a SSI are of poor quality and underpowered.Methods/DesignThis study is a feasibility and pilot trial to examine the feasibility of a full trial that will compare simple dressings, no dressing and tissue-glue as a dressing. It is examining the overall acceptability of trial participation, identifying opportunities for refinement, testing the feasibility of and validating new outcome tools to assess SSI, wound management issues and patients’ wound symptom experiences. It is also exploring methods for avoiding performance bias and blinding outcome assessors by testing the feasibility of collecting wound photographs taken in theatre immediately after wound closure and, at 4–8 weeks after surgery, taken by participants themselves or their carers. Finally, it is identifying the main cost drivers for an economic evaluation of dressing types. Integrated qualitative research is exploring acceptability and reasons for non-adherence to allocation. Adults undergoing primary elective or unplanned abdominal general surgery or Caesarean section are eligible. The main exclusion criteria are abdominal or other major surgery less than three months before the index operation or contraindication to dressing allocation. The trial is scheduled to recruit for nine months. The findings will be used to inform the design of a main trial.DiscussionThis pilot trial is the first pragmatic study to randomise participants to no dressing or tissue-glue as a dressing versus a simple dressing. Early evidence from the ongoing pilot shows that recruitment is proceeding well and that the interventions are acceptable to participants. Combined with the qualitative findings, the findings will inform whether a main, large trial is feasible and, if so, how it should be designed.Trial registrationISRCTN49328913. Registered on 20 October 2015.

Analysis of repeated measures of multiple surgical sites within a factorial randomised controlled trial.

Background In heart bypass surgery, pieces of artery or vein (“conduits”) taken from elsewhere in the patients’ body are used to bypass the blocked or narrowed coronary arteries and improve the blood supply to the heart. The “Hyperplasia and Atherogenesis in Bypass Vein Grafts Following Different Surgical Preparation Techniques” (HArVeST) trial, is a single-centre study investigating if new methods of taking and preparing the conduit can improve the outcome for patients.