Current methodological research
Title: Randomised Trial Of An Incentive to Promote Questionnaire Return.
Authors: Simon Gates, Sallie Lamb, Mark Williams, Esther Williamson
Brief details: Many randomised trials collect data by postal questionnaires to the participants. Poor response rates mean that data will be missing for a significant proportion of the participants, which reduces the sample size and results in loss of statistical power, and may introduce bias. It is therefore essential for the response rate to be as high as possible. Many ways of encouraging participants to return questionnaires have been investigated, but financial incentives have been found to be effective in non-trial questionnaires. There is less evidence for the effectiveness of financial incentives in trials. In this study participants in the Managind Injuries of the Neck Trial (MINT) are randomised to receive or not to receive a £5 gift voucher with their follow-up questionnaire. The main outcome measure will be return or non-return of the completed questionnaire. 2,160 MINT participants will be included in the incentive study and it is expected to be completed in late 2007.
Title: Multiplicity in Systematic Reviews.
Authors: Ralph Bender, Catey Bunce, Mike Clarke, Simon Gates, Stefan Lange, Nathan Pace, Kristian Thorlund
Brief details: When a large number of statistical tests are carried out within a study, the chances of finding spuriously significant results that are due only to chance increases. This may lead to erroneous conclusions. Multiple statistical tests can arise in several different ways in systematic reviews, including multile outcome measures, multiple subgroup analyses, multiple intervention groups and multiple time points. In this project, an international group of statisticians and methodologists from The Cochrane Collaboration, including members from Germany, UK, USA and Denmark, is reviewing the problems and possible solutions of multiple comparisons in systematic reviews. Expected completion date summer 2007.
Title: Cochrane methodology review of double data entry for reducing data transcription errors.
Authors: Simon Gates, Sallie Lamb.
Brief details: Many trials and other research studies collect data on paper forms and enter these into computer databases for processing and analysis. Frequently, double data entry systems are used, in which data are entered twice, independently and by different people, and the resulting data sets compared to identify discrepancies. This is assumed to detect errors that would occur with single data entry. However, double data entry has time and monetary costs, and may detect only a few, unimportant errors. Some research studies have suggested that its use may not be justified. We are conducting a systematic review to summarise studies that have compared single and double data entry. We have registered this title with the Cochrane Methodology Group and a protocol for the review will be submitted in the near future.
Title: Payment to healthcare professionals for patient recruitment to trials: a systematic review.
Authors: J Bryant and John Powell
Brief details: Many randomised controlled trials (RCTs) fail to meet their recruitment targets. One strategy to increase recruitment to trials is to pay healthcare professionals to recruit subjects either by providing financial incentives or by reimbursing excess costs incurred. Many pharmaceutical companies provide inducements but this is not common practice in publicly funded research programmes. Such programmes need to have confidence that payments are worthwhile. We did a systematic review, therefore, to synthesise the evidence on the effectiveness of payment to healthcare professionals for patient recruitment to trials. Further details.