Losing the losses: understanding the reasons for attrition in randomised trials and developing the evidence to prevent it

The purpose of randomising patients in to clinical trials is to improve the evidence-base by recording data on each patient to allow the relevant clinical question to be answered. Ideally once a patient is randomised in to a clinical trial they should continue to participate in the trial until the final follow up visit with data on their progress collected as necessary. Often not all patients randomised in to a trial will contribute data. When this happens there is a discrepancy in the number of patients randomised and the number included in the final analysis. This is termed missing data. Missing data may occur for a variety of reasons, including because the participant (i) dropped out of the trial before completing treatment and did not provide any primary outcome data, (ii) dropped out after completion of treatment but prior to primary outcome data collection, or (iii) provided partial data on the primary outcome e.g. not all questions were answered in a symptom score questionnaire.
Missing data increases the costs of conducting research, decreases the volume of concurrent research possible, and can negatively impact on the validity of research conclusions.
Most trials experience a level of missing data, also called attrition, but levels of missing data higher than 5% give rise to concern in terms of inefficiency of the trial process and potential bias if the reasons for missing data differ according to the arm of the trial.
Due to the negative impact of missing data researchers are interested in ways to promote participant retention so that as many patients as possible randomised in to the trial contribute data. There are many different approaches to maximising participant retention. A survey of clinical trials units identified that a variety of unproven strategies are being used with the aim of maximising retention with consequences on time and costs. The only way to know which ones are effective is to evaluate those approaches and establish an evidence base.

This project will improve understanding of the reasons for the failure to collect complete outcome data on all randomised patients and identifying strategies that might be used to prevent this. It will support the moves to minimise waste in research and to ensure that trials are as reliable and robust as possible, with one of the key ways to achieve this being the retention of all recruited patients in the study and the collection, analysis and reporting of a complete set of outcomes for them. This will be done by using both quantitative and qualitative research methods. This will benefit the research community by establishing a coherent approach to retention research resulting in reduced costs of research, increased capacity to conduct research and increased validity of research results upon which decisions made by all stakeholders can be made reliably.

Missing data is known to negatively impact research. It can increase costs to conduct research, reduce capacity to conduct research either due to financial restrictions or finite patient populations and can bias the conclusions of research.
During the conduct of a clinical trial its performance is usually monitored by the Research Networks in terms of its progress in recruiting to target but there has been a less coordinated emphasis placed on monitoring retention of randomised participants. Recruiting and randomising people who are not subsequently retained for the primary outcome measure can actually be worse for the analysis of the trial than not randomising that patient at all. This means that an emphasis on strategies to monitor and boost recruitment alone could be badly misplaced, if the extra patients recruited are not retained. In a Delphi survey of UKCRC registered clinical trials units retention was highlighted as a priority for research.
Approaches to maximise retention thereby minimising missing data should be informed by reasons for loss of participants to final analysis. Reasons may vary depending on characteristics of condition under study, participant types and settings. This project will:

1)Systematically review reasons for missing data in a cohort of published trials
2) Identify strategies to address missing outcome data by conducting an evaluation of NIHR trials and by collating knowledge and experience related to retention strategies within the network of CTU Directors.
3) Conduct qualitative interviews with clinical trial staff to further understanding of current practice regarding monitoring of missing outcome data in ongoing trials
4)Conduct a Delphi survey across the network of CTU Directors to assess priority interventions to pursue

One of the most frequent problems in clinical trials is the failure to achieve the target number of participants with outcome data recorded, leading to a potential waste of resources and unnecessary delays in resolving the uncertainty that prompted the trial. This project will seek to improve understanding of the reasons for the failure to collect complete outcome data on all randomised patients and identifying strategies that might be used to prevent this. It will support the moves to minimise waste in research and to ensure that trials are as reliable and robust as possible, with one of the key ways to achieve this being the retention of all recruited patients in the study and the collection, analysis and reporting of a complete set of outcomes for them.
The beneficiaries of this work are widespread. In the longer term users of research, for example clinicians, patients and policy makers, will benefit from reduced levels of missing data thereby maximising the validity of research conducted. In the shorter term funders of research and those conducting research will be able to target priority interventions to maximise retention and generate a much needed evidence base to guide activities and placement of available resources. The results of this work will be of equal benefit to trialists within the commercial and publicly funded sectors within the UK and internationally.
Much work has gone in to the creation of the research infrastructure within the UK. This is to establish the UK as an efficient and attractive environment to conduct research by providing the infrastructure to allow high-quality clinical research to take place. This is important in making the UK an attractive proposition for the pharmaceutical, biotechnology, contract research organisation (CRO) and other companies thereby maximising the development of safe and effective medicines and formulations for the benefit of patients. Improving retention will benefit all stakeholders thereby supporting the aims of the UK research infrastructure and attracting industry partners.