The moment a patient is recruited to participate in a clinical trial, we start to collect data. Much of this data derives from the information that only patients can give, that we cannot acquire through any test or scan. Are they experiencing pain? What are the side effects? These important insights are captured in a trial using PROs (patient-reported outcomes).
The moment a patient is recruited to participate in a clinical trial, we start to collect data. Much of this data derives from the information that only patients can give, that we cannot acquire through any test or scan. Are they experiencing pain? What are the side effects? These important insights are captured in a trial using PROs (patient-reported outcomes).
As of this morning, there are 12,892 active randomised controlled trials across the globe. These trials, often complex in their nature, share a common reliance on understanding and evaluating the patient perspective. Many of them use patient-reported outcomes (PROs).
The collection of PROs is essential for the development of pharmaceutical treatments and the implementation of new healthcare interventions. Trial results tell us what is working, and what is not, from the patient’s point of view.
Or at least, they should.
This week, the PRO Research Group at the University of Birmingham published two papers in the scientific journal PLOS ONE, highlighting a disturbing lack of consistency in the design of trials that include PROs.
In the first paper we report that the design of PRO data collection across trial protocols is commonly uneven, and frequently substandard. This poses a real problem: PROs need to be collected in a consistent manner otherwise we are at risk of producing poor quality data that is potentially biased.
Our second paper reveals the potential cause of this inconsistency: a vast number of confusing guidelines spread across various resources, from online archives to academic texts and beyond. Our team trawled through over 20,000 papers to discover 162 separate PRO recommendations. Only five (3%) appeared in more than half of the data, highlighting a lack of consensus.
It is no surprise, therefore, that researchers struggle to achieve a consistency of method and delivery with such a tangled web of information, often contradictory, on how to design, collect and interpret their trial.
The presence of such inconsistency in a trial compromises the reliability and validity of PRO trial results, at best weakening the confidence with which PRO findings may be used to inform patients, clinicians, drug approval committees and policy makers; at worst misleading clinical or health policy decision-making, reducing the value of patient trial participation and wasting research resources.
Waste is a concern: we do not have an excess of funding for medical research. With each tightening of the purse strings we become ever more aware of the necessity for cost-efficient and effective practice.
A comprehensive review of how we collect PROs in trials is required. There is an urgent need to develop clearly defined guidelines on how best to acquire and analyse PRO data which are internationally endorsed, highly visible and easily accessible to researchers, funders and research ethics committees. The lack of such guidelines risks continued deficiencies in PRO study design, undermining important patient-centred trial results.
Sloan and colleagues (2007) aptly sum up the importance of PROs in research: ‘We must do all that we can to make patient-reported outcome assessment feasible and credible. If we fail in our task we will have left out the heart of all healthcare research: the patient.’
A failure to heed this warning would represent a significant backward step by the research community.
Dr Derek Kyte and Professor Melanie Calvert
PRO Research Group, School of Health and Population Sciences,
University of Birmingham
