A Qualitative Study of Screening Logs in Clinical Trials: How Can We Improve Current Practice?
This qualitative study identifies significant challenges in the inconsistent collection and reporting of screening data in clinical trials and advocates for the development and implementation of standardized guidelines to improve data quality and methodological rigor.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Every time a new medical treatment is tested on people, researchers must first find the right volunteers. This process begins long before a patient signs a formal agreement to join a study. Doctors and nurses look through their records and talk to potential participants to see if they fit the strict rules of the experiment. This initial phase is called screening. It is a crucial step because it reveals who is available, who is turned away, and why. If researchers only count the people who eventually join the trial, they miss the full picture of how the study was conducted. They cannot see if they accidentally excluded certain types of people or if the final group of volunteers truly represents the wider population. To fix this, many trials use a simple tool called a screening log, which is essentially a list that tracks everyone who is checked for eligibility, whether they join the study or not.
Despite the clear value of these lists, a new study suggests that the way they are used is often messy and inconsistent. The research team, based at the University of East Anglia, set out to understand why this data is so hard to collect and how to make it better. They looked at hundreds of published medical trials, sent surveys to the people who manage these studies and the doctors who fill out the forms, and held a series of expert discussions to reach a shared agreement on best practices. Their work highlights a gap between the need for accurate information and the reality of how it is currently gathered.
The researchers began by looking at what had already been published. They examined nearly one hundred medical trial reports from the last decade to see how many included a clear picture of the screening process. While most of these reports included a flowchart showing how many people moved through the study, only a small fraction actually defined the very first number on that chart. This top number, which represents everyone who was assessed for eligibility, was often left vague. The team also reviewed older papers that discussed the methods of screening logs. These previous studies consistently pointed out that collecting this data is difficult, time-consuming, and often lacks a standard definition. One major issue identified was that without a clear agreement on what counts as "screened," different sites in the same trial might count people differently, making it impossible to compare results fairly.
To get a direct look at current habits, the team sent surveys to two groups of people: the managers who design the trials and the healthcare professionals who actually do the work at the hospitals. The results showed that while almost everyone agreed that screening data is useful, the vast majority felt the current system could be improved. Many trial managers admitted that they often create their own versions of the log for each study, leading to a lack of consistency. Some sites do not have a template at all, while others have one but do not follow it strictly. The healthcare staff who fill out these forms reported that the process feels burdensome and repetitive. They noted that they are often not given enough time or money to do this extra work, and when they do, the data they collect is sometimes of poor quality because the instructions are unclear.
A significant part of the study involved bringing experts together in a structured series of discussions to agree on specific rules. This method, known as a Delphi survey, allowed the group to refine their ideas over several rounds until they reached a consensus. They tackled difficult questions, such as exactly who should be written down on the list. The group initially proposed a definition that the log should include anyone assessed using the study's rules, even if they never sign a consent form. However, consensus was not reached on this initial definition, as experts felt it was too vague and required greater clarity on what "assessed" meant. The definition was subsequently revised to specify "a list of potential participants and their screening outcomes, assessed using eligibility criteria as per study protocol," which did achieve consensus. Regarding specific details, the group did not reach an agreement in the first round to record demographic items like age, sex, and ethnicity. However, after clarifying in the second round that ethical approval would be obtained to hold this information without consent and that the log could be adapted to protect privacy, the respondents then achieved consensus that sex, age, and ethnicity were critical for inclusion. The group also found that there was no agreement on using alternative sources of data, such as hospital computer systems, to replace the manual lists. While some experts thought this would be more efficient, the group could not reach a consensus on whether these digital sources were reliable enough to use on their own.
The final outcome of this work is a set of draft guidelines designed to bring order to the chaos. The researchers concluded that the current practice is too variable and that the quality of the data suffers because of a lack of funding and clear standards. They argue that if the people who pay for medical research want to know if their trials are fair and representative, they must treat the collection of screening data as a priority. This means providing the necessary resources and time for staff to do it correctly. The study suggests that without these improvements, the scientific community will continue to miss important clues about bias and representation in medical research. The path forward involves creating a standard way to record this information so that every trial can tell the complete story of how its participants were found.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.