Sponsors have long debated whether or not to collect pre-screening and screening data for sites participating in their trials. Understandably this is a burden on the site, having to record screening efforts in real time and upload them to the sponsor on a weekly, bi-weekly or monthly basis. However, the utility of this data in shaping and pivoting your recruitment program in invaluable. Beyond the use of this dataset in the current trial, sponsors can leverage this data to better understand patient populations in this indication as well as better assess sites for future trials.
Inevitably there will be sites that will sporadically enter pre-screening data, generally when nudged by patient recruitment specialist (check out our previous post!) or clinical research associate. This sporadic data will still be useful for understanding their patient population and practices.
Due to the efforts required on the site’s behalf sponsors may receive requests to provide a small monthly stipend to reimburse the site for the time spent. This will pay off in dividends for you as a sponsor when you are pivoting and focusing your recruitment plan / program.
So, what is pre-screening and screening data?
Pre-screening and screening data is selected data collected on each patient being considered at a clinical trial site for participation in the trial.
Typically, you will collect data on which selection criterion the patient fails, though also consider the following:
Refusal of consent
Staff unavailable for enrollment
Native language consent unavailable
IP or clinical supplies unavailable
Patient unavailable- Scheduling
Having a set list from which sites can choose reasons for non-enrollment will reduce the amount of coding required on the sponsors end.
Let’s have a look at a simple pre-screening data set:
Along the left hand side are a selection of the many criteria which may be selected for a patient’s non-enrollment into this study. This is a cumulative listing though these data can be analyzed on a weekly, monthly or annual basis.
An important note is that comparisons of #’s between sites is not advised. Each site has a different definition of a “pre-screened patient”. For one site, it may be any patient that walks through their clinic door, while for another it may be any patient presenting with the top 2-3 selection criteria.
Let’s see what conclusions we could draw from a simple data set such as this.
Yellow: The first thing we notice is that across the board the most common reason for non-enrollment is Exclusion Criterion #5, a co-morbidity. Noticing a trend such as this would prompt a sponsor to revisit that criterion. Why is it in place? Safety? Efficacy? Could we afford to remove this criterion, absorb some risk, and vastly expand out patient pool?
Green: A closer look at site 02 shows us that patients commonly fail as they have moved beyond their maximum timeframe on a concomitant medication. Why is this site not identifying these patients earlier? Furthermore, we see there are 2 instances where staff were unavailable for enrollment. When taken together we can infer that this site may be experiencing some staffing issues. How could we help? Provide a stipend for an additional study coordinator?
Red: Site 05 has a high rate of refusal of consent, relative to the number of patients they are screening. This may be a site which could use some additional counselling on conducting informed consent. A call from the patient recruitment specialist should discuss their local practices for informed consent and offer some suggestions as to how to improve. Perhaps the inexperienced study coordinator is conducting consent discussion, and requires some shadow time with the PI or training on the consent process.
Purple: A simple look at site 10 and we can see three patients were missed due to not having a native language consent available. This is a simple fix. The translation fees may be steep but every patient counts!
For more information on patient recruitment, check out Advantage Clinical's Patient Recruitment for Sponsors and CROs short course.