This article is the second part of a two-part series on forecasting potential recruitment. To read part I, click here.
In the second part of this series, we provide you with an in-depth step-by-step guide on how to build an enrollment funnel that accurately forecasts how many inquiries you would need at the top of your funnel in order to hit your enrollment goal.
Determine the Stages of Your Enrollment Funnel
The first step of building a data-driven enrollment funnel is simple: you have to map out the stages of enrollment.
The stages of enrollment may vary a little depending on your specific process, and may also vary a little depending on the source of the patients. The steps should also take into account any kind of advertising strategy or external recruitment source.
Here’s an example of what a traditional clinical trial enrollment funnel might look like:
Once you’ve established the stages of your enrollment funnel, it’s time to move on to the next step.
Determine the Percentage of Patients Who Will Pass Each Stage
The next step is to determine the percentage of patients who will pass each stage in your enrollment funnel. There is a general hierarchy of data sources you should use to figure out how many patients will drop-out at each stage, which we’ve illustrated in the graphic below:
If your study is underway, then you can simply calculate how many patients are dropping out at each stage of your enrolment funnel from the data you have to date, which should be a reliable estimate moving forward.
If the study hasn’t started yet, then the best thing you can do is look at previous similar studies. You want to start collecting this data to make this process easy. You can get started by recording the number or percentage of people that dropped out at each stage of the enrollment funnel in a spreadsheet.
If you haven’t been collecting this information, you can also find this data from previous studies in sources like phone screen logs and subject logs.
If you can’t get data from previous similar studies then you can rely on benchmarks, like those we’ve provided as average statistics. However, benchmarks are less reliable because there is no guarantee your study is an average one. Of course, you can always rely on gut feel or intuition without consulting any data at all; however, this is when you will fall victim to the optimism bias. We cover optimism bias in Data Doesn’t Lie, Part I: Clinical Trials Enrollment Forecasting—Is It Worth It?.
Although relying on previous studies is a more reliable benchmark, if you really do not have any alternatives, we’ve provided some guidance for estimating your enrollment funnel in the table below.
Be aware that the number of patients that pass phone screening and clinic screening can vary the most. Fortunately, this is purely based on the eligibility criteria, and you’re much more likely to have this data recorded.
Determine the Number of Potential Patients Needed at the Top of Your Recruitment Funnel
Once you’ve determined the percentage of patients who will pass each step, all you have to do is start with the number of enrollments and then just work backward. You can easily do this in a spreadsheet. We will demonstrate how to do this in the following two examples, and what the numbers might look like for an easy study versus a more difficult one.
Example of Recruitment Funnel for an Easy or Average Study
Below is an example of the enrollment funnel breakdown for a fairly straightforward study.
Easy Study Example
Based on the above example, if you require 10 patients to enroll in your study, and your data suggests that only 50% of people will pass the screening visit, then you will need 20 people to attend the screening visit. If you deduce that only 90% of participants will actually attend the screening visit, then you will need 22 participants to pass phone screening. And so on.
By working your way backwards all the way to the number of inquiries needed in the first step, you will know that you need around 91 inquiries for 10 enrolled patients. So, for this example, the recruitment funnel will look something like this:
Recruitment Funnel Example for a Difficult Study
Unfortunately, not all studies are as straightforward as our previous example. Below is an example of a clinical trial enrollment funnel for a more difficult study.
Hard Study Example
Based on the above example, if you require 10 patients to enroll in your study, and your data suggests that only 10% of people will pass the screening visit, then that makes things a bit trickier—you will need a much bigger number of people at the top of your funnel, i.e. 1,984 people.
So let’s say you need 100 people to attend the screening visit, in order to get your 10 enrollments, and that you know only about 20% of those will pass phone screening. This means that you will need 556 people to pass the Interested step of your funnel. What’s more, you know that only 40% of people will continue on to the Interested step—this means that you will need roughly 1,389 people to pass your Successful Call Back step.
If only 70% of people pass the Successful Call Back step, then you will require 1,984 inquiries. And so, this recruitment funnel would look something like this:
Although only 10 enrollments were required for both our example studies, the number of inquiries differed greatly between the two studies. This is why collecting data on how many people pass through every phase of your recruitment funnel is essential for successful recruitment.
Evaluate How Many Subjects Each Potential Recruitment Strategy Will Provide
Once you know how many inquiries you need for your trial, the final step is to assess each recruitment strategy that you might use, and determine how many inquiries each strategy can generate.
You can log all this information in a spreadsheet, as demonstrated below.
For example, let’s say you want to enroll 10 patients in a type 2 diabetes trial. You know that from previous advertising of similar type 2 diabetes studies, you will need about 500 inquiries to reach your enrolment goal. In addition, you have an advertising budget of $5,000 and about three months to recruit. You can then take a look at the different strategies, and how they measure up against each other in terms of time frame and cost and then make a pretty straightforward decision on whether that number of enrollments is realistic, and how you can reach it.
Maintaining the Data Source Hierarchy
It’s much easier to estimate how many inquiries you might get from a newspaper ad, for example than how many randomizations you will get from a newspaper ad—there are simply fewer steps to account for an inquiry versus randomization.
However, the same hierarchy of data sources should be considered when determining how many inquiries a recruitment strategy will generate:
Always use current study data over previous similar studies. If you do not have any data from the current study or any previous study, you can rely on other benchmarks like statistical averages.
Sometimes, advertising vendors might be willing to provide you with an estimate of the number of inquiries; however, it is imperative to understand how the vendors have calculated their estimates. You want to ensure they’ve drawn their calculations from previous data rather than their own intuition or gut feel.
Lastly, while you can rely on your intuition, we strongly recommend that you don’t—relying on gut feeling will only cause you to fall prey to the optimism bias.
Recruitment Does Not Have to Be Hard
Successful recruitment starts with doing your due diligence. Once you start collecting data on how many people pass each stage of your recruitment funnel, and how many inquiries you generate from each recruitment source, forecasting recruitment numbers become a lot more straightforward.
If you collect information from every one of your studies, you will then be able to use data from previous similar studies to forecast your enrolment funnel. You will also be able to determine which recruitment strategies will get you enough patients at the top of your recruitment funnel to reach your clinical trial enrollment goal.
Understanding the numbers behind recruitment not only removes the guesswork involved in forecasting but also allows you to meet your study recruitment goals on time and within budget, as you will know whether your goals are realistic ahead of time.
If you want to find out how many inquiries you would need to successfully recruit for your study, contact us and we will go through our Due Diligence process free of charge.