We’ve mentioned before in a blog post how our data science team downloads the ClinicalTrials.gov database on a weekly basis to keep our clients updated on what’s new in clinical development. Another advantage to collecting these numbers each week, is to analyze clinical trial registration trends over time.
The end of June marked our 26th week this year downloading this dataset, and analyzing the updates. The halfway point of the year seemed like the perfect time to take a step back and ask “How are clinical trial registrations trending in 2016, compare to previous years?”.
When you look at the numbers, it’s clear that there is a downward trend in clinical trial registration happening. After digging into some information about the clinical trial submission process however, I want to show you why this isn’t only a premature conclusion, but potentially a wrong one.
First, look at the data. Here’s new clinical trial registrations by month, spliced by each year for the past 4 years:
While January and February were on par with the registration volume of previous years, March-June figures in 2016 are clearly not.
Why It's Misleading
Looks can be deceiving, and bar graphs are no exception. Because of the ClinicalTrials.gov registration process, trials are dated by when they are submitted to the database, and not when they are approved. In other words, the numbers above reflect only trials that have been approved by the NIH (the team running ClinicalTrials.gov), sorted by the month they were submitted. What is not reflected in the numbers are these two important groups:
- Trials that have been submitted but not yet approved
- Trials that have been denied
When looking at the above chart to find a trend in clinical trial registrations rates, we need to be clear that we are looking at the approval rate of trials, not the submission rate.
If we dive deeper into the submission process, we’ll find that some trials take a few months to be approved. Look at our charts from some other weekly updates.
We see that a number of January 2016 and February 2016 trials are still being added throughout May and into June. This is because those trials have a submission date in January or February, but it took 3-4 months for them to be approved. In the same vein, there should be a number of May and June trials that have been submitted, but won’t show up in the ClinicalTrials.gov database until August or September.
This means it’s inaccurate to look at the most recent 4 months to assess a trend in the rate of clinical trial registrations.
If we look beyond this 4-month window to January and February 2016, we see that there are actually a greater number of new clinical trials registered than January and February of previous years. If the rest of 2016 follows this trend, 2016 will see more new clinical trials (criteria being they’re registered on ClinicalTrials.gov) than previous years, but it's too early to say for sure.
There are 3 key takeaways here:
- Using January and February as indicators, there may be more new clinical trials registered this year than the previous 3 years.
- Our team’s initial assumption that the submission process takes 1 month is wrong. In many cases it actually takes 3-4 months for a submitted trial to be approved.
It's important to look into the qualitative aspects of data when assessing clinical trial trends. In this case, information about NIH’s approval process prevented us from drawing the wrong conclusion about the current trend in clinical trial registration rates.
If you are interested in getting weekly updates on all new submissions in ClinicalTrials.gov including sponsor, enrollment, and contact information about these trials, signup for a free trial our TrialFinder software by submitting the form below. We aim to make it easy for you to keep tabs on the 400 new trials that are approved by the NIH each week.