Clinical Intelligence Analytics - Study Trends

Study Trends

Study Trends dashboard (Figure 4.1) gives us insights into the tends in recruiting country, average registration to enrollment duration and average study duration over the past 2 decades.

Figure 4.1

The data legends are shared between the charts on the first row. Similarly, the legend is same for the 2 charts on the bottom row.
The data is aggregated on a study level and a study is classified as an International (Both US and non-US), US-only and non-US only study based on the countries of patient recruitment. The analysis excludes studies that does not have any recruiting country information which could be for various reasons such as the recruitment might not have started or the study may not have any enrollments yet.
An international study means that it has recruited patients in the US and at least one other country. 
A non-US only study means that the study has recruited the patients in countries other than the US and no patient was enrolled in the US.
For the other 2 charts on the bottom row, the numbers with a minus sign are positive numbers but are shown as negative numbers just for the display purpose as they are on the opposite side. However, this is not intended. The vertical lines are the average lines to show the distance.

The dashboard can provide insights into the following trends:

1. What's the share of studies by recruiting country?
55% of the studies registered in the US have recruited the patients only from countries other than the US. The share of studies that recruited the patients only inside the US is 39%. International studies are just 6%.


2. What is the trend in the study registration by their location of recruitment?
Now that we know something about the share of the studies based on where the patients were recruited, let's take a look at how it has changed over the last 2 decades. The share of international studies has reduced to less than 5% now from 15% in early 2000's. The trend of US-only studies have also declined sharply from lower 90% in early 2000's to 30-35% level now. The trend of non-US studies have increased consistently in past 2 decades from 44% in year 2005 to 66% in 2019.  
Remember that the share of the studies registered in the past years can change based on the studies that are still recruiting or may recruit in future. Hence, the current trend is the snapshot.

3. How has the average time taken study registration to enrollment changed over the last few years?
The chart shows the trend in the time taken from study registration to patient enrollment or study initiation. The chart also compares side-by-side the duration for studies sponsored by industry or non-industry sponsors.
There are many studies that were registered retrospectively, meaning the studies were registered after they were already started (the first patient was already enrolled). Such retrospective studies were excluded from the analysis. Only for the prospective studies, the registration to enrollment duration is calculated in days.  
It appears that it usually took longer, sometimes 1.5-3 times, for non-industry sponsored studies to begin a study after they are submitted. The trend for non-industry sponsored studies is following a parabolic curve. The average registration to enrollment duration for the non-industry sponsored studies is 125 days which is higher than the overall industry average of 107 days. For the most part, the yearly trend is close to the average line except in year 2010 having an average of 145 days, which is also the highest in recent years.
On the other hand, the industry sponsored studies are initiated quickly and the study initiation duration has improved slightly overall. The average for industry sponsored studies is around 80 days which is well below the industry average of 107 days. The average for the last few years has been consistently near the average line.   

4. How has the average study duration changed over the last few years?
The non-industry sponsored studies takes longer to complete as compared to industry sponsored studies. The average study completion duration for non-industry sponsors is close to 3.5 years which is considerably much higher than the overall industry average of 2.6 years. On the other side, the average for industry sponsors is little above 2 years. The good news is that both type of sponsors have made a significant improvement is past 15 years to bring down the average completion duration to lower levels, possibly signalling great improvements in overall operational efficiency in study conduct.
With that positive note, see you till next time. 

Clinical Intelligence Analytics - Study

Study Dashboard

In study dashboard (Figure 3.1), we will look at certain aspects of study at aggregated level as well as at a study level. 

Figure 3.1

The study dashboard will try to answer following questions:

1. What is the average study completion duration for sponsors from Industry and non-Industry?

For all types of studies (All), the sponsors from Industry completed the studies in about 1.9 years. In comparison to that, sponsors from non-industry took almost 3 years to complete the study.

The observational studies (Obs) took longer to complete. The Industry sponsors with an average of 2.3 years performed fairly better than the non-industry sponsors with an average of 3.2 years.

For interventional type of studies (Int), the average study took 1.8 years for Industry sponsor as compared to 2.9 years for non-industry sponsors.

We may further want to look at the study duration by the phase of the study. Phase 3 studies are large scale and complex in nature and hence, it should take longer to complete when compared with phase 1 and phase 2 studies. Let's see what we find. Only interventional studies go through the drug development phases. If you take a look at Avg Study Duration by Phase chart, Phase 2 studies took longest among all the study phases with an average of 3.3 years. Phase 3 studies took an average of 2.9 years to complete where as phase 4 studies took 2.4 years. Early phase 1 studies took longer than the phase 1 studies.

2. What is the share of sponsors from industry and non-industry in interventional or observational studies?

Almost 80% of studies were interventional studies. 56% of 80% which 70% of total interventional studies were sponsored by non-industry sponsors. The industry sponsors have greater share in interventional studies as compared to its share in observational studies.

3. What percentage of studies were completed between 0 to 3 years or between 8 to 10 years?

40% of the studies were completed between 1 to 3 years. Around 29% studies were completed in less than 1 year. 

4. Which studies took the longest to complete?

There are 44 studies (0.02%) that took more than 30 years to complete. The study that took longest was sponsored by Johnson & Johnson to evaluate the efficacy of oral Levofloxacin in the treatment of chronic Bronchitis. This study took 63 years to complete starting in 1931 and completing in 1994 and has enrolled 367 patients. 

5. At study level, how many medical conditions a particular study is conducted?

See the tabular report to view the number of enrollment, medical conditions and the number of study sites and countries of subject recruitment.

6. In how many countries and facilities did a study recruited patients?

See the tabular report.

The dashboard will show the description of the selected study.

Clinical Intelligence Analytics - Trends by Study Attributes

In the last post, we looked at the growth trends of studies registered, initiated, completed and posted results over the period of last 20 years. We looked at yearly trends and then drilled down at quarterly and monthly trends and compared the growth in current year with the previous year.
In this post, we will see the growth trends of registered studies by study attributes like study type, study phase, Drug/Device and DMC flag over the period. I have filtered out the studies where the study attributes were not specified.

Figure 2.4

This dashboard (Figure 2.4) is an extension of growth trends:
1. Study Submission Trend by Study Type-
The chart shows the trend of studies submitted for interventional and observational types.
The share of interventional studies have decreased. In early 2000's, there were around 90-93% interventional studies and 7-10% observational studies. In last few years, the share of observational studies have increased to 21%.

2.  Study Submission Trend by FDA oversight-
The chart shows the trend of studies by the oversight of the FDA, if the study is for FDA regulated drug or a device. 
Until 2008, the share of the studies for FDA regulated devices was less than 10% which has increased to around 25% now.

3. Study Submission Trend by Study Phase-
 Phase 3 studies are important studies since sponsors apply for FDA approval after that. 
Study type NA are the studies that do not have a phase of a study, and I guess it is mainly for the device related studies. These studies share have increased significantly over the period and I think its because the device related studies share have increased as we observed in the previous chart.
There is a slight drop in the share of phase 1 studies. The share of Phase 2 studies in green has decreased from 43% in 2003 to 12% in 2018. Phase 3 studies have also followed the same trend with the share reducing to just 7% in 2018 from 25% in 2005. Phase 4 are post marketing studies and the share has reduced from 16% in 2005 to 7% in 2018.

4. Study Submission Trend by DMC-
The DMC flag tells if the study has a Data Monitoring Committee appointed or not. There were many studies that did not mention if they have DMC or not. As I mentioned in the beginning, the chart is the representation of only studies that has the data and others were filtered out.
The share of DMC appointed studies increased in the first few years and then start to drop until 2008 after which it picked up again a bit and maintaining the 40% level until 2013 but falling down a bit to 34% in 2018.
Feel free to share your thoughts.
See you soon in the next post.