Operational Metrics in Clinical Trials, Part 1
Posted on Thu, Jul 22, 2010 @ 01:23 PM
The Use of Consistently Defined Clinical Trial Metrics – Part 1
Jennifer Price, CDISC CDASH Registered Service Provider (RSP)
If you can not measure it, you can not improve it. –Lord Kelvin
The Metrics Champion Consortium (MCC) develops metrics that allow trial sponsors and CROs to share a common set of clinical trial performance metrics across studies.
This not-for-profit group is solving a very complex problem in a very elegant, straightforward way. They gather industry experts to identify and define a common set of clinical trial metrics that can be used in all studies. See a recent Metrics Champion Consortium press release here.
The idea around using clinical trial metrics sounds so simple, but is not often realized primarily because metrics are not often defined consistently. For example, Study ABC123 might collect number of pages entered, number of queries answered and total subjects enrolled. Study XYZ987 might collect a percentage of expected pages that have been completed, queries resolved and total subjects screened. These are all standard metrics to collect, but are not the SAME metrics across clinical studies.
The MCC has put together common metrics for clinical trials performance, labs, ECGs and Images, and is now working on process improvement metrics.
Clinical Trial Performance Metrics Example
Below is a subset of the details around one of the clinical trial performance metrics: Receipt of Query Response to Database Update Time. There are additional details and columns in the official document provided by MCC for their members that are not shown here.
Metric: Receipt of Query Response to Database Update Time
Definition – This is crucial to define exactly how a clinical trial metric is measured so the same data is compared when looking across studies. I like that the Median number of days is used for this particular metric instead of the mean. The mean could be influenced by a few long outstanding updates, where the median is more representative of how many days it takes to make updates as a result of a query.
Additional analysis on a ‘for cause’ basis – Identify why this metric is important and what process to look at if the metric is out of range.
Reporting Frequency – For some clinical trial metrics such as this one, it is important not to look at the data too often, but to take a ‘big picture’ look at the entire process.
Target – Identify a target range for the particular metric. This may differ based on your company or processes (particularly EDC or paper data collection). By identifying a target goal for this metric, it is easy to see if we are on track, or if we are falling outside our target range.
Business driver / benefit statement – It is important to identify the problem that can be solved by collecting this metric.
By using industry standard metrics across all our clinical studies, the clinical trial manager or data manager can identify potential areas for improvement that will benefit the entire study team.
Part 2 will identify how these clinical trial metrics can be used to maximize their impact.
Jennifer Price is on the MCC ‘Process Improvement Metrics Development Team.’