Metrics were all the rage in clinical research in the 1990s. At the time, I worked for a mid-sized consulting firm that pharma companies hired to measure their clinical trial performance. We would mobilize a small army of consultants to gather data and produce tables, pareto graphs, and spider charts calculating work volume and cycle times for completing clinical studies, and then compare that company to other companies of its size.  Afterwards, we would launch process improvement initiatives to reduce those cycle times. By the time we finished all our initiatives, the vogue for metrics had passed, and we never did go back and measure whether our efforts reduced cycle time, or if reducing cycle time had any positive impact on the business. Times have changed, but as an industry we're still flummoxed by metrics. We know that metrics can help us maintain oversight and gauge quality, but we're not always sure what we're supposed to be reviewing or what to do with it. I frequently recommend an excellent book on this subject, How to Measure Anything by Douglas W. Hubbard, which makes the case for measuring only when "it informs a significant bet of some kind" - in other words, use metrics only when you expect to take action based on the information gained. For this reason, I find it's helpful to think of metrics in terms of what you are trying to accomplish (objective) vs. what you are trying to measure. In running a clinical study, we are generally trying to accomplish one of the following objectives:

• Course-correct by assessing progress against plan and applying adjustments to make up time, optimize resources, or improve quality
• Mitigate risks by obtaining early warning of potentially costly errors and implementing corrective and preventative actions
• Add value by encouraging behavior that improves patient safety and data integrity
• Inform planning by incorporating learnings from other studies

  In the next few blog posts, we'll look at different types of metrics that can help us accomplish objectives.