Lisa Wickert of IMARC Research writes an interesting article for Medical Design and Outsourcing about the FDA’s Clinical Investigator Site Selection Tool (CISST). Wickert reports that the primary goal of the CISST is to develop a science-based approach for site selection to ensure that even with limited resources, sites that pose the highest risk to public health are selected for inspection. Understanding the CISST turns out to be a big deal for all those concerned. Apparently in fiscal year 2017 the FDA performed over 900 inspections through the Bioresearch Monitoring (BIMO) Program. Approximately 70 percent of these inspections targeted clinical sites. Wickert notes that undoubtedly the FDA brings method and approach as to where and when it will announce an inspection during a clinical trial—for both sponsor and clinical site. Interestingly, Wickert introduces the FDA’s CISST.

Wickert provides a valuable reference to Jean Mulinde, senior policy advisor in the Office of Compliance and the FDA’s CISST. Ms. Mulinde utilizes the presentation to explain how the CISST is used to rank clinical sites by compiling data on a number of attributes submitted by sponsors to compile the “Clinsite” data set. Ms. Wickert also introduces us to recent FDA draft guidanc outlining the electronic submission of data in standardized format for CDER (Center for Drug Evaluation and Research) to plan BIMO inspections. Wickert reports that FDA includes study-wide and site-specific information including:

Study Level (clinical trial type, design)

Site Level (enrollment numbers, protocol deviation occurrences, adverse events, investigator track record with prior inspections, financial disclosures and more)

The FDA compiles all of this data; assigns weighted values and then executes a three-part calculation including:

  1. Attribute Raw Values
  2. Risk Functions applied to Attribute Values
  3. Hierarchical Weighting Schema Applied

As noted by Wickert, the final output represents an overall risk score—the higher the score the more likely an inspection. Can FDA inspections become predictable in the future? Undoubtedly sponsors and smart vendors will start working on developing advanced algorithms to do just that.

Source: Medical Design and Outsourcing

Pin It on Pinterest