Facebook
  1. Home
  2. Insights
  3. Blog
  4. Patient Centric Monitoring: Preventing and Learning From Mistakes

Patient Centric Monitoring: Preventing and Learning From Mistakes

Patient Centric Monitoring: Preventing and Learning From Mistakes

When validating the 1,000,000 data points that compose the typical Phase III trial, focusing on errors that don’t matter easily leads to wasted resources. The industry’s initial response — risk based monitoring — more efficiently detects errors and risks, but can still expend resources on errors irrelevant to a trial’s validity and miss opportunities to detect and solve critical quality issues.

At ICON, we’ve drawn upon other industries’ methodologies for risk mitigation to evolve RBM into a more useful approach for protecting your clinical data. One of those methodologies, Human Factor Analysis (HFA), is employed by NASA, Ford, the U.S. EPA, major airlines, and several countries’ militaries to identify and mitigate human error in large, variously trained work forces. HFA uses uniquely structured datasets to reveal underlying behaviours and factors that are otherwise difficult for humans to sense or reconstruct, but ultimately are the root cause of an incident.

For example, take an airline that experienced several mid-air collisions. An HFA investigation, driven by datasets in which behaviours have been classified by analysis, may reveal that in most cases crashes are caused by a subset of miscommunications and gaps in teamwork among pilots and air traffic control. With an accurate understanding of the root cause of these crashes, airlines could mitigate risk by deploying tactical decision-making training or SOP adaptations that address the human errors that ultimately increased collision risk.

ICON has incorporated human factor analysis and into its risk based monitoring approach, which we call Patient Centric Monitoring. ICON’s information platform, ICONIK, systematically classifies and analyses the causes of trial errors to help CRAs deploy the right risk-mitigation resources to the sites that need them most.

 

Patient-Centric-Monitoring

How does ICON use HFA to improve upon traditional RBM?

Let’s consider a site that has multiple issues with informed consent. A traditional monitoring approach would log a problem, potentially classify the site as high risk, and deploy general retraining on informed consent. ICON’s Patient Centric Monitoring draws upon HFA to instead look for trends in the centralised data, revealing that consent errors became more frequent after that site had enroled 100 or more patients. Upon CRA investigation, human factors analysis reveals that the issue was not a deficit in training, but in resources. To mitigate risk, the project team may limit the site to no more than 75 patients and provide additional investigator support focusing on resource allocation.

To identify deeper issues that could jeopardise trial integrity, such as protocol noncompliance, ICONIK transforms site data into visual analyses that enable smarter monitoring.

For example, in a clinical trial for which blood pressure (BP) is a measure of safety or efficacy, Patient Centric Monitoring utilises ICONIK’s real-time data to detect patterns that indicate problems with BP data collection. In the real trial data depicted in Figure 1, a lower-than-expected number of unique BP measurements (Figure 1a, black dots) revealed that some sites, including site 509 (Figure 1b), were deviating from the approved protocol by reporting rounded measurements.

An analysis for the human factors that led to the deviation revealed that the root problem was not a lack of training, but rather a technical issue with that site’s equipment. The BP meters automatically rounded BP values to the nearest 5 mmHg and, due to this being normal practice at the site, hospital staff did not know how to change the equipment settings. With this knowledge, ICON’s CRAs were then able to educate staff on how to change the device settings or replace the devices that could not be adapted, thereby preventing further error in this and future trials. With a traditional monitoring approach, the CRA would see the data and initiate site retraining without further investigation into the BP measurement distribution.

References:

1. Schuler, P., & Buckley, B. Re-Engineering Clinical Trials. pp 9-10. (Elsevier, London, 2015). Clara Heering

In this section

Connect with us