How to Interpret Patient Safety Data—A Guide From the Nation's Largest Event Reporting Database

Abstract

Patient safety metrics are recurrent headlines. Intentional or not, they are often sensationalized, conflated, or misconstrued to tell a story in which patients are frequently harmed by irresponsible, negligent clinicians. The truth is far less dramatic. Although medical error does occur and real people suffer real harm, the vast majority of encounters go as expected. So it’s crucial to understand how to interpret patient safety metrics to distinguish true areas of concern from embellished front-page stories. It is also important to understand that each database has its own reporting criteria and each research study its own methodology, and while there is no universal definition for medical error, medical error is not synonymous with patient harm. Reported events do not necessarily equate to instances of medical error, nor are all instances of harm preventable. For example, a patient may have a serious allergic reaction to a medication that they have never taken previously. The Patient Safety Authority (PSA) is charged with capturing every occurrence of harm or potential harm to patients in Pennsylvania, whether attributable to medical error or not, and providing tools to prevent its recurrence. Since its inception in 2004, more than 3.8 million confidential event reports have been added to the PSA’s database, the Pennsylvania Patient Safety Reporting System (PA-PSRS)—the largest event reporting database in the United States and one of the largest in the world. The number of events reported into PA-PSRS has increased from 2004 to 2018; however, this was anticipated as a result of a maturing safety culture, and one cannot conclude from the data whether the actual number of events went up or the uptick is solely due to increased reporting. Caution should be given to inferences like “medical error is increasing” that cannot be substantiated from event reports. It may seem counterintuitive, but a facility with a low number of reports may be more concerning than one with a higher number, as this could indicate a culture where safety and transparency are not supported. What is certain is that since 2004 in Pennsylvania, the number of reported incidents (events without harm) has increased; the number of reported serious events (events with some level of harm) has not trended up or down; and the number of high- harm events (those causing life-threatening injury, irreversible harm, or death) has declined. Approximately 97% of the reports in PA-PSRS are incidents. These types of events are often overlooked in healthcare, as Pennsylvania continues to be the only state that requires healthcare facilities to report no-harm events. Incidents often indicate potential patient harm, and the difference between a “near miss” and a “serious event” may have been happenstance or an intervention not guaranteed to recur. Though PA-PSRS cannot conclusively address medical error incidence, its millions of datapoints provide insights into emerging trends that are unapparent to individual facilities. As such, it provides the framework for a larger system that transforms data into actionable information to reduce harm. Thorough ongoing analyses drive an education agenda, identify opportunities for collaborative improvement projects, and prioritize issues across Pennsylvania and the United States—in healthcare facilities and in individual practice. The PSA’s work in these areas is published and shared in Patient Safety and elsewhere, and read by healthcare providers in 49 states and 44 countries.