An accident can destroy a hospital's reputation, and figuring out how it happened can be an incredibly complex task. A respected luminary on the subject shared insights on the lessons that sometimes provide a silver lining to these traumatic situations, at the seventh annual Clinical Engineering Association of Illinois (CEAI) conference.

Dr. Larry Fennigkoh, professor of biomedical engineering at the Milwaukee School of Engineering, detailed a number of medical device-related incidents as examples of the intersection between human and technical factors that generally combine to result in these catastrophes.

Fennigkoh, a veteran of health care forensics engineering, exhorted all those who are involved in the design, use, management and maintenance of medical devices to recognize their obligation to learn from such cases. He said any assessment of such tragedies has to be based on intricate understanding of both human and technological elements.

Accidental deaths and device incidents rarely triggered by single causes

Current estimates of accidental deaths associated with hospital care are at 400,000 a year, according to Fennigkoh, and there are an additional 100,000 reports of medical device incidents annually. These numbers, however, may just be the tip of the iceberg, since the Office of the Inspector General suggests that hospital staffs underreport events by up to 86 percent.

Of those reported, more than one-third are ultimately deemed to involve user error. But Fennigkoh argued that a large percentage of user error cases can also be linked to equipment design — so both factors must be taken into account in every instance.

Drawing from his experience investigating numerous injuries — especially shocks, burns, fires, and deaths — associated with a variety of medical and non-medical electrical devices, Fennigkoh described how complex interactions can be when patients connected to multiple machines literally “become one with them.”

Without full recognition of this complexity, incidents often lead to accusations between hospitals over defective design and manufactures asserting user error. Fennigkoh summarized numerous heart-wrenching cases that, upon close examination, proved that the causes are rarely so simple.

Case studies from medical-technical forensic investigations

An infusion pump free-flow incident that once triggered a near-fatal overdose of epinephrine was caused by neither a defective door assembly nor a staff member mistakenly leaving it open, but a separate valve being moved to the open position without a clamp on the line, said Fennigkoh.

Only intricate analysis of the combination of human and technical factors could have brought that particular cause to light, and enabled all parties to work toward a long-term procedural solution.

In another case, an EKG machine provided two reports of a mild conduction defect and one of a heart attack. The family physician ignored the latter, and the patient died three days later from myocardial infarction.

Deep comparative analysis of the three strips demonstrated that the algorithms of interpretive electrocardiographs can result in both false positives and negatives when measurements come in near their mathematically determined diagnostic threshold. Therefore, Fennigkoh asserted, clinicians should not rely exclusively on the machine interpretations of ECGs. They should always be read over by a cardiologist.

The death of an infant during a CPAP sleep study lead to a traumatic search for causes by heartbroken hospital staff, parents, as well as manufacturers. In the end, forensic investigation showed that there were no problems with the medical devices or their application. The infant had a congenital birth defect that likely contributed to an ill-fitting face mask that shifted and caused oxygen desaturation.

Eliminate the tendency to blame

Hazards to patient safety, according to Fennigkoh, are always a combination of inadvertent “holes” lining up despite a series of defenses erected by manufacturers, administrators, clinicians and engineers. The way to minimize them is to collaborate across all these tiers without a culture of mutual recrimination.

The lessons of these and other cases, he concluded, are to treat every adverse event, injury, and accidental death as a precious learning opportunity; to identify the active failures and latent conditions within the system that contributed the event; to break from the habit of managing the last error; and, above all, “eliminate the tendency to blame.”