Our Post-Mortem Was So Good We Never Fixed the Bug

A thorough investigation and a cursory one arrive at the same three inadequate solutions, according to a study. In 2017, researchers led by Kathryn Kellogg examined 302 root cause analyses from a prominent university hospital over eight years. Their findings reveal that only 106 of these investigations resulted in proposed solutions, representing only roughly two-thirds of the total.

The most common proposed remedies were training (20%), process changes (19.6%), and policy reinforcement (15.2%). Over half of the suggested fixes fell into these three categories: teaching people more, adjusting procedures, and reminding staff of rules. The study concludes that the most frequently recommended solutions tend to be weaker interventions, making it less likely for recurrence to be prevented.

The repeated occurrence of the same types of harm, despite multiple root cause analyses, highlights the issue of "organizational forgetting." This phenomenon occurs when institutions remember everything in their documents but nothing in their systems. The title of this essay, "Our Post-Mortem Was So Good We Never Fixed the Bug," encapsulates the inverse relationship between the quality of analysis and the likelihood of implementing effective fixes.

While better post-mortems do not necessarily lead to stronger remedies, they do produce higher-quality documents. The psychological mechanism behind this phenomenon is hypothesized to be that thoroughness in analysis can inadvertently reduce the anxiety driving the response to failure, resulting in the resolution of the investigation when every thread is tied up.