Healthcare is not as safe as it should be.

The Institute of Medicine published the seminal book To Err is Human 20 years ago which highlighted the fact that 98,000 people died each year from medical mistakes. It called on the healthcare industry and government to begin making changes. Specific attention was given to ensuring that well-designed processes of care were established to prevent, recognize, and quickly recover from errors so that patients were not harmed.

Fast forward to 2016 at the Johns Hopkins School of Medicine where they just completed a comprehensive study evaluating U.S. hospital mortality records and determined that over 250,000 patients died each year from adverse medical incidents in hospitals. The figure was 10% of all U.S. deaths per year and meant it was more dangerous to be admitted to the hospital than to drive a car in traffic.

In the past two decades, medical technology has improved, diagnostics have improved, and medical training has improved, yet the adverse incident rate has remained high.

Contrast this with aviation. From 1970 to 2018 the commercial aviation industry reduced its fatal accident rate from 6.35 per million departures to 0.39 per million. That is a 16-fold decrease in the accident rate. 

How did they accomplish this? As an industry, emphasis was placed on creating safer aircraft and navigational systems, and improving crew training. 

Experts agreed that advances in engineering were important, but they felt aircrew behavioral changes were the most important factor on incident rate reduction. The difference between these two industries was that the aviation industry encouraged consequence-less reporting of potentially unsafe practices and crew members responded with reams of information. As an industry, aviation aggregated this data and recognized that pilot errors were causal factors in 60% of all accidents. Special focus was placed on training crew members on the identification of leading indicators that developed prior to the accidents. The intent was to give the crews the information they needed to break the error chain prior to the accident occurring.

Arkenstone is applying this model for healthcare consumers by digitally collecting adverse incident leading indicators in real-time, applying machine learning to identify potential accident factors and then providing hospital caregivers with timely information to allow them to break the adverse incident error-chain.

Our solution uses multiple technologies to accomplish this. The current standard for adverse incident recognition is IHI’s Global Trigger Tool, which identifies adverse events that have already happened. With standard EHR accessibility, it is possible to analyze patient data through the digitized IHI measures and apply a risk score to individual patients. However, this will only tell you that an event is happening or has happened. We are focused on preventing the events from happening. Our solution has identified leading indicators of these events and we capture real-time consumer experience data from the facility. This allows us to accurately predict events that may happen in time to take action and break the error chain.

To learn more, contact me at