Failure of a sepsis predictive model used in over 50% of the US hospitals

Tackling our widespread laboratory data usability problems has not become a universally shared priority (as evidenced for example, by a recent  failure to perform of one of the most widely implemented sepsis prediction models in U.S. hospitals).[i, ii, iii] Achieving true interoperability of clinical laboratory data will require a careful use case analysis of the sepsis prediction model being used by more than half of the hospitals in the U.S.

[i]. Wong A, Otles E, Donnelly JP, et al. External Validation of a Widely Implemented Proprietary Sepsis Prediction Model in Hospitalized Patients. JAMA Intern Med.
Published online June 21, 2021. doi:10.1001/jamainternmed.2021.2626
[ii]. Habib AR, Lin AL, Grant RW. The Epic Sepsis Model Falls Short—The Importance of External Validation. JAMA Intern Med.
Published online June 21, 2021. doi:10.1001/jamainternmed.2021.3333
[iii]. Shielded from scrutiny, Epic algorithms deliver inaccurate information
https://www.statnews.com/2021/07/26/epic-hospital-algorithms-sepsis-investigation/

Ref iii provided by Andrea Pitkus