An automated real-time method for the detection of patients at risk for malignant hyperthermia

Background: Malignant hyperthermia (MH) is a rare anesthetic pharmacogenetic disorder that can be difficult to detect in its earliest phases. Prompt treatment is known to improve outcomes. The modern anesthesia information management systems (AIMS) collect enormous amounts of data. However, data lack context and are not able to provide real-time guidance. Utilizing our AIMS, we developed the capacity to incorporate decision support. Aims: We describe the creation and evaluation of a real-time detection tool for MH. Methods: Cases of MH from 2003 to 2013 were retrospectively reviewed to confirm the diagnosis of MH and to calculate a MH clinical grading scale score. The index cases were utilized to develop four electronic data Rules, based on endtidal CO₂(ETCO₂) and temperature criteria. The Rules were then applied retrospectively to the index cases and to a full cohort of general operating room (OR) patients from January 2012 to June 2012. If criteria for possible MH was met, the detection tool generated an alert on the monitor at the patient's bedside. Results: We identified seven patients with confirmed MH with MH Scores ranging from 28 to 70. Using four data Rules, all of our seven index cases were captured and generated an appropriate alert. Median time from MH computer alert time to dantrolene administration time among our index cases was 23 min (IQR 17–77). There were 938 false-positive alerts for possible MH (1.8%) when the Rules were applied to a general OR cohort of 51 579 total cases from January 2012 to June 2012. Conclusions: We demonstrated a real-time MH detection tool based on established physiologic criteria that is sensitive enough to capture cases suspicious for MH, while limiting false positives to prevent alarm fatigue. This has the potential to notify the provider of possible MH such that treatment may be rapidly initiated.