To estimate the ability of commonly measured laboratory variables to predict imminent (within the same or next calendar day) medical emergency team (MET) calls, ICU admission or death.Methods
We performed a retrospective observational study of ED patients. We estimated the ability of each laboratory variable or combination of variables together with patient age to predict imminent MET calls, ICU admission or death. We externally validated our findings in patients from a different hospital.Results
We studied 160 341 batches in 71 453 ED patients (average age: 59.9 ± 22.1 years) for a total of 1 million individual measurements. There were 341 MET calls, 160 ICU admissions from the wards and 858 deaths. Multivariable modelling achieved a receiver operating characteristic area under the curve (ROC-AUC) of 0.69 (95% CI 0.63–0.74) for imminent MET call with prediction occurring a mean of 11.9 h before the call. Additionally, it achieved a ROC-AUC of 0.82 (95% CI 0.73–0.87) for imminent ICU admission. Finally, it achieved a ROC-AUC of 0.90 (95% CI 0.87–0.91) for imminent death. When tested using an additional 37 367 batches from a cohort of 21 430 ED patients from a second teaching hospital, the multivariate model achieved a ROC-AUC of 0.70 (95% CI 0.66–0.73) for imminent MET call, a ROC-AUC of 0.84 (95% CI 0.78–0.90) for imminent ICU admission. Finally, it achieved a ROC-AUC of 0.89 (95% CI 0.86–0.91) for imminent death.Conclusions
Commonly performed laboratory tests can help predict imminent MET calls, ICU admission or death in ED patients. Prospective investigations of the clinical utility of such predictions appear desirable.