Using routine blood test results to predict the risk of death for emergency medical admissions to hospital: an external model validation study

Abstract

The Biochemistry and Haematology Outcome Model (BHOM) relies on the results from routine index blood tests to predict the patient risk of death. We aimed to externally validate the BHOM model. Method We considered all emergency adult medical patients who were discharged from Northern Lincolnshire and Goole (NLAG) hospital in 2014. We compared patient characteristics between NLAG (the validation sample) and the hospital where BHOM was developed. We evaluated the predictive performance, according to discriminative ability (with a concordance statistic, c), and calibration (agreement between observed and predicted risk). Result There were 29 834 emergency discharges of which 24 696 (83%) had complete data. In comparison with the development sample, the NLAG sample was similar in age, blood test results, but experienced a lower mortality (4.7% vs 8.7%). When applied to NLAG, the BHOM model had good discrimination (c-statistic 0.83 [95% CI 0.823 - 0.842]). Calibration was good overall, although the BHOM model overpredicted for lowest (<5%, observed = 229,predicted =286) and highest (≥50%, observed = 31, predicted = 49) risk groups, even after recalibrating for the differences in baseline risk of death. Conclusion Differences in patient case-mix profile and baseline risk of death need to be considered before the BHOM model can be used in another hospital. After re-calibrating for the baseline difference in risk the BHOM model had good discrimination but less adequate calibration.