Predictors of acute hospital mortality from a 23-year database of emergency medical admissions

There is much focus on the management of the acute emergency medical admission; however, comparative outcome data over long time periods are sparse.

We report all the outcomes of emergency admissions over 23 years (2002-2024).

Demographic and clinical details are described. Predictors of 30-day in-hospital mortality were analysed with logistic and Cox regression.

The 2002-2024 cohort consisted of 186,004 admissions in 95,192 patients. The 30-day in-hospital mortality per admission averaged 3.8% (95% CI 3.7% to 3.9%); there was a relative risk reduction (RRR) between 2002 and 2024 of 55.3%, from 5.5% to 2.4% (p = 0.001), with a calculated number needed to treat (NNT) of 33. Per patient mortality (single admission only considered; last admission if > 1) averaged 7.5% (95% CI 7.4% to 7.7%); there was a RRR of 77.9% between 2002 and 2024, from 13.1% to 2.9% (p = 0.001) with NNT of 9.9. Total in-hospital deaths were 14,092 at respective age for males of 73.6 (SEM 0.12) and females of 79.4 (SEM 0.11). The strongest predictors of short-term 30-day in-hospital mortality were being older (> 70 years)-OR 2.05 (95%CI 1.94, 2.18) and the acute illness severity score 1.91 (95%CI 1.85, 1.96). Other predictors were the Charlson Index 1.32 (95% CI 1.28, 1.37) and a primary neurological MDC (major disease category) 1.18 (95% CI 1.09, 1.27).

The 30-day in-hospital mortality for emergency medical admissions has improved over time.