Hospital mortality, withdrawal of life-sustaining therapy decisions and early secondary brain insults for critically ill traumatic brain injury patients in England, Wales and Northern Ireland (2009-2024): an observational cohort study.

BACKGROUND

Recent epidemiological studies reported conflicting results regarding mortality trends for traumatic brain injury (TBI) patients. Mortality trends for the critically ill TBI population, and their drivers of changes, remains understudied. Particularly, withdrawal of life-sustaining therapy (WLST) decisions were rarely evaluated concurrently. In this study, we aimed to describe hospital mortality and WLST trends over the past 15 years in England, Wales and Northern Ireland for TBI patients admitted to an intensive care unit (ICU).

METHODS

Observational cohort study, involving 235 adult ICUs participating in the Intensive Care National Audit & Research Centre (ICNARC) Case Mix Programme (CMP). From April 1, 2009 to March 31, 2024, all TBI patients were included. Comparator cohorts consisted of patients with trauma, sepsis, and vascular brain injury recorded in the CMP. The primary outcome was hospital mortality. The secondary outcome was the incidence of WLST decisions. We also examined the proportion of patients experiencing predefined early secondary brain insults.

FINDINGS: Of the 2,324,961 ICU admissions, we identified 45,684 unique TBI patients. Over the study period, hospital mortality for TBI patients increased from 25.6% (1021/3988) to 35.0% (1306/3727). The proportion of WLST decisions rose from 7.5% (301/4024) to 19.7% (759/3850). After adjustment for main confounders, multivariable analyses confirmed these trends. No similar trends were observed among the comparator cohorts. TBI patients were exposed to hypotension, hypocapnia, hypercapnia and hyperglycaemia in 49.8% (22,559/45,298), 29.9% (12,356/41,262), 33.6% (13,869/41,262) and 29.2% (12,127/41,505) of cases, respectively. Half of patients (50.3%, 20,747/41,265) were exposed to hypoxaemia, and this proportion increased markedly from 36.9% (1359/3684) to 61.2% (2186/3572) over time.

INTERPRETATION: For critically ill TBI patients, hospital mortality and WLST decisions rates increased over time. These findings raise important questions regarding the processes and ethical frameworks underpinning WLST decisions.

FUNDING: UKRI, NIHR, UK Ministry of Defence, Alzheimer's Research UK, French Society of Anaesthesiology and Critical Care, Gueules Cassées Foundation, INNOVEO donation fund.