Invasive ventilation and mortality in critically ill nonagenarians: a retrospective cohort study

TL;DR

This study examines the use of invasive ventilation in ICU patients aged 90 or older and finds that it is associated with higher mortality and illness severity.

Contribution

The study identifies specific predictors of mortality in nonagenarians receiving invasive ventilation and suggests ventilation as a trigger for goals-of-care discussions.

Findings

Invasive ventilation was associated with higher ICU and hospital mortality in patients aged ≥90 years.

Prolonged invasive ventilation (>72 hours) and elevated SOFA scores were significant predictors of hospital mortality.

One-year mortality was significantly higher in ventilated patients compared to non-ventilated patients.

Abstract

ICU admissions among very old patients are increasing. Invasive ventilation (IV) is common, but its benefit in patients aged ≥ 90 years is uncertain given high mortality and ethical concerns.

This retrospective cohort study analysed all consecutive ICU patients aged ≥ 90 years admitted between 2008 and 2023 at a tertiary care centre in Germany. Demographic, clinical, and outcome data were extracted from electronic health records. Multivariable logistic regression was used to identify predictors of hospital mortality, while Kaplan–Meier survival analysis and Cox proportional hazards regression were used to assess predictors of 1-year mortality.

Of 113,950 patients, 1422 (1.25%) aged ≥ 90 years were identified (median 92 years, IQR 91–94; 66% female). IV was administered to 434 patients (31%), while 988 (69%) were not invasively ventilated. Median ICU length of stay was 1.7 days (IQR 1–4) overall. Among ventilated patients, the median duration of IV was 13 h (IQR 4–44), and 66% received IV for less than 24 h. IV was associated with higher illness severity at admission (SOFA score 9 [IQR 7–11] vs. 2 [IQR 1–4], p < 0.001), longer ICU stays (2.9 days [IQR 1.1–7.4] vs. 1.5 days [IQR 0.9–3.1], p < 0.001), as well as higher requirement of vasopressors at admission (0.112 µg/kg/min [IQR 0.056–0.278] vs. 0.072 µg/kg/min [IQR 0.038–0.133], p < 0.001) and renal replacement therapy (3.5% [n = 15] vs. 1.7% [n = 17], p = 0.042). In patients requiring IV, ICU and hospital mortality were 35.7% (n = 155) and 49.3% (n = 214) vs. 11.5% (n = 114) and 21.5% (n = 212) in non-IV patients, respectively (both p < 0.001). Independent and significant predictors of hospital mortality included prolonged IV duration (> 72 h: OR 4.01), peak lactate ≥ 4 mmol/l within 72 h (OR 6.67), as well as elevated SOFA scores (4–7: OR 1.96, ≥ 8: OR 3.46), and CCI ≥ 3 (OR 1.74) at admission. One-year mortality risks were 73.2% (95% CI 68.5–77.3%) and 53.4% (95% CI 50.0–56.6%) for IV and no-IV patients (p < 0.001), respectively.

In this selected cohort of ICU patients aged ≥ 90 years, invasive ventilation–particularly beyond 72 h–identified a subgroup with very high ICU and hospital mortality and greater illness severity. These observational data support using invasive ventilation in nonagenarians as a trigger for early, patient-centred goals-of-care discussions, rather than as evidence that ventilation itself causes excess mortality.

The online version contains supplementary material available at 10.1186/s40001-026-03928-6.

Figures