Human Development Index and outcomes in older critically ill patients: A European multicentre study

Older adults represent an increasing proportion of intensive care unit admissions, but the relationship between country-level human development and outcomes after critical illness remains incompletely understood.

We conducted a secondary analysis of three prospective multicentre registries, VIP1, VIP2, and COVIP, including acutely admitted older ICU patients with available Clinical Frailty Scale assessment, country-level Human Development Index (HDI), and 30-day vital status. VIP1 and VIP2 enrolled patients aged ≥80 years, whereas COVIP enrolled patients aged ≥70 years. The primary exposure was exceptionally high human development, defined as HDI ≥ 0.90 versus <0.90. The primary outcome was 30-day mortality. Associations were assessed using logistic regression with robust standard errors clustered by country, adjusting for age, sex, SOFA score, frailty, admission diagnosis, organ support modalities, and treatment-limitation decisions. Exploratory mediation analyses examined selected ICU management variables as potential pathways linking HDI to mortality.

Among 9920 patients included in the primary analysis, 8324 (83.9%) were treated in countries with HDI ≥ 0.90 and 1596 (16.1%) in countries with HDI < 0.90. Thirty-day mortality was lower in high-HDI countries than in lower-HDI countries (40.0% vs. 53.3%). In unadjusted analysis, HDI ≥ 0.90 was associated with lower 30-day mortality (OR 0.58; 95% CI 0.38-0.90; P = 0.016). This association persisted after multivariable adjustment (adjusted OR 0.49; 95% CI 0.31-0.80; P = 0.004) and was similar after additional adjustment for study cohort (aOR 0.49; 95% CI 0.31-0.77; P = 0.002) and ICU bed capacity (aOR 0.50; 95% CI 0.29-0.86; P = 0.013). When modelled continuously, higher HDI was associated with lower mortality after full adjustment (OR 0.33 per 0.10-unit increase; 95% CI 0.19-0.56; P < 0.001). Exploratory mediation analyses suggested that lower use of invasive mechanical ventilation in high-HDI countries may partially contribute to the observed association (NIE OR 0.86; 95% CI 0.83-0.89). Mediation analyses involving treatment-limitation decisions were more difficult to interpret because these decisions are closely linked to prognosis, clinical trajectory, and end-of-life practice.

In this large European cohort of older critically ill patients, treatment in countries with exceptionally high human development was associated with lower 30-day mortality. The association persisted after adjustment for patient-level severity, frailty, treatment limitation, organ support, study cohort, and ICU bed capacity. These findings suggest that country-level development and ICU management patterns, particularly invasive ventilation practices, may contribute to outcome differences. Because this was an observational secondary analysis using country-level exposure data, causal interpretation should remain cautious.