Prolonged ICU Stay Doubles Hospital Mortality, Global Prediction Model Falls Short in Indian Patients: Study

While previous research in high-income countries defines persistent critical illness as the point when pre-existing characteristics out-predict acute admission diagnoses, it remains unclear if this pattern holds in lower-resource settings; consequently, Dr. Bharath Kumar Tirupakuzhi Vijayaraghavan and colleagues from the Indian Registry of Intensive Care (IRIS) evaluated this phenomenon to determine how acute and antecedent patient factors impact long-term prognostic accuracy in India.

Therefore, in the registry-embedded cohort study evaluating 42,925 critically ill adult patients across 56 Indian ICUs, investigators utilized generalized mixed-effects models to compare the predictive value of acute physiologic derangements versus antecedent characteristics like age and comorbidities on hospital mortality over time.

Key Clinical Findings of the Study Includes:

Diminishing Acute Predictors: Investigators found that the prognostic accuracy of acute illness characteristics for predicting hospital mortality declined significantly over time, dropping from an Area Under the Receiver Operating Characteristic (AUROC) curve of 0.76 at admission to 0.69 by day seven.

Poor Baseline Predictors: Researchers noted that antecedent characteristics, such as age and the Charlson Comorbidity Index, consistently demonstrated poor prognostic performance throughout the ICU stay, maintaining a low AUROC of 0.56.

Escalated Mortality Risk: Authors showed that patients remaining in the ICU beyond seven days faced a significantly higher hospital mortality rate of 33.1% versus 14.6% for those with shorter stays, alongside an increased ICU mortality of 30.3% compared to 13.2%.

Intensified Organ Support: Scientists reported that prolonged ICU stays correlated with a substantially higher need for interventions, including invasive ventilation (51.7% vs. 19.5%), noninvasive ventilation (18.4% vs. 7.6%), vasopressors (34.7% vs. 19.1%), and kidney replacement therapy (11.4% vs. 6.1%).

The results suggest that as intensive care lengths of stay increase, the predictive power of acute physiological derangements wanes without a corresponding rise in the predictive value of baseline health traits.

Although extended stays beyond a week lead to escalated mortality rates reaching 33.1% and greater reliance on life support, the traditional high-income definition of persistent critical illness is not replicated in this region.

Thus, the study concludes clinicians that standard prognostic models relying heavily on pre-existing comorbidities may not accurately predict outcomes for patients experiencing prolonged critical illness in Indian intensive care units, highlighting the need for tailored predictive approaches in this demographic.

While the research provides vital insights, the exact reasons why the high-income model of persistent critical illness failed to replicate in this cohort remain to be fully elucidated, gently pointing toward the need for future localized studies to better define prognostic markers for long-term critical care patients in developing nations.

Reference

Tirupakuzhi Vijayaraghavan, B. K., Rashan, A., Ramakrishnan, N., Haniffa, R., Beane, A., Adhikari, N. K., Lone, N., de Keizer, N., & the Indian Registry of Intensive Care (IRIS) Collaborators. (2025). Persistent Critical Illness Among Intensive Care Patients in India: A Registry-Embedded Cohort Study. Critical Care Medicine, 53(8), e1641-e1649.