Predicting the failure of noninvasive ventilation in hypoxemic patients by updated HACOR score

ABSTRACT

Background: Acute hypoxemic respiratory failure is a major cause of morbidity and mortality requiring ventilatory support. Non-invasive ventilation (NIV) is an alternative to invasive ventilation, and the HACOR score is widely used to predict NIV failure. The updated HACOR score incorporates additional baseline variables such as ARDS, septic shock, immunosuppression, organ failure, and the SOFA score, but evidence on its predictive utility remains limited.

Methods: This prospective observational study included adult ICU patients who were started on NIV for hypoxemic respiratory failure. Baseline data included age, sex, indication for NIV, comorbidities, SOFA score, and the presence of COVID-19, septic shock, and ARDS. Non-invasive ventilation failure prediction was assessed using original and updated HACOR scores at baseline, 1–2 hours, 12 hours, and 24 hours. Diagnostic performance was compared using the area under the curve (AUC).

Results: A total of 63 patients were included. The AUCs for original vs updated HACOR scores were 0.522 vs 0.404 at baseline, 0.899 vs 0.404 at 1–2 hours, 0.862 vs 0.957 at 12 hours, and 0.485 vs 0.500 at 24 hours. The difference in AUCs was statistically significant at 1–2 hours and 12 hours (p = 0.0001), but not at baseline or 24 hours.

Conclusion: The updated HACOR score, which incorporates six baseline variables and the five original scale items demonstrated improved predictive performance for NIV failure compared to the original HACOR score at certain timeframes. This suggests that the performance of both scores varies over time in predicting NIV failure.

Keywords: Acute hypoxemic respiratory failure, Non-invasive ventilation (NIV), updated HACOR score, NIV failure