Abstract
BACKGROUND: Pneumonia is a major cause of mortality and morbidity, but the development of new antimicrobials is lacking. Radiological assessment of pneumonia severity may serve as an effective intermediate endpoint to reduce barriers to successful completion of antimicrobial trials. We sought to determine whether the Radiologic Severity Index (RSI) correlated with mortality and healthcare resource utilisation in patients with acute leukaemia undergoing induction chemotherapy. METHODS: We measured RSI (range 0-72) on all chest radiographs performed within 33 days of induction chemotherapy in 165 haematological malignancy patients with pneumonia. Peak RSI was defined as the highest RSI score within 33 days of induction. We used extended Cox proportional hazards models to measure the association of time-varying RSI with all-cause mortality within the first 33 days after induction chemotherapy, and logistic regression or generalised models to measure the association of RSI with total daily cost and healthcare resource utilisation. RESULTS: After adjustment for clinical variables, each one-point increase in RSI was associated with a 7% increase in all-cause 33-day mortality (HR 1.07, 95% CI 1.05 to 1.09, p<0.0001). Peak RSI values of 37.5 or higher were associated with 86% higher daily direct costs (p<0.0001), more days in intensive care unit (9.9 vs 4.8 days, p=0.001) and higher odds for mechanical ventilation (OR 12.1, p<0.0001). CONCLUSIONS: Greater radiological severity as measured by RSI was associated with increased mortality and morbidity in acute leukaemia patients with pneumonia. RSI is a promising intermediate marker of pneumonia severity and is well suited for use in antimicrobial trials.