A New Prognosis Score to Predict Mortality After Acute Pneumonia in Very Elderly Patients

Acute pneumonia (AP) induces an excess of mortality among the elderly. We evaluated the value of a new predictive biomarker index compared to usual prognosis scores for predicting in-hospital and 1-year mortalities in elderly inpatients with AP.

Retrospective study in 6 clinical departments of a university hospital.

Burgundy university hospital (France).

All patients aged 75 and over with AP and hospitalized between January 1 and June 30, 2013, in the departments of medicine (5) and intensive care (1) of our university hospital.

A new index, which we named UBMo, was created by multiplying the uremia (U in the formula) by the N-terminal-pro-brain natriuretic peptide (NT-proBNP) plasmatic rate (B), divided by the monocyte count (Mo).

Among the 217 patients included, there were 138 community-acquired pneumonia, 56 nursing home-acquired pneumonia, and 23 hospital-acquired pneumonia. In-hospital and 1-year mortality rates were respectively 19.8% and 43.8%. In multivariate analysis, Pneumonia Severity Index (PSI), unlike CURB-65 (confusion, urea >7 mmol/L, respiratory rate ≥30 breaths/min, blood pressure <90 mmHg systolic or ≤60 mmHg diastolic, age ≥65) score, was associated with in-hospital and 1-year mortalities. UBMo index performed better than PSI and CURB-65 scores in predicting both in-hospital and 1-year mortalities. For in-hospital mortality, the areas under the receiver operating characteristic curves (AUCs) were 0.89 (95% CI = 0.84-0.94), 0.72 (95% CI = 0.65-0.80), and 0.63 (95% CI = 0.54-0.72), respectively, for the 3 scores. For 1-year mortality, the AUCs were 0.93 (95% CI = 0.89-0.98), 0.66 (95% CI = 0.59-0.74), and 0.58 (95% CI = 0.50-0.66), respectively, for the 3 scores. The cut point for the UBMo index of 20,000 × 10^-9 ng·mmol/L had a sensitivity of 93.1% and 80.9% and a specificity of 76.3% and 95.8%, respectively, for in-hospital and 1-year mortalities.

If confirmed by prospective studies, the UBMo index appears very efficient in identifying patients at high risk of in-hospital and 1-year mortalities after an AP.