Doxycycline for the empiric treatment of low-severity hospital acquired pneumonia

Abstract

Background In the U.K., doxycycline is widely recommended as first-line empiric treatment for mild HAP, however this practice is based on expert opinion and we are aware of no data describing the outcomes of patients treated with doxycycline. Here we describe the outcomes of non-ICU patients treated empirically with doxycycline for low-severity HAP. Methods 1680 inpatient chest x-rays were manually screened to identify cases of HAP. HAP was defined as new or progressive CXR consolidation occurring ³48 hours after admission, combined with compatible symptoms or signs. Treatment failure was defined as requirement for antimicrobial escalation, HAP recurrence or mortality attributed to HAP. We compared groups according to treatment outcome using the Kruskal-Wallis test and receiver operator characteristic (ROC) analysis. Results Forty-nine patients who received doxycycline were included in the analysis. The median age was 78 years and 63% had >2 co-morbidities. Hypoxia was common (57%) but extra-pulmonary organ dysfunction was uncommon. 71% of patients were successfully treated with doxycycline as first-line empiric therapy. Treatment failure was associated with increased duration of hospitalisation prior to HAP onset (median 21.5 vs. 10 days, p=0.03) and higher neutrophil count (10.1 vs. 6.1 x109L-1, p=0.04). ROC analysis identified HAP onset >14 days after admission as the optimum cut-off for predicting treatment failure. Conclusions In this cohort, the majority of patients with low-severity HAP were successfully treated with doxycycline. Treatment failure was associated with prolonged hospitalisation prior to HAP onset with 14d identified as a potential surrogate marker for patients requiring antimicrobials against Gram-negative organisms.