Clinical outcomes of doxycycline, azithromycin and chloramphenicol for the treatment of patients with severe scrub typhus.

• All individual complications including shock, ARDS, meningitis and acute kidney injury were found to be significantly associated with treatment failure and time to defervescence, after adjusting for the effects of age, sex, initial antibiotic type, and interval from symptom onset to start of antibiotic treatment. • Treatment failure risk was higher for azithromycin than doxycycline for patients with meningitis; and treatment failure risk was higher for chloramphenicol than doxycycline for patients with acute kidney injury, pneumonia or shock. • Antibiotic resistance occurred in 2.4% of doxycycline group, 6.3% of azithromycin group and 10.4% of chloramphenicol group. Deaths occurred in 29.4% (5/17), 0% (0/5) and 10.0% (1/10) of patients, respectively, with no significant differences between groups (P =0.232). Previous studies have evaluated treatment efficacy of various antibiotics for patients with mild-to-moderate scrub typhus (ST). However, the efficacy of different antibiotics for treating severe ST remains uncertain. A retrospective study of patients with severe ST was undertaken in China. The treatment efficacy rates of doxycycline, azithromycin and chloramphenicol were compared, using treatment failure and time to defervescence as primary outcomes. In total, 876 patients with severe ST who initially received doxycycline, azithromycin or chloramphenicol were recruited. The treatment failure rate did not differ significantly between patients receiving doxycycline and patients receiving azithromycin (6.0% vs 11.4%; P =0.109). However, a higher treatment failure rate was observed for chloramphenicol compared with doxycycline (14.6% vs 6.0%; P =0.004). No significant difference in time to defervescence was observed between patients receiving doxycycline, azithromycin or chloramphenicol. Further subgroup analysis revealed a higher risk of treatment failure for chloramphenicol compared with doxycycline in patients with acute kidney injury, pneumonia and shock; and a higher risk of treatment failure for azithromycin compared with doxycycline in patients with meningitis. Significant correlation was found between azithromycin resistance and meningitis (P =0.009), and between chloramphenicol resistance and acute respiratory distress syndrome (ARDS) (P <0.001) using Cramer's V correlation coefficient. Multi-variate Cox regression analysis revealed significant associations between time to defervescence and presence of ARDS, shock, myocarditis, meningitis and acute kidney injury. Azithromycin and doxycycline were found to have significant therapeutic effects in patients with severe ST. In contast, chloramphenicol was less efficacious for the treatment of these patients. [ABSTRACT FROM AUTHOR]