Résumé
We have performed a case-control analysis to determine the significance of clinical, laboratory and epidemiological features as predictive factors of rickettsioses among patients in Sangkhla Buri, Thailand (Thai-Myanmar border). Fifteen serologically-confirmed rickettsiosis patients including Spotted Fever Group (SFG) rickettsioses, scrub typhus, and murine typhus were classified as 'cases'; one hundred and sixty-three acutely febrile patients presenting to the same hospital during the same time period, who had no serological evidence of acute rickettsiosis, were classified as 'controls'. Patients' report of rash/arthropod bite [Odds ratio (OR) 22.90, 95% CI (confidence interval) 6.23, 84.13] and history of jungle trips (OR 5.30, 95% CI 1.69-16.62) were significant risk factors. Elevated ALT (OR 3.04, 95% CI 1.04, 8.88) and depressed platelet count (OR 3.38, 95% CI 1.13, 10.10) were also useful differentiating markers of rickettsioses in this population. Definitive diagnosis of rickettsioses is difficult without specialized diagnostic capabilities that are rarely available in remote areas such as Sangkhla Buri, where other acute febrile illnesses with similar presentation are commonly found. The relative importance of predictive factors presented here may provide clinicians with some useful guidance in distinguishing rickettsioses from other acute febrile illnesses. Timely administration of empiric treatment in highly suspicious cases can deter potential morbidity from these arthropod-borne infections.