1494Scrub typhus among patients with acute undifferentiated febrile illness in a non-tropical, endemic area

Background. Scrub typhus may be a life-threatening infection that usually presents as acute undifferentiated febrile illness (AUFI). Because of the similarity of clinical features and absence of sensitive point-of-care testing, however, clinicians have difficulty in differentiating scrub typhus from other etiologies among patients presenting with AUFI in acute healthcare settings. Methods. This cross-sectional study included adult patients hospitalized with AUFI via the emergency department of a suburban Korean hospital between 2009 and 2013. AUFI was defined as any febrile illness with duration of ≤14 days, without any indication of localized infection at the time of initial presentation. Using epidemiological, clinical, and laboratory data, cases of scrub typhus, which were confirmed by indirect immunofluorescence antibody of ≥1:80 in a single serum or a ≥4-fold increase in paired sera, were compared with those of other AUFI etiologies. A multiple logistic regression model identified risk factors associated with scrub typhus to develop a clinical prediction rule (CPR) by summing up their regression coefficients. Its performance was checked by c-statistic. Results. Of 382 AUFI cases, etiology was identified in 262 (68.6%): influenza (26.7%) was the most common cause, followed by acute hepatitis A (24.0%), scrub typhus (20.6%), hemorrhagic fever with renal syndrome (4.6%), primary bacteremia (4.2%), and Plasmodium vivax malaria (3.4%). Risk factors associated with scrub typhus included age of ≥65 years (simplified regression coefficient, 1), outdoor activity within the preceding 30 days (1), onset of the illness during the outbreak season of scrub typhus (2), myalgia (1), or eschar (1) (P < 0.05 for all). The c-statistic of CPR was 0.971 (95% CI, 0.951–0.991), and that with bootstrapping samples of 1,000 repetitions was 0.971 (95% CI, 0.952–0.990). For the cut-off value of ≥3, our CPR showed the sensitivity of 94.4% (95% CI, 83.7–98.6%) and specificity of 82.2% (95% CI, 76.2– 87.0%). Conclusion. Scrub typhus is one of the leading causes of AUFI in a non-tropical, endemic area. Our CPR can assist clinicians to identify immediately cases with scrub typhus among adult patients presenting AUFI in acute healthcare settings, with a high sensitivity and specificity. Disclosures. All authors: No reported disclosures.