Estimating the burden of scrub typhus: A systematic review.

BACKGROUND: Scrub typhus is a vector-borne zoonotic disease that can be life-threatening. There are no licensed vaccines, or vector control efforts in place. Despite increasing awareness in endemic regions, the public health burden and global distribution of scrub typhus remains poorly known. METHODS: We systematically reviewed all literature from public health records, fever studies and reports available on the Ovid MEDLINE, Embase Classic + Embase and EconLit databases, to estimate the burden of scrub typhus since the year 2000. FINDINGS: In prospective fever studies from Asia, scrub typhus is a leading cause of treatable non-malarial febrile illness. Sero-epidemiological data also suggest that Orientia tsutsugamushi infection is common across Asia, with seroprevalence ranging from 9.3%-27.9% (median 22.2% IQR 18.6-25.7). A substantial apparent rise in minimum disease incidence (median 4.6/100,000/10 years, highest in China with 11.2/100,000/10 years) was reported through passive national surveillance systems in South Korea, Japan, China, and Thailand. Case fatality risks from areas of reduced drug-susceptibility are reported at 12.2% and 13.6% for South India and northern Thailand, respectively. Mortality reports vary widely around a median mortality of 6.0% for untreated and 1.4% for treated scrub typhus. Limited evidence suggests high mortality in complicated scrub typhus with CNS involvement (13.6% mortality), multi-organ dysfunction (24.1%) and high pregnancy miscarriage rates with poor neonatal outcomes. INTERPRETATION: Scrub typhus appears to be a truly neglected tropical disease mainly affecting rural populations, but increasingly also metropolitan areas. Rising minimum incidence rates have been reported over the past 8-10 years from countries with an established surveillance system. A wider distribution of scrub typhus beyond Asia is likely, based on reports from South America and Africa. Unfortunately, the quality and quantity of the available data on scrub typhus epidemiology is currently too limited for any economical, mathematical modeling or mapping approaches.