Research Article: A Spatiotemporal Database to Track Human Scrub Typhus Using the VectorMap Application

Date Published: December 17, 2015

Publisher: Public Library of Science

Author(s): Daryl J. Kelly, Desmond H. Foley, Allen L. Richards, Archie C. A. Clements.

Abstract: Scrub typhus is a potentially fatal mite-borne febrile illness, primarily of the Asia-Pacific Rim. With an endemic area greater than 13 million km2 and millions of people at risk, scrub typhus remains an underreported, often misdiagnosed febrile illness. A comprehensive, updatable map of the true distribution of cases has been lacking, and therefore the true risk of disease within the very large endemic area remains unknown. The purpose of this study was to establish a database and map to track human scrub typhus. An online search using PubMed and the United States Armed Forces Pest Management Board Literature Retrieval System was performed to identify articles describing human scrub typhus cases both within and outside the traditionally accepted endemic regions. Using World Health Organization guidelines, stringent criteria were used to establish diagnoses for inclusion in the database. The preliminary screening of 181 scrub typhus publications yielded 145 publications that met the case criterion, 267 case records, and 13 serosurvey records that could be georeferenced, describing 13,739 probable or confirmed human cases in 28 countries. A map service has been established within VectorMap ( to explore the role that relative location of vectors, hosts, and the pathogen play in the transmission of mite-borne scrub typhus. The online display of scrub typhus cases in VectorMap illustrates their presence and provides an up-to-date geographic distribution of proven scrub typhus cases.

Partial Text: VectorMap is a spatial database of vector and host collection records, distribution models, and vector-borne disease (VBD) maps that provide a “one-stop shop” for exploring vector hazard and VBD risks on a global scale. Based on MosquitoMap [1] and building on SandflyMap [2], VectorMap uses ESRI’s ArcMap Server 10 (Redlands, CA) and MS Silverlight (Redmond, WA) to serve distribution data on mosquitoes, sand flies, ticks, fleas, mites, and hosts and reservoirs. As part of the establishment of the new mite vector disease service (aka MiteMap) on the VectorMap Silverlight viewer that maps the location of mite (chigger) collection data, we have endeavored to include published data on scrub typhus distribution. Scrub typhus (aka mite-borne typhus or tsutsugamushi disease) is an acute febrile disease endemic within the countries of the Asia-Pacific Rim. It is caused by the bite of Orientia tsutsugamushi-infected Leptotrombidium spp. larvae (i.e., chiggers). Scrub typhus is regarded as “probably the single most prevalent, under-recognized, neglected, and severe but easily treatable disease in the world” [3]. In recent years there have been reports of confirmed or probable scrub typhus well outside the classic diagnostic triangle for this disease [4–9]. Based on the results of a literature review, we make available in MiteMap the locations of cases of human scrub typhus after applying stringent diagnostic criteria to categorize probable or confirmed scrub typhus cases from the World War II era to the present. Data from VectorMap will assist investigators in selecting potential sites for vaccine trials, as well as identifying at-risk areas for spraying to control vectors. In addition, the use of VectorMap will assist clinicians who must diagnose and treat cases as endemic areas expand. We aim to: (1) describe and develop a comprehensive updatable case mapping system for gathering human scrub typhus information, (2) propose standards for what constitutes a human case and recommend how it should be reported in the literature, (3) discuss the problem of geolocating transmission sites from case study data, (4) improve our understanding of the global geographic area of transmission (i.e., where is the “endemic” area, and is this a useful concept?), (5) demonstrate the value in mapping cases, and (6) recommend possible directions for future studies of the geography of mite-borne scrub typhus.

This literature review was not intended to be exhaustive, but rather was meant to identify and present a subset or cross section of reports from the majority of geographic areas and years, including confirmed and probable human scrub typhus cases.

This examination of case data compiled from the literature and displayed on VectorMap demonstrates the dramatic geographical and topological range of scrub typhus case investigations throughout the last 70-plus years. This range includes the pre- (effective) antibiotic era of World War II, with thousands of cases in soldiers and hundreds of deaths reported [8], to the present day, when US military interest remains high [19–21]. Our examination underscores the difficulty in identifying the actual distribution and incidence of the human disease, for which there are several reasons. Publications usually do not give precise location information for the likely site of transmission, which resulted in large uncertainty estimates (up to 1,200 km in radius). Thus, vague regional descriptions (e.g., Northern New Guinea), or country and state or province information only are given. Using the likely patient catchment area of a hospital results in lower estimates of spatial uncertainty for civilian cases, but this becomes problematic for military hospitals that receive patients airlifted over a potentially larger area. The value in having an error estimate is that it allows the database user to determine how accurate the georeference is, affecting decisions such as what spatial resolution of remotely sensed data is appropriate for spatial modeling of scrub typhus occurrence data. Each paper also needed to be judged according to the likelihood of the clinician’s diagnosis. Even when modern diagnostic tools were available, we were dependent on the clinician’s suspicions, the differential diagnosis, and the endemicity of the disease [22]. The objective laboratory confirmation becomes paramount. In order to accurately score cases identified in the literature, we found that several different serologic and direct agent detection tests for laboratory confirmation of diagnosis were used (S1 Data, “Identification Method For Parasite”).



0 0 vote
Article Rating
Notify of
Inline Feedbacks
View all comments