Scrub typhus in urban areas of Wardha district in central India

Scrub typhus usually presents as acute undifferentiated fever and is caused by Orientia tsutsugamushi, a bacterium of the order Rickettsiales. Human transmission of the bacterium is caused by the bite of infected larval stage (called chiggers) of trombiculid mites1. The disease spectrum ranges from fever to lethal complications involving various organs. Early clinical manifestations of the illness are fever, headache, myalgia, and rash, thus making it difficult to differentiate scrub typhus from other acute febrile illnesses. The pathognomonic feature, i.e. eschar, develops when the small lesion at the bite site becomes large, undergoes necrosis at the centre and develops blackened crust. Eschars are usually found in the parts of the body that are warm and damp, e.g. groin, axilla and waistband. However, the proportion of scrub typhus patients showing eschar ranges from 1-4 to >60 per cent1. One new species, Orientia chuto was identified in Dubai in 20102. Emerging issue of scrub typhus is the increasing burden on public health within the endemic region and its expansion outside it3. The disease has also emerged in newer areas in the endemic zone, for example, in northern China4, and continuous outbreaks in several countries including India5. There have been efforts to update epidemiological data on the precise disease burden in the last decade, and the information is available on various aspects of the disease3,6,7. Climatic factors have been shown to affect the larval response to temperature and humidity, thus affecting disease occurrence to changes in temperature and rainfall. There has been evidence that climatic factors that affect the temporal occurrence of the number of cases in some areas1. Conventionally, scrub typhus is considered to be a disease associated with agricultural activities in rural areas8; however, slowly, the disease is progressing to urban areas9.

The present hospital-based study was carried out to know the magnitude and distribution of scrub typhus in the Wardha district in central India, and also the climatic factors affecting the disease.

Material & Methods

This hospital-based surveillance of zoonotic and vector-borne zoonotic diseases was conducted in the tertiary care Kasturba Hospital, Sewagram, located at Wardha district in central India, from July 2015 to June 2016. Patients presenting with fever of five days or more were tested for malaria, dengue, and enteric fever and those found negative were recruited for the study after obtaining informed written consent. Information on sociodemographic variables were collected by personal interviews. A total of 3454 individuals were screened and 1680 were included in the study. Five millilitres of blood was drawn from each patient into a vacutainer and transferred to the microbiology laboratory immediately. Serum was separated and IgM ELISA for scrub typhus was performed on the serum samples. For the present study, all the IgM ELISA-positive samples and 10 per cent of negative samples were tested by IgM indirect immunofluorescence assay (IFA). Positivity for both IgM ELISA and IgM IFA was considered as a diagnostic confirmation for scrub typhus. IgM ELISA-negative samples also tested by IFA were found to be negative by IFA. This study was carried out after the permission obtained from the Institutional Ethics Committee wide letter number MGIMS/IEC/ICMR/118/2014.

IgM ELISA: Commercial scrub typhus Detect™ IgM ELISA Kit (InBios International Inc., Seattle, WA 98109, USA) was used, which included a recombinant p56-kDa type-specific antigen for the detection of IgM antibodies against O. tsutsugamushi strains, Karp, Kato, Gilliam and TA716. The procedure was performed as per the manufacturer’s recommendation. Absorbance was read at 450 nm using Multiskan FC reader (Thermo Scientific, Waltham, Massachusetts, USA). Cut-off value of 0.5 was decided5,10.

IgM IFA: O. tsutsugamushi IFA IgM antibody kit OTM-100 (1312 E Valencia 92831, Fullers Laboratories, USA) was used to test IgM IFA. Four different prototype antigens of O. tsutsugamushi, namely Karp, Kato, Gilliam and Boryong, were coated on the IFA slide. The assay was performed as per the manufacturer’s instructions. The slides were read using 490 nm excitation wavelength for fluorescein isothiocyanate (FITC) under fluorescent microscope Evos FL (Thermo Scientific, Waltham, Massachusetts, USA), and the results were recorded as positive or negative. End-point titres up to 1:512 were calculated. The criteria for positivity as recommended by the kit literature and Indian studies on the determination of IFA titre were kept at 1:64 dilution10,11. We tried to differentiate among serotypes using the highest IFA IgM titres, but most serum samples showed the same highest titre for at least two and sometimes all serotypes. Thus, serotype differentiation was not possible in this study.

Climatic variables: India Meteorological Department (IMD) and National Centre for Environmental Prediction (NCEP) provided data on weather variables (rainfall, relative humidity and average temperature). The IMD provided the daily rainfall data and temperature at a surface grid resolution of 0.25 × 0.25° and 1 × 1°, respectively12,13. The gridded data for relative humidity were obtained from the NCEP (https://psl.noaa.gov/data/gridded/data.ncep.html). The data were available in the grid format and further processed using the tools Climate Data Operator and Grid Analysis and Display System14 (www.ncep.noaa.gov) . The area-weighted average was applied to extract the district-wise data for Maharashtra State on a monthly basis and further complied to get the annual accumulated rainfall15.

Statistical analysis: The magnitude of scrub typhus was expressed in percentages with 95 per cent confidence interval given. Pearson’s correlation coefficient was used to identify the time lag of a climatic variable, and a lag time with the highest correlation coefficient has been presented here with P<0.05. Association between climatic condition and magnitude of scrub typhus was studied using correlation coefficient. The climatic determinants were analysed using time-series Poisson regression. Spatial analysis (point pattern analysis) was done using CrimeStat v 3.316 (software for spatial analysis, especially used for analysis of crime incident location data). The hotspots of scrub typhus were identified using k-mean clusters analysis using an online software SaTScan™ v10.1 (https://www.satscan.org/). K-mean clusters quantify geographic variation patterns in the incidence of scrub typhus. QGIS 2.18.2 (https://www.qgis.osgeo.org) was used to plot the maps.

Results

During the study period, 1680 individuals were tested for scrub typhus who had fever of five days or more duration. Majority of them (n=636, 37.86%) were in the age group of 21-40 yr and 22.50 per cent (n=378) were 41-60 yr old. Males were 53.75 and 28.39 per cent individuals were working in farms. Half of the individuals presented in monsoon while 20 per cent each presented in post-monsoon and winter. Three fourth of them belonged to the Wardha district (Table I).

During the year, magnitude of scrub typhus was 15.89 per cent. The highest proportion of scrub typhus was observed to be 23.01 per cent in 41-60 yr, females had slightly higher positivity than male (16.60 vs. 15.28%), farm workers (23.69%) had higher positivity as compared to non-farm workers (12.88%) and positivity was very high in monsoon and post-monsoon seasons (22.84 and 19.40%, respectively) (Table I).

Significant correlation of magnitude of scrub typhus with rainfall was found in the previous month (r=0.802, P=0.002) and with relative humidity in the previous month (r=0.868, P<0.001). Correlation (r=0.321, P=0.309) with average temperature was not found to be significant (Table II). The month-wise number of cases with average climate conditions, viz. temperature, rainfall and humidity, starting from three months before the study is given in Table III.

We did spatial analysis of 196 cases, who were from Wardha district. Figure and Table IV describe spatial distribution of the scrub typhus in Wardha district. Nearest neighbour distance was 1028.2 m [standard deviation (SD) = 3028.9 m] and nearest neighbour index was 0.3578 which showed significant clustering of scrub typhus cases (P=0.001). Blue circle shows one SD of scrub typhus cases, indicating 68 per cent of scrub typhus cases resided in this area with mean centre near Wardha town. Three hotspots were discovered using k-means clustering as indicated by black circles. These were areas surrounding Wardha, Hinganghat and Deoli towns of Wardha district.

Close

Figure

Spatial distribution and clustering of scrub typhus.

Export to PPT

Discussion

Initial studies on acute undifferentiated fever in central India were conducted by Joshi et al17,18. The present study was initiated to look for zoonotic agents of acute undifferentiated fever in addition to routine agents. This year-long study revealed that up to 16 per cent of cases of acute febrile illness had IgM antibodies against O. tsutsugamushi detected by capture ELISA and confirmed by IgM IFA. We also looked for epidemiology and spatiotemporal distribution of the cases in the Wardha district, as was done in the recently published study from mainland China19.

In our study, high positivity was observed in monsoon and post-monsoon seasons while it was minimal in the winter and summer seasons. Peak period was of August and September. The rainfall and the relative humidity in the previous month were significantly associated with magnitude of scrub typhus in our area, indicating enhanced activity of mites20.

An Indian case–control study looking for scrub typhus in acute encephalitis cases has reported various risk factors in children as well21. Our study has demonstrated a marginal increased risk of cases in females (female 16.60% and male 15.28%), which may be linked to their field duration exposures in this semiarid area and open defecation habits. Various other studies, especially from rice-growing areas, have reported females to be exposed more to vector mites and having more scrub typhus1.

Three hotspots were found using k-means clustering, and these areas were surrounding Wardha, Hinganghat, and Deoli towns of Wardha district which are urban areas. However, this tertiary care hospital-based study may not be considered as a true representative of the Wardha district. The analysis has been helpful to identify that there was clustering of cases in urban areas around the hospital, and this could be because of access to the only diagnostic facility for scrub typhus in the district. Peripheral areas of the district may have similar burden of the disease, but because of lack of diagnostic facility the situation was not clear.

Worldwide also, the disease is becoming urbanized and may not be considered as a problem restricted to rural areas22,23. Reasons for extension of scrub typhus to urban areas have been hypothesized by Ranjan and Prakash9 from south India.

The contact with chigger mite is especially increased when the conditions are conducive. In our study, more cases were detected when the average temperature was 28°C and humidity was more than 45 per cent reaching up to 74 per cent in August, with maximum number of cases detected in September. This association accounts for overgrowth of vegetation, increased activity of rodents and thriving of chigger mite.

Our study was complemented by the study of rodents and vector mites by Bhate et al24. A total of 50 rodents were tested for mites by presence of O. tsutsugamushi using PCR for 56 kDa type-specific antigen gene. The rodents caught from Nagpur railway station harboured mites and a new vector Ornithonyssus bacoti was also demonstrated to harbour the bacterium24.

It was hypothesized that at least three factors were responsible for the emergence of scrub typhus in urban areas of district Wardha: (i) clearing of trees and construction activities, thus creating more suitable habitat for mite; (ii) infected chiggers belonging to other genera may be expanding to urban areas as was reported from Nagpur24; and (iii) changes in the behaviour of urban residents with increased outdoor recreational activities, leading to more contact with mites9.

There were certain limitations of this study. We could not identify the serotypes of O. tsutsugamushi as the IgM IFA test was not able to differentiate among them. Second, the hospital-based study may not be representative of whole of Wardha district/central India.

To conclude, scrub typhus was found mainly in monsoon and post-monsoon periods in Wardha district of Maharashtra with an overall positivity of 15.89 per cent and peak during monsoon. Economically productive age group, farmworkers and females were affected more. Significant clustering occurred in three urban areas in the centre of the district. Among the climatic determinants, rainfall and relative humidity during the previous month were significantly associated with the occurrence of scrub typhus.