Study Of Epidemic Features And Related Factors Of Autumn-winter Type Scrub Typhus

BACKGROUNDScrub typhus, or tsutsugamushi disease, is known as a febrile zoonosis caused by Orientia tsutsugamushi, with clinical manifestations characterized by typical eschars or ulcers, high fever, lymnopathy, and rash. It can also cause many complications or even death.Scrub typhus has been getting great prevalence in the Asia-Pacific triangle areas which involves Japan in the north, Australia in the south and Pakistan and Afghanistan in the west. China is one of the most important epidemic areas for scrub typhus. Before1985, scrub typhus was confined geographically south of Yangtze River of China (south to north latitude31°) with summer as the main epidemic season. Since1985, it expanded across Yangtze River to Northern China and new endemic areas were continuously confirmed. The prevalence of scrub typhus usually occurs during the period of autumn and winter time, so it is called autumn-winter type scrub typhus compared to summer type in South. In Northern China, Shandong Province is regarded as the typical epidemic focus for autumn-winter type scrub typhus. This kind of disease got its first outbreak in Mengyin county in1986, and then has spread rapidly to80%areas of Shandong province.Scrub typhus, which can do great harm to the health of people, has been spreading rapidly as a kind of new-born infectious disease in Northern China, with its intensity growing. Based on data of the typical focus of Shandong province, we described the epidemic features and explored related factors of autumn-winter type scrub typhus in our study to provide important information on its prevention and control. OBJECTIVE1. To describe the epidemic features of scrub typhus by methods of traditional analysis, spatial analysis, and time series analysis.2. To explore related factors of scrub typhus by a case-control study.Both of the two parts were aimed to provide scientific information for its prevention and control.MATERIALS and METHODS1. Source of materials(1) Descriptive study:Based on Shandong Diseases Reporting Information System of year2006to2011, information of1746cases were collected for analysis.(2) Case-control study:Based on Shandong Diseases Reporting Information System (SDIS) of year2010, a total of128cases and256controls were enrolled from9counties during the period of January to March of2011.2. Methods(1) Descriptive study:Distributions by people, time and place were described with statistical tests of x2, Kruskal-WallisH rank sum. The significance level was set to0.05. Based on data base of geographic information system, and softwares of ArcGIS9.3and SatScan7.0, methods of disease mapping, inverse distance weighted, spatial autocorrelation analysis, purely spatial analysis, purely temporal analysis and spatial variation in temporal trends were applied to explore the temporal and spatial distribution of scrub typhus. Spectral analysis, moving average analysis, seasonal decomposition analysis and establishion of ARIMA were used here to analyze the data of scrub typhus from2006to2011.(2) Case-control study:All the subjects were investigated with questionnaire. Logistic regression analysis was adopted to explore related factors of scrub typhus with software of SPSS16.0after all the data had been adjusted. The relationship of cause-result was expressed by indicators of odds ratio (OR),95%confidential interval (CI) and Population Attribute Risk Percent (PAR%). RESULTS1. Epedemic features(1) Three-dimension distribution of scrub typhus:A total of1476cases were analyzed. The onset age of scrub typhus ranged from1to92years old and51.1%(892/1746) of the cases were beyond55years;48.3%(843/1746) of cases were men and51.7%(903/1746) were women, and the ratio was1:1.07; farmers, preschoolers, students and workers were involved in scrub typhus cases, and farmers were the predominant population accounting for84.6%(1477/1746). This disease occurred mainly in period of October to November with October as the peak time, and71.0%(1240/1746) of all cases happened in October. Of all the17cities in Shandong province, Linyi, Taian, Rizhao, Laiwu, and Qingdao were considered to be the most susceptible places for scrub typhus. And22.9%(399/1746),18.1%(316/1746),14.7%(256/1746),14.1%(246/1746) and13.1%(229/1746) of all the cases occurred in the former5cities respectively.(2) Spatial analysis:Global Moran’sI index was0.324(P<0.01). There were8locations with statistical significance (P<0.01) of Local Moran’sI, of which Gangcheng, Laicheng, Xintai, and Donggang were the most important locations with case numbers of38.154,160,105and Local Moran’ sI of2.111,1.642,1.277,0.775respectively. They were all high-high (H-H) clusters. Temporal clusters of scrub typhus were from2006.09.23to2006.11.20(202cases),2007.10.02to2007.11.11(197cases),2008.09.30to2008.11.07(302cases).2009.09.25to2009.11.10(204cases), and2010.10.05to2010.11.13(226cases) with RRs of45.55、34.60、50.64、53.09and79.84respectively (P<0.01). Two spatial clusters were found in our study, and they were inland areas of mountains and hills and costal areas of hills, with different centers of Xintai of Taian and Donggang of Rizhao, and different radius of58.28and22.68kilometers respectively; the RRs (and case numbers) were4.52(542) and3.96(134) respectively (P<0.01). The area with Mengyin county as the center and a radius of45.82kilometers was considered to be the most susceptible locations with an increasing incidence rate of45.04%per year, and the RR was3.68(P<0.01). (3) Time series analysis:The results of spectral analysis found that the prevalence of autumn-winter type scrub typhus had a periodicity of3years. A long-term up-trend was confirmed by method of moving average analysis, with annually case numbers of310,337and366forecasted for2012to2014respectively, and the annually increasing rate was9%per-year. The multiple seasonal decomposition analysis indicated that the prevalence of this disease possessed a typical autumn-winter type. The seasonality indexes for scrub typhus in October and November were8.454and2.230respectively, while others were no more than1. The ARIM4(0,1,1)(0,1,0)12model of (1-B)(1-B12)Xt,=(1-0.811B)u, which was used to forecast the prevalence of autumn-winter type scrub typhus was constructed with the residual error of16lags as white noise. The Box-Ljung test statistic for the model was3.116, giving a P value of0.999. The model fitted the data well. Good accordance was achieved between the observed values and the forecasted values of scrub typhus in November and December of2011produced by the ARIMA model, and all observed values were within the forecasted95%CI2. Related factors of scrub typhusResults of logistic regression analysis indicated that Multiple analysis by conditional logistic regression model demonstrated that surroundings with weeds or crops growing, living in outskirts of a village, a bitten history of unknown insects, firewood or straws piled in yards, living in damp conditions, ill personal health habits, and living in cottages were risk factors for scrub typhus, and their ORs (95%CI) were2.815([1.493.5.307]),2.250([1.208,4.190]),4.500([2.206,9.180]),2.100([1.163,3.793]),5.788([1.819,18.418]),1.943([1.000,3.810]) and4.409([1.062,15.434]) with PAR%of44.04%、41.45%、29.77%、27.44%、19.56%、13.95%、7.68%respectively. Recognizing that rodents transmit diseases was a preventive factor with OR (95%CI) and PAR%were0.523([0.292,0.936])and47.57%.CONCLUSIONS1. Epedemic featuress:Farmers of middle and old age are the predominant population for scrub typhus, and the prevalence time is between the last10days of September to the first10days of November. A positive spatial correlation of the disease is found and they are all high-high clusters, of which Gangcheng, Laicheng, Xintai dan Donggang are the most important. Two spatial clusters were found, which were inland areas of mountains and hills and costal areas of hills, and area around Mengyin was the most susceptible location with highest yearly increasing incidence rate. The prevalence of autumn-winter type scrub typhus has a periodicity of3years probably, with a long-term up-trend, and the case numbers of2012to2014are forecasted to be310,337and366respectively with an increasing rate of9%per-year. It occurs seasonally with October as the peak time for every year. A ARIMA(0,1,1)(0,1,0)12model is quite available for the forecast of autumn-winter type scrub typhus in short coming months.2. Related factors:Living environments, personal health habits and recognition on health knowledge are related to the prevalence of scrub typhus. surroundings with weeds or crops growing, living in outskirts of a village, a bitten history of unknown insects, firewood or straws piled in yards, living in damp conditions, ill personal health habits, and living in cottages were included in risk factors for scrub typhus. Recognizing that rodents transmit diseases was considered as a preventive factor.