| Background: Scrub typhus(ST) is a natural focus infection disease caused by Orientia tsutsugamushi. The transmission of the etiological agent to the rodent host or the human incidental host occurs following the bite of infected mite vectors. The clinical manifestation of the disease usually includes acute fever, eschar or ulcer, lymphadenopathy, rash, enlargement of liver and spleen, and its associated complications such as pneumonia, meningitis, disseminated intravascular coagulation(DIC), etc., may be caused even to death. Scrub typhus is endemic to a 13,000,000 km2 area of the Asia-Pacific rim, extending from Afghanistan to China, Korea, the islands of the southwestern Pacific, and northern Australia. And it posed particular threat for China and Southeast Asia.From as early 313 A.D. and throughout China’s history, there have been numerous references to “chigger fever” on the clinical features, as well as prevention and treatment methods. It was in 1948 that Chinese scientist isolated the etiological agent at the first time in Guangzhou, Guangdong Province. ST case reports starts from 1952 in China. It was endemic only in provinces south of Yangtze River before 1985, including Hainan, Fujian, Guangdong, Guangxi, Yunnan, Sichuan, Zhejiang, Hunan, south of Tibet and Taiwan, with average annual incidence of 1000. Since the beginning of 1980 s, it began to attack the areas north of Yangtze River, with a start of Mengyin County, Shandong Province, and Dongtai City, Jiangsu Province in 1986. Then cases emerged in Anhui, Henan, Hebei, Tianjin, Beijing, Shanxi, Jilin, Liaoning and Heilongjiang. It was listed in routine reporting system as notified disease until 1989 in China; nevertheless the increasing emdemicity trend was noticeable recently. As of December 2013, scrub typhus cases had been reported in 29 provinces of China. Annual incidence was on increase, and it was over 10 thousand confirmed cases reported in the year of 2013.Along with the extension of natural focus and the increase of annual incidence, endemic heterogeneity was shown between the foci north and south of the Yangtze River. The majority ST cases in the foci south of Yangtze River occurred in summer and autumn, and peaked between June—August. On the other hand, most of ST cases reported in autumn and winter in the foci north of Yangtze River, and peaked between November—December. Dominant vectors and animal hosts of ST are heterogeneous for different regions, e.g., Leptotrombidium deliense is the most important vector responsible for ST transmission in the south, while Leptotrombidium scutellare in the north. And Rattus losea, Rattus flavipectus, Rattus norvegicus, Niviventer confucianus and Apodemus agrarius are the main animal hosts; while Apodemus agrarius, Rattus norvegicus and Tscherskia triton in the north.As an ancient natural focus infection disease, spatiotemporal expansion during a long period and in recent years, endemic characteristics for different natural foci, and risk factors for endemicity are still unknown. It is important to carry out researches on nationwide endemic features, understand the heterogeneity between foci north and south of Yangtze River, clarify the spatiotemporal distribution of ST in China, determine the high-risk population, the high-risk season and high-risk areas, assess the environmental risk factors and forecast the incident risk.Objective: ⑴ To clarify pattern of spatiotemporal expansion during a long period and in recent years. ⑵ To summarize the heterogeneity of the endemic characteristics and related potential factors. ⑶ To further mapping the risk distribution of the disease by apllying BRT models separately for the two different natural foci, north and south foci of Yangtze River. ⑷ To identify the endemic features of ST in typical natural foci south and north of Yangtze River respectively, and to quantitatively evaluate the environment risk factors.Methods: ⑴ Historical data of reported cases(1980–1989) and the first reported time and location(1950–1980) were collected from China Information System for Disease Control and Prevention(CISDCP), and references involving reports of ST emergence were also collected to describe the historical procedure of the spatiotemporal expansion. ⑵ Reported ST cases in China from January 1, 2006 to December 31, 2013 were obtained from the CISDCP, and multivariate mathematical statistics and geographic information system technology were combined to describe and analysis the epidemiology features of ST nationwide. A preliminary experiment was conducted based on the reported data of 1980–1989 and 2006–2013, and the South and North Natural focus were classificated by Yangtze River. ⑶ Based on reported ST cases and data concerning climate and land cover collected, a comprehensive spatial database of ST was constructed. Two Boosted Regression Tree(BRT) models were applied at the county level respectively for South and North Natural focus to assess the risk factors associated with the occurrence of ST, and then a risk map of ST was created in China. ⑷ Reported ST cases in Guangzhou from January 1, 2006 to December 31, 2014 were obtained from the CISDCP. The endemicity of ST in the typical natural foci south of Yangtze River was described to details. Data regarding the climate, land cover and demographics were collected as well. And a Negative binomial regression model was applied to explore the associations between ST incidence and environmental factors at the township level. Based on an epidemiological investigation performed on a random sample of local hospitals during January 1, 2012 to February 28, 2013 in Guangzhou, clinical features of typical Summer-type ST patients were summerized. ⑸ Based on all reported cases in Shandong, Jiangsu and Anhui during 2006–2013, endemic features of ST was characterized, and environmental factors associated with the endemicity was calculated using Negative binomial regression at county level. ⑹ The software used in this study including Microsoft office 2010, Arc GIS 9.3, Epidata 3.1, Origin Pro 8, STATA 9.1, R language 3.1.1, etc.Results: ⑴ With a start in Guangzhou City, Guangdong, on 1948, natural foci of scrub typhus had been identified subsequently in most provinces south of Yangtze River in China: Fujian Province in 1951, Guangxi Province in 1952, Zhejiang Province in 1954, Yunnan Province in 1956, Sichuan Province in 1966, Tibet Autonomous Region in 1973, and Hainan Island in 1981, long before the Hainan Province established. In 1986, scrub typhus strode north over Yangtze River at the first time, and was documented to occur in Shandong Province and Jiangsu Province. With the investigations of the disease in northeast China, it had been found over latitude 44°N by 1990 s and the natural foci were reported in majority of provinces of mainland China. The northernmost reach of the disease was Aihui County of Heilongjiang Province dated to 2013.⑵ The noticeable increase of the annual incidence had been shown. The ST incidence of foci south of Yangtze River rose faster than that in the north. There were a quicker increase of the incidence during 2006–2013 than 1980–1989, and the majority of cases(89.7% during 1980–1989 and 75.7% during 2006–2013) were distributed in foci south of Yangtze River. The seasonality heterogeneity of the disease between the two foci is also evident. Foci in south of Yangtze River represented a seasonal endemic peak in summer-autumn(July–September) and those in north of Yangtze River usually showed a later endemic peak with 1–3 months lag. The endemic peak of the disease shifted from summer to autumn with the latitude to north. Yunnan, Guangdong, Fujian, and Hainan had higher incidence in south of Yangtze River, and Anhui, Jiangsu, as well as Shandong were the most important epidemic areas in north of Yangtze River. All above provinces had an increasing pattern during the recent years.⑶ 38,603 confirmed ST cases were reported in China from 2006 to 2013, distributing in 917 counties of 28 provinces. The staple constituent of patients was farmer in both foci, increased from 58.5% in 2006 to 70.7% in 2013 in total. The group of scattered and preschool children was also to be reckoned, and the percentage was 16.9% in south provinces. There were more female cases than male cases in either focus. Median age of the patients was 50(IQR: 33–61). The majority of cases aged 40 and above(68.54%), indicating that the elderly were at higher risk. The highest annual average age-specific incidence occurred in the age group above 60, which also rose fastest among all age groups in the both foci during the study period. The variation in patients’ population structure and demography was illustrated by either focus. The situation differed between the two foci: little boys got more chance to get scrub typhus than girls at early age(age group 0–9) but situation reversed at mid-age(age group 20–59, chi-sqaure test, P < 0.05) in the foci north of Yangtze River; while in the foci south of Yangtze River, though boys were at higher risk than girls(age group 0–19, chi-square test, P < 0.05), the aged women had more probability to be hospitalized than men(age group above 40, chi-square test, P < 0.05). It should be extra pointed out that the incidence of the children under 10 in foci south of Yangtze River was 6.98 per 1,000,000 population per year, much higher than the same figure that 0.45 in north, also the increment speed of this age group was fast in the south, which required more attention to children in southern provinces.⑷ Using the BRT models, importance of risk factors assessed by the estimated weight and the fitted function of each risk factor both showed heterogeneity between foci north and south of the Yangtze River. In the foci north of Yangtze River, the influential factors were precipitation, sunshine hour, temperature, percentage coverage of crop field, and relative humidity; though in the foci south of Yangtze River, those of significance were temperature, sunshine hour and relative humidity, all with average BRT weights > 5.0%. To the preference of the disease occurrence, precipitation 700 mm, sunshine hour 140–180h, temperature 9°C–14°C, percentage coverage of crop field 47–80%, and relative humidity 62–65% were favorable in north provinces; while temperature above 15°C, and sunshine hour around 150 h, as well as relative humidity below 65% in south provinces. The ranking of the relative contribution of all land types was as follows: in foci north of Yangtze River, crop field, grass land, broadleaf forest, coniferous forest, shrub, and mixed broadleaf-conifer forest; in foci south of Yangtze River, grass land, shrub, crop field, broadleaf forest, coniferous forest, and mixed broadleaf-conifer forest. The AUC value was 0.838(95% CI: 0.812–0.864) for the BRT model of north and 0.841(95% CI: 0.819–0.863) for that of south. With the values above 0.8 the results suggested a good predictive accuracy of the predicted models. The high risk areas overlapped very well with the observed scrub typhus cases. In the south foci, areas at high risk of scrub typhus presence included Yunnan, Hainan, Guangdong, Fujian, Zhejiang, Guangxi and south of Sichuan, all with long epidemic histories of scrub typhus. Hunan, Jiangxi, north of Chongqing, and west of Sichuan, was at quite risk as well. There was also large probability of scrub typhus presence when we cast an eye northward, especially in Jiangsu, Anhui and Shandong. Emerging foci should be noted such as Hebei, Liaoning, Jilin, Heilongjiang, south of Gansu, and west of Xinjiang.⑸ A total of 4,821 confirmed cases were reported during 2006–2014 in Guangzhou. All 12 counties of Guangzhou and 98.8% of the towns(159/161) were affected. 11 out of 12 counties showed increasing incidences during the nine years except for Liwan County. The majority of cases(86.8%) occurred during May–October, and the incidence peaked in either June/July or October with dual peaks in some years. The annual incidence increased slowly but steadily from 2006 to 2011, then jumped in 2012, and remained high through 2013 and 2014. Rural areas had 2–4 times higher incidences and a more rapid increase of the incidence than urban areas during the study period. The newly-established epidemic towns mostly occurred during 2007–2008 in both rural and urban areas. Compared with urban areas, ever-affected towns in rural areas were more likely to see reemergence of the disease each year. Farmers constitute the majority of the cases over the study period, accounting for 33.9% in urban and 61.6% in rural areas, followed by housekeepers who accounted for 19.6% in urban and 12.5% in rural areas. The proportion of housekeepers tended to increase, over the study years and more so in urban areas than in rural areas. The senior population older than 50 years had the highest average annual incidence among all age groups, especially the senior females. The age difference in average annual incidence seems more eminent in rural areas than in urban areas. In both rural and urban areas, average annual incidences in males were higher than that in females among the population younger than 50, but the direction of gender difference flipped among the older population, i.e., elderly females were more prone to infection with scrub typhus than elderly males. Average atmospheric pressure with 1-month lag, average relative humidity with 2-month lags, percentage coverage of broadleaved forest and type of township, are shown to be independent predictors for the spatiotemporal distribution of the disease in the multivariate negative binomial regression model. Each one hundred Pa increase of average atmospheric pressure was associated with an 11%(95% CI: 10%–12%) decrease in the incidence of scrub typhus in the next month, while a 10% rise in monthly average relative humidity corresponded to an 4%(95% CI: 3%–5%) increase in the scrub typhus incidence in the month after the next. for every 10% increase in the percentage coverage of broadleaved forest, the incidence of the disease went up by 5%(95% CI: 3%–6%). On average, a rural town had an 81%(95% CI: 43%–129%) higher risk than a urban town. A total of 571 confirmed patients were interviewed in the epidemiology study. Of these patients, 108(18.9%) were classified as severe scrub typhus. Common symptoms regardless of severity included fever(Fisher exact test, P-value = 0.218), eschar or ulcer(chi-square test, P-value = 0.586), headache(chi-square test, P-value = 0.321), and delirium(chi-square test, P-value = 0.361). Irregular fever was noted as the most common fever type among both severity groups. Eschar or ulcer was usually distributed in the urogenital/peri-urogenital area, the axilla and the chest. However, severe patients suffered from significantly longer fever duration(Wilcoxon rank-sum, P-value < 0.001), higher maximum body temperature(Wilcoxon rank-sum, P-value = 0.003), and more frequent presence of cough(chi-square test, P-value < 0.001), nausea(chi-square test, P-value = 0.014), abdominal discomfort(chi-square test, P-value < 0.001), splenomegaly(chi-square test, P-value = 0.021), coma(chi-square test, P-value < 0.001), and lymphadenectasis(chi-square test, P-value < 0.001). Severe patients were less likely to be correctly diagnosed initially, compared with mild patients(36.1% vs. 70.4%, chi-square test, P-value < 0.001). In addition, severe patients tended to have higher WBC count, lymphocyte count, aspartate aminotransferase, alanine aminotransferase and direct bilirubin, but lower serumcreatinine(Wilcoxon rank-sum, P-value < 0.05). The combination of multiple antimicrobials was also not rare(14.5%), and the majority of them were doxycycline coupled with other antimicrobials.⑹ A total of 2,968, 2,331 and 3,447 scrub typhus cases were reported during 2006–2013 in Shandong, Jiangsu and Anhui respectively. The average annual incidence was 0.39, 0.38 and 0.94 per 100,000 populations. The uptrend was observed in Shandong and Jiangsu, beside a slight rollback in Anhui. Cases were clustered in autumn with a single epidemic peak at October and November in the 3 provinces. Natural foci expansion was found in the 3 provinces, with affected counties respectively accounted for 38.0%, 48.2% and 46.5% in Shandong, Jiangsu and Anhui in 2013. Female patients accounted for 53.1%, 52.1%, and 56.7% in Shandong, Jiangsu and Anhui. The population older than 60 years had the highest average annual incidence among all age groups(9.57/100,000 in Shandong, 1.24/100,000 in Jiangsu and 2.21/100,000 in Anhui). The incidence of females was significantly higher in age group 30–59 in Shandong, 50–59 in Jiangsu and age group above 20 in Anhui. The majority of the cases came from farmers during the study period. And farmers also showed an increasing incidence in these provinces, followed by housekeepers in Shandong and scattered and preschool children in Anhui. The transmission risk of scrub typhus in all the 3 provinces was shown to be negative associated with monthly precipitation, and had an “inverted-U” pattern in association with monthly temperature. Furthermore, a positive relationship between the transmission risks of the disease with monthly relative humidity was found in Shandong and Anhui, whereas an “inverted-U” pattern between that and monthly sunshine hour was examined in Shandong and Jiangsu provinces. The incidence of ST in Shandong was also positive related to the percentage coverage of forest as well.Conclusions: Epidemiological characters of scrub typhus in China were retailed. The incidental population, high-risk areas and the high-occurrence seasons were located in both Northern and Southern China, with the heterogeneity elucidated. Environmental influencial factors were quantatively assessed, and a risk map for scrub typhus was plotted in mainland China.⑴ The natural foci expansion of scrub typhus in China: Based on the literature and historical data of reported cases, this depicted a detailed picture of the foci expansion of ST in China.⑵ The epidemiological heterogeneity of scrub typhus in Northern and Southern China: Based on the Geographic Information System(GIS) technology, contrastive analysis of the epidemiology and risk factor heterogeneity of ST between Northern and Southern China was summarized.(1) Spatio-temporal distribution: High-occurrence seasons were summer and autumn in the foci south of Yangtze River, while autumn and winter in the foci north of Yangtze River.(2)Population distribution: Though high-risk populations structured differently between the two foci, those were the elderly, females and farmers in general.(3)Environmental influential factors: At the macro-level, risk factors in the north were precipitation, sunshine hour, temperature, percentage coverage of crop land and relative humidity; while thus were temperature, sunshine hour, relative humidity in the south; at the micro-level, risk factors associated with monthly incidence in Guangzhou, a typical “summer-autumn type” ST natural focus, were atmospheric pressure, relative humidity, percentage coverage of broadleaved forest and feature of town; while associated factors in Shanong, Jiangsu and Anhui, typical “autumn-winter type” ST natural foci, were precipitation and temperature, also relative humidity was related in Shandong and Anhui, while sunshine hour in Shandong and Jiangsu. The incidence of ST in Shandong was also positive related to the percentage coverage of forest as well.⑶ Risk forecast was assessed through conducting the Boosted Regression Tree models separately for foci in the north and the south, with higher risk shown in the south. High-risk area included Yunan, Hainan, Guangdong, Fujian, Zhejiang, Guangxi and south of Sichuan in the south; and Shandong, Anhui, Jiangsu and mid Shaanxi in the north. Prevention and control in Hunan, Jiangxi, north Chongqing, west Sichuan, Hebei, Liaoning, Jilin, Heilongjiang, south Gansu and west Xinjiang should be strenthern as well, though fewer cases were reported there recently. |
PDF Full Text Request