Spatio-temporal Transmission Patterns Of Tuberculosis In Songjiang District,Shanghai

Tuberculosis (TB) is a serious chronic disease which is infected by Mycobacterium tuberculosis (Mtb). According to the World Health Organization (WHO), nearly1/3(about20million) of the world population have been infected with Mtb, with9.4million TB patients each year and14million prevalent cases. The5th National Tuberculosis Epidemiological Sampling Survey showed there were5million active tuberculosis patients who were elder than15years old, including1,290,000smear-positive tuberculosis cases. It has been showed that China has high rates of infection, high prevalence of drug resistant cases and high mortality of TB. China ranked2nd regarding disease burden of TB in the22TB worldwide burden countries. With the growing TB/HIV co-infection, multi-drug resistant TB (MDR-TB) and disadvantaged problems from floating population, China faces with severe challenges to prevent and treat tuberculosis.A correct understanding of the transmission of tuberculosis is of great significance to develop effective measures to control the TB epidemic. Due to latent infection characteristics of TB, however, its transmission mechanism remains unclear, especially in areas with high disease burden. New theories and methods of molecular epidemiology and spatial epidemiology are increasingly used to understand the TB transmission combined with traditional epidemiological investigation.Songjiang District was selected as the study site. This study attempted to describe the prevalence and transmission of new TB cases registered between2006and2011in Songjiang District, using spatial epidemiology, molecular epidemiology combined field epidemiology. Specifically, we aimed to describe the prevalence of TB in Songjiang District by descriptive study; to obtain the spatial and temporal characteristics of local resident and rural-to-urban migrant TB cases in Songjiang District by use the space-time scan statistic method; to identify the gathering area, and analysis the risk factors affecting gathering by logistic regression; to determine the recent transmission by the16loci Mycobacterial Interspersed Repetitive Unit-Variable Number of Tandem Repeat (MIRU-VNTR) genotyping of Mtb, using a combination of field epidemiological investigation and molecular epidemiology research methods; to determine the risk factors of recent transmission by comparing demographic and clinical treatment characteristics of clustered and non-clustered cases; to explore the spatial and temporal transmission pattern of tuberculosis by case tracing and analysing spatial aggregation of clustered cases; and provide a theoretical basis for TB prevention and control strategies in Songjiang District.The main research methods and results are as follows:1. Descriptive study of TB in Songjiang DistrictThe research objects are newly registered TB patients in Songjiang District from2006to2011. A total of2916TB cases were registered in Songjiang District during the study period, including9011local resident patients and2015rural-to-urban migrant patients. Of them,1545(52.98%) were bacteriologically confirmed (sputum smear-positive and/or sputum culture-positive), including481in local resident and1064in floating patients.2.Space-time cluster analysis of TB by space-time scan statisticSpace-time cluster analysis by purely spatial scan statistic and space-time scan statistics shows there were spatial and temporal clusters that have been identified, other than a random pattern of the cases.Purely spatial scan statistics identified clusters that were all located in Che Dun both for active TB cases and bacteriologically confirmed cases. For local resident active TB and local bacteriologically confirmed cases, clusters were identified in five communities/towns (Xinbang, Shihudang, Xiao Kunshan, Maogang and Yongfeng) in the southwest areas. The urban migrant patients were clustered in Che Dun and urban migrant sputum positive cases were clustered in Che Dun, Yexie and Mao Gang. Active TB cases and bacteriologically confirmed cases clustered almost in the same communities/towns. The communities/towns where urban migrant patients clustered are the same with that all cases clustered.The cluster of TB patients in Songjiang District were mainly influenced by urban migrant patients.The results of space-time scan statistical analysis showed clusters existed in Che Dun and XinQiao between2007and2008for all cases and migrant patients. This shows the high incidence year of TB in Songjiang District is2007-2008, and the cluster of TB patients in Songjiang District are mainly influenced by the urban migrant patients. Local resident patients were clustered in Xinbang, Shihudang, XiaoKunshan, Mao Gang and Yongfeng between2007and2009. For all bacteriologically confirmed cases and urban migrant bacteriologically confirmed cases, the Spatio-time cluster was found to exist at Che Dun for the year2008-2010. This shows the high incidence year of bacteriologically confirmed TB is2008-2010. Local resident bacteriologically confirmed cases had no clusrer.For urban migrant patients clusters were located in Eastern Industrial Zone, and for local resident patients clusters were identified in five communities/towns in southwestern of Songjiang Disttict. We should focus on these two areas and urban migrants for TB control.3. Influencing factors of TB spatial clusterLogistic regression analysis shows that being the farmers and treatment delays are independently associated with clustering for local residents patients. For urban migrant patients, being blue-collar workers and migration from western region were the risk factors of spatial clustering.4. Mtb genotype and cluster analysisWe used the16loci MIRU-VNTR methods to genotype the Mtb isolates.[11] Between July2009and December2011, there were615strains from culture-positive patients that were genotyped. Most of the isolates (73.82%,454/615) had unique patterns, while the remaining161belonged to57individual clusters. The clusters varied in size from two to fourteen patients. Assuming that genotypic clusters represent recent transmission, approximately seventeen percent of the TB patients in Sonjiang District were due to recent transmission.We performed cluster analysis with urban migrants and local residents. Of the57clusters,22(38.6%) was mixed among urban migrants and local residents. Thus, the transmission of Mtb happened across urban migrants and local residents. The proportion of clustered TB cases among urban migrants was significantly lower than that among local residents (23.76%vs32.37%, p=0.029).5. Risk factors of recent transmissionDemographic and clinical characteristics of patients with clustered versus unique Mtb DNA fingerprints were analysed by univariate and multivariate logistic regression. For the bacteriologically confirmed cases, those who living in towns are more likely to be clustered (OR=1.558,95%CI:1.022-2.375); sputum smear-positive patients were more likely to be clustered (OR=1.666; CI1.087-2.553). In addition, TB patients infected with a Beijing strain of Mtb were more likely to be clustered than those with non-Beijing strains (OR=3.019,95%CI:1.76-5.179). The odds of being in a cluster were2.93times higher among TB cases with MDR TB than among TB cases with pan-susceptible TB.6. Mechanism of TB recent transmissionThe contact tracing of clustered patients found, of the161clustered patients from57clusters, only15TB cases from7clusters had epidemiological links with other TB cases in the same cluster. Purely spatial scan analysis of the clustered patients shows that the clustered patients clustered in Shihudang and Xinbang area (RR=2.32, p=0.035). TB cases in the same cluster distributed dispersedly, indicating that the spread of TB may be due to some patients causal contact or contact with other unidentified TB patients.