Study On The Spatio-temporal Clustering Rules And Influence Factor Of Pulmonary Tuberculosis In Dongguan City

Objective:To understand the epidemiological and spatiotemporal clustering rules of pulmonary tuberculosis in Dongguan City from 2011 to 2020,so as to provide data support for scientific decision-making of pulmonary tuberculosis prevention and control.To evaluate the care-seeking delay of pulmonary tuberculosis patients and its influencing factors,so as to provide a scientific basis for improving the care-seeking delay in pulmonary tuberculosis.To master the drug resistance status of pulmonary tuberculosis patients and its influencing factors status,so as to provide important reference for guiding clinical rational use of drugs.To explore TB/HIV dual infection among pulmonary tuberculosis patients and its risk factors,in order to provide scientific evidence for generate preventive intervention strategies and reducing the burden of disease.Methods:This study adopted the monitoring research method for 10 consecutive years,and the pulmonary tuberculosis patients in Dongguan City from2011 to 2020 were included as study participants,with a total of 37248patients with pulmonary tuberculosis.The combination of spatial auto-correlation and spatio-temporal multi-component model was used to explore the spatio-temporal clustering of tuberculosis in different regions.The analysis idea of"sequence stabilization→model identification→model diagnosis"was established,and the ARIMA model was used to predict the number of cases of pulmonary tuberculosis in 2020.The combination of the rank-sum test and multiple linear regression were used to explore the influencing factors of care-seeking days.The combination of theχ²test and multivariate Logistic regression were used to analyze the influencing factors of care-seeking delay and multidrug-resistant.The combination of the Lasso regression and rare-event Logistic regression were used to explore the risk factors of TB/HIV dual infection.Results:1.Demographics characteristic of study participantsA total of 37,248 TB patients were included,ranging from 10 to 96years old,with an average age of 35.4±13.0 years.Most of them were males（66.2%）,Han nationality（90.0%）,floating population（86.0%）,living in suburban area（85.8%）,workers（57.8%）,referrals（86.9%）,and newly treated new pulmonary tuberculosis patients（96.7%）.2.Spatio-temporal clustering and incidence prediction of new pulmonary tuberculosis from 2011 to 2020（1）A total of 36037 cases of new pulmonary tuberculosis patients were reported from 2011 to 2020.The average incidence was 43.22 per100,000.Cases were mainly males（65.9%）and floating population（86.4%）.However,the local population of pulmonary tuberculosis patients was significantly increased in 2020.The main occupation was workers（58.1%）,followed by household/unemployed workers（21.1%）.The onset time was mainly from March to October,and the cases were mainly concentrated in Chang’an,Humen and Dongcheng.A total of 10249 cases of new infectious pulmonary tuberculosis patients were reported from2011 to 2020.The average incidence was 12.26 per 100,000.These cases were mainly males（65.6%）and floating population（83.0%）.However,the local population of infectious pulmonary tuberculosis patients was significantly increased in 2020.The main occupation was workers（55.1%）,followed by household/unemployed workers（23.2%）.There was no obvious seasonal trend,and the cases were mainly concentrated in Chang’an,Humen and Dongcheng.（2）Global auto-correlation analysis revealed that the incidence of pulmonary tuberculosis in Dongguan was clustered（P<0.05）,but the incidence of infectious pulmonary tuberculosis was not clustered（P>0.05）.Local auto-correlation analysis revealed that the high-incidence cluster areas of pulmonary tuberculosis and infectious pulmonary tuberculosis were northern areas.The spatial-temporal multi-component model suggests that the pulmonary tuberculosis epidemic was mainly affected by the local characteristic components and the infectious pulmonary tuberculosis epidemic was mainly affected by the spatial epidemic components.（3）ARIMA（0,1,1）（0,1,1）₁₂ was the best model for fitting the sequence of the number of pulmonary tuberculosis cases from 2011 to2019,which can be used to predict the number of cases in 2020.The results showed that the number of cases of pulmonary tuberculosis was about 292 cases less than the predicted number.ARIMA（0,1,1）（0,1,1）₁₂was the best model for fitting the sequence of the number of infectious pulmonary tuberculosis cases from 2011 to 2019,which can be used to predict the number of cases in 2020.The results showed that the number of cases of infectious pulmonary tuberculosis was about 427 cases less than the predicted number.3.Analysis of the care-seeking delay and influencing factorsThe median care-seeking day among pulmonary tuberculosis patients was 21 days,and the prevalence of the care-seeking delay was 60.3%.Multiple linear regression indicated that the risk factors for increased care-seeking days were females（β=0.16）,increasing age（β=0.14）,farmers（β=0.21）,referral due to symptom（β=0.12）,trace/other source（β=0.18）,patients complicated with other tuberculosis（β=0.54）,complication（β=0.11）,pathogen-positive patients（β=0.53）,retreatment patients（β=0.09）,severe patients（β=0.22）and patients from 2016-2020（β=0.12,all P<0.05）.Multivariate Logistic regression from 2011 to 2020 indicated that the risk factors for care-seeking delay were females（OR=1.24,95%CI=1.18～1.29）,increasing age（OR=1.22,95%CI=1.17～1.27）,farmers（OR=1.31,95%CI=1.18～1.45）,trace/other source of patients（OR=1.30,95%CI=1.22～1.40）,patients complicated with other tuberculosis（OR=2.10,95%CI=1.58～2.79）,complication（OR=1.39,95%CI=1.16～1.66）,pathogen-positivepatients（OR=1.81,95%CI=1.72～1.90）,pure tuberculous pleurisy（OR=1.29,95%CI=1.06～1.56）and patients from 2016-2020（OR=1.12,95%CI=1.08～1.17,all P<0.05）.4.Analysis of the drug resistance characteristics and influencing factorsAmong the 5271 isolates of M.tuberculosis,the total resistance rate was 33.2%,the single resistance rate was 19.6%,the poly-drug resistance rate was 6.4%and the multi-drug resistance rate was 4.0%.The resistance rates were high in streptomycin（22.5%）and isoniazid（11.0%）.The rates of resistance to anti-tuberculosis drugs were significantly higher in multidrug resistance isolates than in non-multidrug resistance isolates（P<0.001）.The Logistic regression of influencing factors for multidrug resistance indicated that the risk factor was retreatment（OR=7.30,95%CI=5.34～9.98）,the protect factor was the absence of other tuberculosis（OR=0.27,95%CI=0.10～0.75,all P<0.05）.Multiple correspondence analysis revealed that susceptibility to all anti-tuberculosis drugs was related to new treatment patient and≤30 years old patient,and resistance to 1-3 anti-tuberculosis drugs was related to31～60 years old patients,>60 years old patients,and absence of other tuberculosis patients.Resistance to 4-6 anti-tuberculosis drug was related to retreatment patient.Non-drug resistance is related to new treatment patient,≤30 years old patients,and no other tuberculosis patients,single drug resistance is related to 31～60 years old patients,poly-drug resistance was related to>60 years old patients,and multi-drug resistance was related to>60 years old patients and retreatment patients.5.Analysis of TB/HIV dual infection and influencing factorsLasso regression was subjected to 10-fold cross-validation,and 20variables were screened out and included in the model.Multivariate and rare-event Logistic regression indicated that the risk factors for TB/HIV dual infection were males（OR=3.91,95%CI=1.13～13.46）,source of patients due to symptoms（OR=8.26,95%CI=1.02～66.92）,miliary tuberculosis（OR=64.26,95%CI=7.67～538.45）,total protein increased（OR=1.17,95%CI=1.08～1.27）and eosinophil percentage increased（OR=1.22,95%CI=1.13～1.31）.And the protective factors for TB/HIV dual infection were the hemoglobin increased（OR=0.95,95%CI=0.91～0.99）,serum calcium increased（OR=0.001,95%CI=0.000～0.062）and platelet distribution width increased（OR=0.76,95%CI=0.63～0.92,all P<0.05）.Conclusions:1.The incidence of pulmonary tuberculosis in Dongguan is in the process of dynamic change,indicating that the problems of prevention and control of pulmonary tuberculosis in Dongguan is still very serious,and more attention should be paid to the prevention and control of pulmonary tuberculosis,especially for infectious pulmonary tuberculosis.We should focus on strengthening the prevention and control of pulmonary tuberculosis in cluster areas with high incidence.The number of pulmonary tuberculosis cases in Dongguan City in 2020 was slightly decrease compared with the predicted number,indicating that the prevention and control measures of COVID-19 may reduced the transmission of pulmonary tuberculosis.2.The problem of care-seeking delayed for pulmonary tuberculosis in Dongguan City needs urgent attention.It was necessary to focus on increasing tuberculosis health education for females,olders and farmers in the future,and strengthen pulmonary tuberculosis screening of suspected patients and the tracking and follow-up of the treatment of patients.3.The problems of drug resistant pulmonary tuberculosis in Dongguan should not be ignored.Special attention should be paid to the resistant problems of absence of other tuberculosis,olders and retreatment patients.The drug resistance rate and multidrug resistance rate of M.tuberculosis should be monitored continuously.4.In the future,it is necessary to focus on the problems of TB/HIV dual infection among males,patients treated for symptoms,and miliary tuberculosis.The increase of total protein,the increase of percentage of eosinophils,the decrease of absolute value of neutrophils,the decrease of hemoglobin,the decrease of serum calcium and the decrease of platelet distribution width can provide a reference for the screening and treatment of TB/HIV dual infection,and improve the screening efficiency.