Résumé
In recent years, with the improvement of various surveillance network, surveillance system has become an important data source for ecological study. Different data types, including cross-sectional data, time series data and panel data, containing abundant information involving exposure, outcome and confoundings. Gradually, some new statistical methods have been developed or improved for the special structural characteristics of surveillance data. In this paper, we summarized the principles of these models, preconditions, as well as their advantages and limitations.
Résumé
Objective@#To study the lag effect of temperature and the source of heterogeneity on other infectious diarrhea (OID) in Zhejiang province, so as to identify related vulnerable populations at risk.@*Methods@#Data on OID and meteorology in Zhejiang province from 2014 to 2016 were collected. A two-stage model was conducted, including: 1) using the distributed lag non-linear model to estimate the city-specific lag effect of temperature on OID, 2) applying the multivariate Meta- analysis to pool the estimated city-specific effect, 3) using the multivariate Meta-regression to explore the sources of heterogeneity.@*Results@#There were 301 593 cases of OID in Zhejiang province during the study period. At the provincial level, temperature that corresponding to the lowest risk of OID was 16.7 ℃, and the temperature corresponding to the highest risk was 6.2℃ (RR=2.298, 95%CI: 1.527- 3.459). 16.7 ℃ was recognized as the reference temperature. P5 and P95 of the average daily temperature represented low and high temperature respectively. When the temperature was cold, the risk was delayed by 2 days, with the highest risk found on the 5th day (RR=1.057, 95%CI: 1.030-1.084) before decreasing to the 23rd day. When the temperature got hot, the risk of OID occurred on the first day (RR=1.081, 95%CI: 1.045-1.118) and gradually decreasing to the 8th day. Differences on heterogeneous sources related to the risks of OID in different regions, presented on urban latitude and the rate of ageing in the population.@*Conclusions@#Both high or low temperature could increase the risk of OID, with a lag effect noticed. Prevention program on OID should be focusing on populations living in the high latitude and the elderly population at the low temperature areas.
Résumé
Objective To understand the influence of diurnal temperature range (DTR) on influenza incidence in the elderly in Beijing and to conduct a subgroup analysis.Methods The incidence data of daily influenza cases in the elderly and daily meteorological data from 2014 to 2016 in Beijing were collected for this study.A generalized additive model (GAM) was used to explore whether the relationship between daily influenza cases and DTR is a linear one.A distributed lag non-linear model (DLNM) was established to quantify the lagged effect of DTR on daily influenza incidence in the elderly.The model was also used to estimate the effects of DTR on daily influenza incidence among various subgroups.Results A total of 4 097 influenza cases in the elderly were notified during study period.The mean DTR was 10.153 ℃.A linear relationship between daily influenza incidence and DTR was detected by using GAM.DTR was significantly associated with daily influenza incidence between lag0 and lag5 with a maximal effect at lag0.An 1 ℃ increase of DTR was associated with a 2.0％ increase in daily influenza incidence in the elderly (95％CI:0.9％-3.0％).The RR values of males,females,people aged 60-69 years,people aged ≥70 years were 1.018 (95％CI:1.005-1.032),1.021(95％CI:1.007-1.035),1.012 (95％CI:1.002-1.022),1.025 (95％CI:1.012-1.039),respectively.The influencing time of DTR on females (lag6) was longer than males (lag2).Conclusions DTR was associated with increased risk of influenza in the elderly in Beijing.It is necessary to take targeted measures in the elderly to control the incidence of influenza when DTR becomes greater.
Résumé
Objective To understand the influence of diurnal temperature range (DTR) on influenza incidence in the elderly in Beijing and to conduct a subgroup analysis.Methods The incidence data of daily influenza cases in the elderly and daily meteorological data from 2014 to 2016 in Beijing were collected for this study.A generalized additive model (GAM) was used to explore whether the relationship between daily influenza cases and DTR is a linear one.A distributed lag non-linear model (DLNM) was established to quantify the lagged effect of DTR on daily influenza incidence in the elderly.The model was also used to estimate the effects of DTR on daily influenza incidence among various subgroups.Results A total of 4 097 influenza cases in the elderly were notified during study period.The mean DTR was 10.153 ℃.A linear relationship between daily influenza incidence and DTR was detected by using GAM.DTR was significantly associated with daily influenza incidence between lag0 and lag5 with a maximal effect at lag0.An 1 ℃ increase of DTR was associated with a 2.0％ increase in daily influenza incidence in the elderly (95％CI:0.9％-3.0％).The RR values of males,females,people aged 60-69 years,people aged ≥70 years were 1.018 (95％CI:1.005-1.032),1.021(95％CI:1.007-1.035),1.012 (95％CI:1.002-1.022),1.025 (95％CI:1.012-1.039),respectively.The influencing time of DTR on females (lag6) was longer than males (lag2).Conclusions DTR was associated with increased risk of influenza in the elderly in Beijing.It is necessary to take targeted measures in the elderly to control the incidence of influenza when DTR becomes greater.
Résumé
Objective To analyze the influencing factors on condom use among 35-yearsor-older female sex workers (OFSWs) in Qingdao by using the three-level logistic model.Methods From March to June 2014,OFSWs were recruited in Qingdao,using respondent-driven sampling.Related information on OFSWs and their recent five sexual partners (not including husband) were obtained by conducting a questionnaire survey on OFSWs.A Three-level logistic regression model was conducted to analyze the influencing factors of condom use between OFSWs and their sexual partners.Results A total of 420 OFSWs participated the survey as well as information on 2 100 sexual partners.Results from the empty model showed that the use of condoms among OFSWs having an aggregation that related to the levels of working sites and their own behaviors.Results from the three-levels of logistic model analysis showed that,OFSWs that having had junior middle school education (OR=1.450,95％CI:1.054-1.994)/high school education or above (OR=2.264,95％ CI:1.215-4.222),knowing the function of condom use (OR=2.004,95％ CI:1.273-3.154) would have higher rates of condom use.OFSWs with higher score of attitude on condom use (OR=0.796,95％CI:0.745-0.849),having had syphilis infections in the past (OR=0.657,95％ CI:0.478-0.902) would have lower rate of condom use.For the sexual partners,the rate of condom use among OFSWs' regular partners were higher than that of OFSWs' boyfriends (OR=15.291,95％CI:8.441-27.700;OR=29.032,95％ CI:15.413-54.682).Conclusion Condom use of OFSWs was affected by behaviors of both OFSWs themselves and their sexual partners.Prevention and control programs should focus on OFSWs and their sexual partners at the same time.The key intervention contents should include target populations as:OFSWs with low level of education,having had infections of syphilis,those who do not use condoms with their trusted partners.
Résumé
Objective To analyze the influencing factors on condom use among 35-yearsor-older female sex workers (OFSWs) in Qingdao by using the three-level logistic model.Methods From March to June 2014,OFSWs were recruited in Qingdao,using respondent-driven sampling.Related information on OFSWs and their recent five sexual partners (not including husband) were obtained by conducting a questionnaire survey on OFSWs.A Three-level logistic regression model was conducted to analyze the influencing factors of condom use between OFSWs and their sexual partners.Results A total of 420 OFSWs participated the survey as well as information on 2 100 sexual partners.Results from the empty model showed that the use of condoms among OFSWs having an aggregation that related to the levels of working sites and their own behaviors.Results from the three-levels of logistic model analysis showed that,OFSWs that having had junior middle school education (OR=1.450,95％CI:1.054-1.994)/high school education or above (OR=2.264,95％ CI:1.215-4.222),knowing the function of condom use (OR=2.004,95％ CI:1.273-3.154) would have higher rates of condom use.OFSWs with higher score of attitude on condom use (OR=0.796,95％CI:0.745-0.849),having had syphilis infections in the past (OR=0.657,95％ CI:0.478-0.902) would have lower rate of condom use.For the sexual partners,the rate of condom use among OFSWs' regular partners were higher than that of OFSWs' boyfriends (OR=15.291,95％CI:8.441-27.700;OR=29.032,95％ CI:15.413-54.682).Conclusion Condom use of OFSWs was affected by behaviors of both OFSWs themselves and their sexual partners.Prevention and control programs should focus on OFSWs and their sexual partners at the same time.The key intervention contents should include target populations as:OFSWs with low level of education,having had infections of syphilis,those who do not use condoms with their trusted partners.
Résumé
Objective To evaluate the risk factors of fluil-term infants with low birth weight.Methods All related Chinese and English literatures published from 1980 to February 25,2016 were collected from CBM,CNKI,Wang Fang Data,Medline and Embase databases,and screened with inclusion and exclusion criteria and Stata13.0 software was used in this Meta-analysis.Results Twenty three studies were included and there were 278 020 subjects.Female infants (pooled OR=1.60,95％CI:1.49-1.72),less antenatal care visits (pooled OR=1.81,95％CI:1.54-2.11),maternal passive smoking (pooled OR=1.49,95％CI:1.08-2.06),pregnancy-induced hypertension (pooled OR=2.96,95％CI:1.85-4.74) and hypamnion (pooled OR=2.71,95％CI:1.87-3.93) were the risk factors for fluil-term infants with low birth weight.Conclusion Departments of maternal and health care should encourage pregnant women to have antenatal care visits to find and treat their pregnancy complications,and avoid passive smoking actively through health education for the purpose to prompt the birth quality of infants.
Résumé
Objective To understand the influence of floods on bacillary dysentery in Liaoning province.Methods The monthly surveillance data of bacillary dysentery,floods,meteorological and demographic data in Liaoning from 2004 to 2010 were collected.Panel Poisson regression analysis was conducted to evaluate the influence of floods on the incidence of bacillary dysentery in Liaoning.Results The mean monthly morbidity of bacillary dysentery was 2.17 per 100 000 during the study period,the bacillary dysentery cases mainly occurred in during July-September.Spearman correlation analysis showed that no lagged effect was detected in the influence of floods on the incidence of bacillary dysentery.After adjusting the influence of meteorological factors,panel data analysis showed that the influence of floods on the incidence of bacillary dysentery existed and the incidence rate ratio was 1.439 4 (95％CI:1.408 1-1.471 4).Conclusion Floods could significantly increase the risk of bacillary dysentery for population in Liaoning.
Résumé
Objective To evaluate the risk factors of fluil-term infants with low birth weight.Methods All related Chinese and English literatures published from 1980 to February 25,2016 were collected from CBM,CNKI,Wang Fang Data,Medline and Embase databases,and screened with inclusion and exclusion criteria and Stata13.0 software was used in this Meta-analysis.Results Twenty three studies were included and there were 278 020 subjects.Female infants (pooled OR=1.60,95％CI:1.49-1.72),less antenatal care visits (pooled OR=1.81,95％CI:1.54-2.11),maternal passive smoking (pooled OR=1.49,95％CI:1.08-2.06),pregnancy-induced hypertension (pooled OR=2.96,95％CI:1.85-4.74) and hypamnion (pooled OR=2.71,95％CI:1.87-3.93) were the risk factors for fluil-term infants with low birth weight.Conclusion Departments of maternal and health care should encourage pregnant women to have antenatal care visits to find and treat their pregnancy complications,and avoid passive smoking actively through health education for the purpose to prompt the birth quality of infants.
Résumé
Objective To understand the influence of floods on bacillary dysentery in Liaoning province.Methods The monthly surveillance data of bacillary dysentery,floods,meteorological and demographic data in Liaoning from 2004 to 2010 were collected.Panel Poisson regression analysis was conducted to evaluate the influence of floods on the incidence of bacillary dysentery in Liaoning.Results The mean monthly morbidity of bacillary dysentery was 2.17 per 100 000 during the study period,the bacillary dysentery cases mainly occurred in during July-September.Spearman correlation analysis showed that no lagged effect was detected in the influence of floods on the incidence of bacillary dysentery.After adjusting the influence of meteorological factors,panel data analysis showed that the influence of floods on the incidence of bacillary dysentery existed and the incidence rate ratio was 1.439 4 (95％CI:1.408 1-1.471 4).Conclusion Floods could significantly increase the risk of bacillary dysentery for population in Liaoning.
Résumé
Several new varients related to the case-control designs have been developed in the recent decades,and this article briefly summarized four new designs:two-stage design,case-specular study,exposure-crossover study and case-case-time-control study.This paper involved principles of study design,requisites for application,advantages and disadvantages on all the studies.
Résumé
BACKGROUND: Short-form-36 health survey(SF-36) has good reliability and validity, which is applicable to assess quality of life (QOL) in European and American population.OBJECTIVE: To study the reliability, validity and feasibility of SF-36 scale in the rural elder and evaluate their QOL.DESIGN: It is a cross-sectional study, all participants are sampled by two-phase random sampling.SETTING: Institute of Epidemiology and Hygienic Statistics, China Union Medical University and institute of epidemiology and hygienic statistics in a university.PARTICIPANTS: Totally 687 old people were randomly selected from six villages in Bao' an District of Shenzhen from May to July 2002. Inclusion criteria: All the people were residents and aged 60 years or above. Exclusion criteria: People with severe auditory or mental obstacles were excluded. Totally 666 questionnaires were valid, among them 264 were male and 402 were female.METHODS: All participants were interviewed face to face with SF-36 questionnaire by trained investigators. The feasibility, reliability and validity were the main measures of SF-36 scale.MAIN OUTCOME MEASURES: Split-half reliability, internal consistency, criterion-related validity, discriminant validity and construct validity.RESULTS: SF-36 scale had a good internal consistency, and all Cronbach' α coefficients were equal or more than 0. 8 in 8 dimensions. The split-half reliability of every dimension was preferable and Pearson correlation coefficients were 0.7 or higher except mental health(MH) . Discriminant validity and criterion related validity were obtained and satisfied. Two factors were got by factor analysis, which could account for 63.14% of total variance. Every dimension held a moderate loading in its relevant factor except role-emotion (RE) dimension(≥0.4).CONCLUSION: SF-36 scale is applicable for QOL assessment in the rural elder after some item revisions. The vulnerable elder should be given more attention and care.
Résumé
Objective: To study the reliability, validity and feasibility of SF-36 Health Survey Scales in elderly living in rural area around Shenzhen of south China. Method: 666 old people aged 60 or above were assessed with SF-36. Split-half reliability, internal consistency, criterion-related validity, discriminant validity and construct validity of the scale were tested. Result: All Cronbach alpha coefficients were 0.8 or above in eight subscales. The internal consistency was good. The split-half reliability of every subscale was preferable and Person correlation coefficients were 0.7 or higher, except that of Mental Health. Good discriminant validity and criterion related validity were obtained. Factor analysis showed two factors, which could account for 63% of total variance. Every subscale held moderate loading in its relevant facto, except that of Role-Emotion.Conclusion:SF-36 was applicable for life quality assessment in country-dwelling elderly population after some item revisions.