Variance estimation considering multistage sampling design in multistage complex sample analysis / 中华流行病学杂志

Multistage sampling is a frequently-used method in random sampling survey in public health.Clustering or independence between observations often exists in the sampling,often called complex sample,generated by multistage sampling.Sampling error may be underestimated and the probability of type Ⅰ error may be increased if the multistage sample design was not taken into considerationin analysis.As variance (error) estimator in complex sample is often complicated,statistical software usually adopt ultimate cluster variance estimate (UCVE) to approximate the estimation,which simply assume that the sample comes from one-stage sampling.However,with increased sampling fraction of primary sampling unit,contribution from subsequent sampling stages is no more trivial,and the ultimate cluster variance estimate may,therefore,lead to invalid variance estimation.This paper summarize a method of variance estimation considering multistage sampling design.The performances are compared with UCVE and the method considering multistage sampling design by simulating random sampling under different sampling schemes using real world data.Simulation showed that as primary sampling unit (PSU) sampling fraction increased,UCVE tended to generate increasingly biased estimation,whereas accurate estimates were obtained by using the method considering multistage sampling design.

Simulation on design-based and model-based methods in descriptive analysis of complex samples / 中华预防医学杂志

<p><b>OBJECTIVE</b>To compare design-based and model-based methods in descriptive analysis of complex sample.</p><p><b>METHODS</b>A total of 1 000 samples were selected and a multistage random sampling design was used in the analysis of the 2010 China chronic disease and risk factors surveillance. For each simulated sample, cases with probability proportional age were randomly deleted so that sample age structure was deviated systematically from that of the target population. Mean systolic blood pressure (SBP) and prevalence of raised blood pressure, as well as their 95% confidence intervals (95%CI) were determined using design-based and model-based methods (routine method and multi-level model). For estimators generated from those 3 methods, mean squared error(MSE) was computed to evaluate their validity. To compare performance of statistical inference of these methods, the probability of 95%CI covering the true parameter(mean SBP and raised blood pressure prevalence of the population) was used.</p><p><b>RESULTS</b>MSE of mean estimator for routine method, design-based analysis and multilevel model was 6.41, 1.38, and 5.86, respectively; and the probability of 95%CI covering the true parameter was 24.7%, 97.5% and 84.3%, respectively. The routine method and multi-level model probably led to an increased probability of type I error in statistical inference. MSE of prevalence estimator was 4.80 for design-based method, which was far lower than those for routine method (20.9) and multilevel model (17.2). Probability of 95%CI covering the true prevalence for routine method was only 29.4%, and 86.4% for multilevel model, both of which were lower than that for design-based method (97.3%).</p><p><b>CONCLUSION</b>Compared to routine method and multi-level model, design-based method had the best performance both in point estimation and confidence interval construction. Design-based method should be the first choice when doing statistical description of complex samples with a systematically biased sample structure.</p>

Drinking behaviors and patterns among floating population aged 18-59 years old in China, 2012 / 中华流行病学杂志

<p><b>OBJECTIVE</b>To understand the drinking behaviors and patterns among floating population in China.</p><p><b>METHODS</b>Floating population were selected through multistage clustering sampling method, stratified by 6 major industries in 170 counties and districts from 31 provinces/autonomous regions, and Xinjiang Construction Corps, in mainland China. Design-based methods were adopted to analyze the drinking behavior in subpopulations.</p><p><b>RESULTS</b>A total of 48 697 floating population aged 18-59 years were included in the study. The overall prevalence rate of drinking was 51.7% with 71.9% in men and 24.7% in women, in the last 12 months. Among those who ever drank, the weekly drinking prevalence and daily alcohol intake were 53.9% and 18.7 g in males while 16.7% and 4.1 g in females, respectively, among those floating population. Weekly drinking prevalence rates, for both genders, increased along with the increase of age (P < 0.01), but declining with having higher education level (P < 0.01). Among male drinkers, daily alcohol intake increased with age(P < 0.01) but declining with those having higher education (P < 0.01). Prevalence rates on excessive drinking, hazardous drinking and harmful drinking among male drinkers were 23.6% , 6.6% and 6.9%, respectively, but among female drinkers the figures were 6.3% , 2.3% and 1.5% , respectively. More male and older drinkers had excessive drinking, hazardous drinking or harmful drinking than the younger drinkers (P < 0.01), while less male drinkers in those having lower education (P < 0.01). 71.2% male and 33.1% female drinkers experienced at least once binge drinking in the last 12 months. Distribution of binge drinking frequencies were statistically different among male drinkers by age groups, education levels, or occupation they engaged in (P < 0.01). Among all the 6 occupations, floating population from the construction industry had the highest prevalence rates on items as weekly drinking, daily alcohol intake and prevalence of unhealthy drinking behaviors, except for hazardous drinking behaviors.</p><p><b>CONCLUSION</b>Drinking behavior was prevalent among floating population in China. Significant difference was seen between genders. Unhealthy drinking behaviors varied greatly among male drinkers at different age groups or education levels as well as among those working in the different industries.</p>

Smoking behavior among the Chinese employed floating population aged 18-59 in 2012 / 中华流行病学杂志

<p><b>OBJECTIVE</b>To understand the prevalence of cigarette smoking among employed floating population in China and to provide evidence for developing prevention and control measures targeted at cigarette smoking among them.</p><p><b>METHODS</b>On the basis of the Disease Surveillance Points (DSPs)System and employed distribution among floating population, the China Non- communicable and Chronic Disease Risk Factor Surveillance in 2012 sampled from 170 DSPs, multi-stage stratified equal-sized cluster sampling method was used and stratified by six occupational groups. Cigarette smoking related information among the employed floating population was collected by face-to-face interviews. The analytical method accounted for the complex sampling design.</p><p><b>RESULTS</b>There were 48 699 subjects, aged 18-59 years old, eligible and included in this analysis among the surveyed employed floating population. Among them, the prevalence of current smoking was 32.5% (95%CI: 32.0%-33.0%). A significantly higher percentage of men 55.3% (95%CI:54.6%-56.0%)relative to women 1.9% (95% CI:1.7%-2.1%)reported current smoking(χ(2) = 2.6, P = 0.103 7). Among men, differences in current smoking rates remained similar across age groups (χ(2) = 2.6, P = 0.103 7), but the current smoking rate tended to be lower with higher education(χ(2) = 140.7, P < 0.000 1). The current smoking rate in men was the highest(58.6% , 95% CI:57.3%-60.0%)in the floating population working in the construction industry. The prevalent of daily cigarette smoking was 27.9% (95% CI: 27.4%-28.4%), with a significantly higher proportion among men 47.8% (95% CI:47.1%-48.5%) than among women 1.3% (95%CI:1.1%-1.4%). Among daily smokers, the mean number of cigarettes smoked was 15.6(95%CI:15.5-15.8)per day(men:15.7 per day; women: 10.3 per day). The number was higher in men 15.7(95% CI:15.6-15.9)than in women 10.3(95% CI:9.3-11.3). Among daily smoking men, the daily consumed cigarettes number increased with the age increased (t = 34.89, P < 0.000 1), whereas the daily consumed cigarettes number decreased with the increase of education level (t = -14.63, P < 0.000 1). The daily consumed cigarettes number in men was the highest (18.2, 95%CI:17.9-18.6)in the floating population working in the construction industry. There were 47.1% of the daily smokers who smoked more than 20 cigarettes per day in the floating population aged 18-59, with a higher percentage among men (47.6%) than women (21.9%). The percentage in men increased with age and education level and the heavy smoking rate was the highest in the construction industry (60.4%). Among current smokers, the percentage of attempting to quit was 10.3% (95%CI: 9.7%-10.8%). A higher proportion of women 14.8% (95% CI:11.1%-18.5%), compared to men 10.1% (95% CI:9.6%-10.7%), reported attempting to quit. The rate of attempting to quit among current smokers increased with age. No significant difference in prevalence of attempting to quit by education level or occupation was seen. The percent of successfully quitting smoking was 6.1% (95%CI:5.7%-6.5%). The proportion was higher in women 7.2% (95%CI:4.6%-9.8%)than in men 6.1% (95%CI:5.7%-6.5%). The percentage of quitting smoking successfully increased with age (χ(2) = 269.0, P < 0.000 1). The percent of people who exposed to secondhand smoking was 68.7% (95%CI: 68.1%-69.3%). The percent was higher among men 76.4% (95%CI:75.5%-77.3%) than in women 64.0% (95%CI:63.2%-64.9%).</p><p><b>CONCLUSION</b>Smoking was more common among men in the employed floating population aged 18-59 years old, especially in the construction industry. Low percent was seen among those who attempting to quit, and the ones who successfully quit were very few.</p>

Fruit and vegetables intake among the Chinese migrant population aged 18 to 59 years old in 2012 / 中华流行病学杂志

<p><b>OBJECTIVE</b>To describe the intake of fruit and vegetables among employed migrant population aged 18 to 59 year-olds in China.</p><p><b>METHODS</b>Data from the Migrant Population Survey related to China Chronic Disease and Risk Factor Surveillance that conducted in 170 counties/districts in 31 provinces, 2012, was used. Information on non-communicable diseases and related risk factors among migrant population were collected through face-to-face questionnaire interview, physical measurement and lab tests. A total of 48 704 subjects aged 18 to 59 years old were included in our study. Sample was standardized by age and sex. Information on average daily fruit and vegetables intake, prevalence of low fruit and vegetables intake, grouped by sex, age, industries, and education level were analyzed.</p><p><b>RESULTS</b>The average daily intakes of vegetables and fruits were 353.7 (95%CI:351.3-356.2) g and 125.1 (95%CI:123.4-126.9) g respectively, among the employed migrant population aged 18-59 years old in China. Prevalence of low fruit and vegetables intake was 44.1% (95% CI:43.5%-44.6% ) among employed migrant population, 46.2% (95% CI: 45.5%-47.0%)for males and 41.2% (95% CI:40.3%-42.0%)for females (χ(2) = 82.19, P < 0.05). Among different professions, the prevalence of low fruit and vegetables intake was the highest among people working in accommodation and restaurants (46.2%, 95%CI:45.0%-47.3%) while the lowest seen among those working in social services (42.5%, 95%CI:41.4%-43.7%,χ(2) = 15.81, P < 0.05). The prevalence of low fruit and vegetables intake showed a decrease along with the increase of education levels (χ(2) = 22.29, P < 0.05).</p><p><b>CONCLUSION</b>In 2012, more than 40% of the employed migrant population aged 18 to 59 years old in China had low fruit and vegetables intake. Being male and with low education level were risk factors linked with the higher prevalence of low fruit and vegetables intake.</p>

Physical activities and sedentary behavior among the Chinese floating population aged 18-59 in 2012 / 中华流行病学杂志

<p><b>OBJECTIVE</b>To evaluate the prevalence of leisure-time physical activity and sedentary behaviors among Chinese floating population aged 18-59.</p><p><b>METHODS</b>Data was from the Chinese Floating Population Chronic Disease Surveillance (2012). Of the 48 704 subjects aged 18-59 years old, information on frequency and duration of leisure activity was collected by face-to-face interview. Time spent on watching TV, reading, using computers or playing games after work was also calculated.</p><p><b>RESULTS</b>19.4% (95%CI:19.0%-19.7%) of the subjects took part in moderate or vigorous intensity physical activities at least 10 minutes for three days per week, with percentage higher in males than in females and increasing among younger males (P < 0.01). Data showed that it was opposite among females (P < 0.01). Physical activities increased in those with higher education for both males and females (P < 0.01) and showed significant differences on occupations. Among those who engaged in different occupations, subjects working on social service and manufactory presented the top two highest percentages on often engaging in leisure time activities and they spent 4.9 (95%CI:4.8-4.9) hours daily on average in sedentary behavior decreasing along with aging but increasing with having received higher education.</p><p><b>CONCLUSION</b>The percentage of taking part in moderate and vigorous activities among floating population was generally low. Prevention should target on floating population, especially on females, those with lower education or at older age.</p>