| Objectives: The first aim of this study is to evaluate the representativeness of sample and quality of the data obtained from HIV screening population of the AIDS demonstration area in Henan province of “Twelfth five-year” major science and technology project, and to estimate the positive ratio of HIV and morbidity accurately. The second aim is to explore the problems existing in the data and their causes, so as to provide reference basis for further field epidemiological investigation in Henan province.Methods: A multi-stage random cluster sampling was performed to choose 2112766 residents from Shangcai, Yongcheng, Queshan, Shenqiu and Jiyuan of Henan province during September 2012 to June 2013. The questionnaire survey and physical examination were conducted. People who had first physical examination need take part in examination for the second time during October 2014 to March 2015. Sample representativeness, data integrity, accuracy and entry timeliness were applied to evaluate the quality of the data. Sample representativeness was analyzed from the two aspects of goodness fit of test and Myer’s index. Integrity evaluation included missing rates and coincidence rate. Accuracy was analyzed from the repetition rate and logic error rate. In addition, composite index method was used to evaluate and compare the quality of the data obtained from five regions.Results: 1 Representativeness The composition of the population was same as that of the sixth census of the counties in 2010(2χ =8.202, P=0.879). There were no statistically significant differences between five regions and sixth census with respect to their compositions of the population(Shangcai:2χ =8.630,P =0.854;Yongcheng:2χ =8.228,P =0.877;Queshan:2χ =14.126,P =0.440;Shenqiu:2χ =12.238,P =0.587).The overall Myer’s index was 4.146 for two selected age groups in 10-69 and 20-79 years, indicating that the distribution of age was proportional. The Myer’s index in Shangcai, Yongcheng, Queshan, Shenqiu and Jiyuan were 2.884, 3.844, 8.333, 6.251 and 4.747, respectively. 2 Integrity The overall integrity rate of the data from questionnaire and AIDS files were 82.356% and 87.462%, respectively. The missing rates of questionnaire for five counties were as follows: Shenqiu(10.324%), Jiyuan(13.513%), Queshan(16.582%), Shangcai(19.495%), and Yongcheng(23.680%). The missing rates of AIDS files were as follows: Queshan(8.364%), Jiyuan(10.119%), Shangcai(12.809%), Shenqiu(13.333%), and Yongcheng(20.269%). The coincidence rate of two physical examinations was 85.793% as a whole. The coincidence rate for five counties were as follows: Queshan(97.465%), Jiyuan(84.545%), Shangcai(83.883%), Yongcheng(83.824%), and Shenqiu(83.283%). 3 Accuracy The rates of duplicate entry were 0.132% and 0.637% based on personal information and physical examination information, respectively. The percentage of people who had the history of AIDS, but HIV screening results showing negative was 7.328%. The inconsistent rate of gender based on ID card and self-writing was 0.209%. The logic error rate of check-up date later than entry was 1.171%. 4 Timeliness The time interval from the first and second physical examinations to inputting ranged from 0 to 386 day and 0 to 388 day, respectively. The median time was 0 and 20 days, separately. 5 The composite index for five rejions were as follows: Shangcai(2.914), Jiyuan(3.507), Yongcheng(4.319), Queshan(7.964) and Shenqiu(10.919).Conclusions:(1) The representativeness of sample from the AIDS demonstration area in Henan province meets the preliminary requirement, and there is no artificial sampling bias. Data integrity, accuracy and entry timeliness meet the requirement of task design, while it still needs to be improved further.(2) The quality of data in sequence from high to low is Shenqiu, Queshan, Yongcheng, Jiyuan, Shangcai. |
PDF Full Text Request