论文部分内容阅读
目的为了解在现场调查中采用自报高血压、糖尿病、体重、腰围的可行性。方法利用2013年昆山市慢性病行为危险因素调查数据,获得体重、腰围的自报值与实测值,计算自报准确率并分析影响准确率的因素;以实际测量得的诊断血压、血糖水平为金标准,计算自报高血压、糖尿病的灵敏度、特异度、ROC曲线下面积(AUC)等。结果调查人群体重、腰围的知晓率分别为96.8%、83.0%,自报准确率分别为62.7%、69.5%;通过Logistics回归分析发现,性别、年龄、教育水平、家庭人均月收入、BMI分级等因素对体重自报准确率没有影响,而性别、教育水平、家庭人均月收入对腰围自报准确率有影响;自报高血压的灵敏度为89.0%,特异度为77.6%,诊断价值中等(AUC=0.83),自报糖尿病的灵敏度为85.3%,特异度为95.1%,诊断价值较高(AUC=0.90);上述因素的不同亚组,2种病的自报诊断价值亦不同。结论昆山市成年人体重、腰围自报准确率,高血压、糖尿病自报的诊断价值均在可接受范围。要更有效的运用自报数据,仍需在指标选择、研究设计、分析方法上不断改进。
Purpose To understand the feasibility of using self-reported hypertension, diabetes, weight, and waist circumference in field surveys. Methods According to the survey data of behavioral risk factors of chronic diseases in Kunshan City in 2013, the self-reported values and measured values of body weight and waist circumference were obtained, the self-reported accuracy rate was analyzed and the factors influencing the accuracy rate were analyzed. Based on the measured blood pressure and blood glucose level Criteria, calculate self-reported hypertension, diabetes sensitivity, specificity, area under the ROC curve (AUC) and so on. Results The awareness rates of body weight and waist circumference were 96.8% and 83.0% respectively, and the self-reported accuracy rates were 62.7% and 69.5% respectively. According to the logistic regression analysis, gender, age, educational level, per capita monthly household income, BMI classification, etc. Factors of self-reported accuracy rate of body weight had no effect, while gender, education level, per capita monthly household income on waist circumference self-reported accuracy rate of 89.0% reported self-reported hypertension, specificity was 77.6%, the diagnostic value of the medium (AUC = 0.83). The sensitivity of self-reported diabetes mellitus was 85.3%, specificity was 95.1%, and the diagnostic value was high (AUC = 0.90). The diagnostic value of different subgroups and 2 diseases of these factors were also different. Conclusion Kunshan City, adults, self-reported accuracy of body weight, high blood pressure, diabetes, the diagnostic value is within the acceptable range. To use self-reported data more effectively, we still need to continuously improve on the selection of indicators, research design and analysis methods.