Menon Chandrakala, Gipson Katrina, Bowe Whitney P, Hoffstad Ole J, Margolis David J
Department of Dermatology and Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, Philadelphia, PA, USA.
Dermatology. 2008;217(2):164-8. doi: 10.1159/000136655. Epub 2008 Jun 5.
BACKGROUND/OBJECTIVE: Acne prevalence studies often use subject self-report as data source. Our aim was to evaluate the validity of acne self-report.
Responses of university students to an acne questionnaire were compared to the trained observer's concurrent examination of acne. The validity of self-report was measured by sensitivity, specificity, positive predictive value and negative predictive value. Agreement was measured by Cohen's kappa and correct classification percentage.
The sensitivity of self-report was 0.55 (95% CI 0.47-0.63), the specificity was 0.72 (95% CI 0.63-0.80), the positive predictive value was 0.70 (95% CI 0.61-0.78), and the negative predictive value was 0.57 (95% CI 0.49-0.65). Cohen's kappa was 0.26 (95% CI 0.15-0.38) and correct classification percentage was 63.
Validity of self-report was moderate at best and agreement was fair, indicating that college students could not accurately report that they have acne. This is likely not sufficient for clinical or research activities or to assure that individuals who self-guide their acne therapy actually have acne.
背景/目的：痤疮患病率研究通常将受试者自我报告作为数据来源。我们的目的是评估痤疮自我报告的有效性。
将大学生对痤疮问卷的回答与经过培训的观察者对痤疮的同步检查结果进行比较。自我报告的有效性通过敏感性、特异性、阳性预测值和阴性预测值来衡量。一致性通过科恩kappa系数和正确分类百分比来衡量。
自我报告的敏感性为0.55（95%置信区间0.47 - 0.63），特异性为0.72（95%置信区间0.63 - 0.80），阳性预测值为0.70（95%置信区间0.61 - 0.78），阴性预测值为0.57（95%置信区间0.49 - 0.65）。科恩kappa系数为0.26（95%置信区间0.15 - 0.38），正确分类百分比为63。
自我报告的有效性充其量为中等，一致性为一般，这表明大学生无法准确报告自己是否患有痤疮。这对于临床或研究活动可能是不够的，也无法确保自行指导痤疮治疗的个体确实患有痤疮。
