School of Commerce, University of South Australia, Adelaide, Australia.
Chennai Mathematical Institute, Chennai, India.
PLoS One. 2018 Aug 27;13(8):e0202786. doi: 10.1371/journal.pone.0202786. eCollection 2018.
Researchers interested in the effect of health on various life outcomes (such as employment, earnings and life satisfaction) often use self-reported health and disease status as an indicator of true, underlying health status. Self-reports appear to be reasonable measures of overall health. For example, self-assessed overall health has been found to be a reliable predictor of mortality. However, the validity of self-reports is questionable when investigating specific diseases such as diabetes and hypertension. A small and nascent body of research comparing self-reported status on certain diseases with the true status based on clinical diagnoses has found significant gaps. These validation exercises predominantly use data from high-income countries. In this paper, we use survey data from India to compare self-reports of disease prevalence to diagnostic tests conducted on the same individuals. We focus on hypertension and lung disease, two of the primary causes of death in India. We find that self-reported measures substantially understate the true disease burden for both conditions. The attenuation bias from using self-reports is over 80 percent for both diseases, and bigger than estimates from high-income countries. We test and reject the hypothesis that self-reports of the disease status are identical to the true disease status in expectation. We identify characteristics associated with false negative reporting (reporting not having the disease but testing positive for it) for both diseases. The large awareness gap between self-reports and true disease burden indicates multiple deficiencies in India's public health policy. The survey data depicts limited access to medical facilities, high levels of health illiteracy, low rates of health insurance, and other barriers related to poverty and lack of equity in the delivery of health services. These factors prevent timely intervention for managing health and controlling disease, invariably leading to morbidity and often to premature death.
研究人员对健康对各种生活结果（如就业、收入和生活满意度）的影响感兴趣，他们通常使用自我报告的健康状况和疾病状况作为真实潜在健康状况的指标。自我报告似乎是衡量整体健康状况的合理方法。例如，自我评估的总体健康状况被发现是死亡率的可靠预测指标。然而，当研究特定疾病（如糖尿病和高血压）时，自我报告的有效性值得怀疑。少数比较某些疾病的自我报告状况与基于临床诊断的真实状况的新兴研究发现存在显著差距。这些验证工作主要使用来自高收入国家的数据。在本文中，我们使用来自印度的调查数据将疾病流行的自我报告与对同一人群进行的诊断测试进行比较。我们关注高血压和肺部疾病，这是印度主要死亡原因中的两种。我们发现，自我报告的疾病患病率大大低估了这两种疾病的真实疾病负担。两种疾病的自我报告都存在超过 80%的衰减偏差，比高收入国家的估计值还要大。我们检验并拒绝了自我报告的疾病状况与真实疾病状况在预期中相同的假设。我们确定了与两种疾病的假阴性报告（报告没有疾病但测试结果呈阳性）相关的特征。自我报告和真实疾病负担之间的巨大差距表明，印度公共卫生政策存在多种缺陷。调查数据描绘了获取医疗设施的机会有限、健康知识水平低、健康保险率低以及与贫困和提供卫生服务的公平性相关的其他障碍。这些因素阻碍了及时干预来管理健康和控制疾病，不可避免地导致发病率增加，并且经常导致过早死亡。
