Ohio University, United States of America.
Ohio University, United States of America.
J Psychosom Res. 2022 Aug;159:110948. doi: 10.1016/j.jpsychores.2022.110948. Epub 2022 May 25.
Over the last decade, the use of online labor markets to collect data in health science has grown exponentially. However, self-identification remains the most common method for recruiting specific clinical sub-populations, and this may adversely affect data validity among respondents motivated to feign a condition for financial gain.
Online respondents who professed taking medication for a specific medical condition (sample 1: diabetes: N = 307; sample 2: pain: N = 506) were asked to upload an image of their prescribed medication. These images were then evaluated to identify authentic and inauthentic responders based on the images submitted. Authentic and inauthentic respondent groups were then compared on a series of condition-specific health measures and attention checks.
In the diabetes sample, respondents whose photos were deemed inauthentic passed fewer attention checks and reported poorer physical (e.g., number of comorbidities) and mental health (e.g., diabetes distress) across a wide variety of measures (η = 0.014-0.159). Similarly in the pain sample, respondents whose photos were deemed inauthentic reported poorer physical (e.g., pain interference) and mental health (e.g., depression) across a wide variety of measures (η = 0.008-0.129).
The present findings suggest that there may be substantial exaggeration of adverse health among online survey respondents who feign health conditions such as diabetes and chronic pain. Hence, in the absence of procedures to verify health status claims, the validity of data from online survey respondents should be viewed with skepticism.
在过去十年中，利用在线劳动力市场在健康科学领域收集数据的方式呈指数级增长。然而，自我认同仍然是招募特定临床亚人群的最常见方法，而这可能会对那些为了经济利益而故意伪装某种疾病的受访者的数据有效性产生不利影响。
自称正在服用某种特定药物治疗特定疾病的在线受访者（样本 1：糖尿病：N=307；样本 2：疼痛：N=506）被要求上传其处方药物的图像。然后，根据提交的图像，评估这些图像以识别真实和虚假的响应者。然后，在一系列特定于疾病的健康指标和注意力检查中比较真实和虚假响应者群体。
在糖尿病样本中，被认为是不真实的照片的受访者通过的注意力检查较少，并且在各种指标上报告了较差的身体（例如，合并症数量）和心理健康（例如，糖尿病困扰）（η=0.014-0.159）。同样，在疼痛样本中，被认为是不真实的照片的受访者在各种指标上报告了较差的身体（例如，疼痛干扰）和心理健康（例如，抑郁）（η=0.008-0.129）。
本研究结果表明，在那些假装患有糖尿病和慢性疼痛等健康状况的在线调查受访者中，可能存在对不良健康状况的大量夸大。因此，在缺乏验证健康状况声明的程序的情况下，应该对来自在线调查受访者的数据的有效性持怀疑态度。
