Abstract
IMPORTANCE: Patient failure to disclose medically relevant information to clinicians can undermine patient care or even lead to patient harm. OBJECTIVE: To examine the frequency of patients failing to disclose to their clinicians information that is relevant to their care and their reasons for doing so. DESIGN, SETTING, AND PARTICIPANTS: Two national nonprobability samples were recruited to participate in an online survey, one using Amazon's Mechanical Turk (MTurk) from March 16 to 30, 2015 (2096 respondents), followed by one using Survey Sampling International (SSI) from November 6 to 17, 2015 (3011 respondents). Data analysis was conducted from September 28 to October 8, 2018. After dropping respondents meeting the exclusion criteria, the final sample sizes were 2011 (MTurk) and 2499 (SSI). MAIN OUTCOMES AND MEASURES: The primary outcome measures were self-reported nondisclosure of 7 types of information to their clinician (eg, did not understand instructions, medication use) and reasons for nondisclosure (eg, embarrassment, not wanting to be judged). RESULTS: There was a total of 4510 overall respondents. Of 2096 respondents, 2013 completed the MTurk survey (96.0% completion rate) and 2011 were included in the analysis. Of 3011 respondents, 2685 completed the SSI survey (89.2% completion rate) and 2499 were included in the analysis. The mean (SD) age of the participants was 36 (12.4) years for MTurk and 61 (7.59) years for SSI. Both samples were predominantly white (MTurk: 1696 [84.3%]; SSI: 1968 [78.8%]). A total of 1630 MTurk participants (81.1%) and 1535 SSI participants (61.4%) avoided disclosing at least 1 type of information. Disagreeing with the clinician's recommendation (MTurk: 918 of 2010 respondents [45.7%]; SSI: 785 of 2497 respondents [31.4%]) and not understanding the clinician's instructions (MTurk: 638 of 2009 respondents [31.8%]; SSI: 607 of 2497 respondents [24.3%]) were the most common occurrences. The most commonly reported reasons for nondisclosure included not wanting to be judged or lectured (MTurk: 81.8% [95% CI, 79.8%-83.9%]; SSI: 64.1% [95% CI, 61.5%-66.7%]), not wanting to hear how harmful the behavior is (MTurk: 75.7% [95% CI, 73.5%-78.0%]; SSI: 61.1% [95% CI, 58.5%-63.8%]), and being embarrassed (MTurk: 60.9% [95% CI, 58.9%-62.9%]; SSI: 49.9% [95% CI, 47.8%-52.1%]). In both samples, participants who were women (MTurk: odds ratio [OR], 1.88 [95% CI, 1.49-2.37]; SSI: OR, 1.38 [95% CI, 1.17-1.64]), younger (MTurk: OR, 0.98 [95% CI, 0.97-0.99]; SSI: OR, 0.98 [95% CI, 0.97-0.99]), and with worse self-rated health (MTurk: OR, 0.87 [95% CI, 0.76-0.99]; SSI: OR, 0.80 [95% CI, 0.72-0.88]) were more likely to report withholding information. CONCLUSIONS AND RELEVANCE: Many respondents in these surveys intentionally withhold important information from their clinicians and were most likely to do so when they disagreed with or misunderstood their clinician's instructions. A better understanding of how to increase patients' comfort with reporting this information may improve the clinician-patient relationship and patient care.