In a national survey of 959 US adults, 27.9% said they had exaggerated or made up symptoms to obtain antibiotics, most often cold-like symptoms, with implications for outpatient stewardship.
Nearly 30% of US adults reported exaggerating or fabricating symptoms to a clinician to obtain antibiotics, according to the authors of a survey study published in JAMA Network Open.¹ Cold-like symptoms were misrepresented most often. Of those who misrepresented symptoms, 76.9% reported obtaining antibiotics this way at least once.¹
Outpatient settings account for most human antibiotic use in the US, with approximately 255.9 million prescriptions issued annually.¹ Clinicians rely on patient-reported symptom onset, severity, and duration when deciding whether antibiotics are warranted. Inaccurate reporting can therefore lead to inappropriate prescribing and drive antimicrobial resistance (AMR).¹
The investigators recruited a national, nonprobability sample of US adults for an online survey conducted from January 8 to February 13, 2026.¹ Of 1371 eligible participants, the final sample included 959 respondents with a mean (SD) age of 50.7 (17.6) years. Of these, 478 (49.8%) were female, and quotas ensured representation across age, race and ethnicity, and US Census region.¹
Overall, 268 respondents (27.9%) reported ever misrepresenting symptoms to a clinician to get antibiotics.¹ Among them, 108 (40.3%) only exaggerated existing symptoms, 19 (7.1%) only fabricated symptoms, and 141 (52.6%) did both.¹
Cold-like symptoms were misrepresented most often (176 [65.7%]), followed by stomach symptoms (132 [49.3%]) and joint or muscle pain (124 [46.3%]).¹ Most respondents who misrepresented symptoms (177 [66.0%]) were at least moderately confident the tactic would work. They learned what to say from their own experience (43.1%), friends or family (34.5%), clinicians (30.7%), social media (21.7%), and AI chatbots (11.6%).¹
A companion study of pediatric encounters found 27% of respondents had misrepresented symptoms to get antibiotics for a child in their care, according to a University of Utah Health news release.²
The most commonly endorsed reasons were:
More than half (54.1%) said they know their body and medical history better than a health care practitioner. Nearly one-third (32.3%) said they do not think antibiotic resistance is real.¹
In an exploratory multiple logistic regression, respondents aged 18 to 33 years (odds ratio [OR], 3.20; 95% CI, 1.58-6.63) had higher odds of misrepresentation than those 65 years or older. So did respondents aged 34 to 55 years (OR, 3.05; 95% CI, 1.55-6.17).¹ Higher odds were also linked to often feeling clinicians do not take symptoms seriously (OR, 2.19; 95% CI, 1.49-3.22) and needing help reading health materials (OR, 2.40; 95% CI, 1.62-3.54).¹
Believing AMR is a hoax (OR, 1.98; 95% CI, 1.34-2.93) and higher lying-acceptability scores (OR, 1.49; 95% CI, 1.23-1.80) also raised the odds.¹ Greater antibiotic knowledge (OR, 0.73; 95% CI, 0.58-0.91) and dislike of taking antibiotics (OR, 0.89; 95% CI, 0.81-0.99) were associated with lower odds.¹
Only 77 respondents (28.7%) regretted misrepresenting symptoms. Their regret stemmed more often from adverse effects or lack of efficacy than from concerns about dishonesty.¹ Fewer than half (45.1%) named someone they would listen to if told not to misrepresent symptoms again.¹
“It should be a priority for us to identify strategies for reassuring people, re-earning trust, and finding the right ways to address their concerns,” said co-first author Alistair Thorpe, PhD, research assistant professor of population health sciences at University of Utah Health, in a statement.²
Co-first author Andrea Gurmankin Levy, PhD, MBe, professor of social sciences at Connecticut State Community College Middlesex, warned clinicians against responding with distrust.
“The practical takeaway is not for clinicians to become more suspicious of their patients,” Levy said.² Instead, she noted, clinicians should build interactions where patients feel heard enough to have no reason to exaggerate.²
The authors noted the nonprobability sample is not fully representative of the US population, so estimates should be viewed as hypothesis-generating.¹ Social desirability bias and inaccurate recall may have led to underreporting. The authors called for strategies to improve antibiotic knowledge and patient-clinician communication.¹
