Antibiotics survey: 77% of adults who overstated symptoms got the drugs

Of the 268 US adults who told researchers they had exaggerated or invented symptoms to get antibiotics, 206 (76.9%) said it had worked at least once, according to a survey published on October 30 in JAMA Network Open. Headlines will focus on the 27.9% who admitted doing it, but the authors call their estimates preliminary, because the sample was not a probability sample of the country.

Exaggerating was far more common than inventing

The survey ran online from January 8 to February 13, 2026, through the market research firm Dynata, and 959 responses were analysed after the team removed 26 likely bots and 28 people whose written answers raised validity concerns.

Most of the behaviour was stretching rather than pure invention: 249 people (26.0%) had exaggerated real symptoms, and 160 (16.7%) had fabricated some. Only 19 people had fabricated without ever exaggerating, while 141 had done both.

Cold-like symptoms were the usual vehicle, named by 65.7% of those who had misrepresented. The authors note that antibiotics do nothing against viruses, which cause most cold- and flu-like illness.

Who says they did it, and why

The top reasons sat close together: feeling really sick (62.2%), believing antibiotics were the only thing that would help (61.8%), and needing to get back to responsibilities sooner (61.0%).

Respondents who often felt clinicians did not take their symptoms seriously had about twice the odds of misrepresenting (adjusted odds ratio 2.21). So did those who needed help reading health materials (2.38). Adults aged 18 to 33 and 34 to 55 had roughly three times the odds of those 65 and older, while people who knew more about antibiotics were less likely to report it (0.72).

The regret figures are modest. Of those who misrepresented, 28.7% regretted doing so, and 27.2% regretted taking the resulting antibiotics.

A companion paper in the same journal found that 27% of respondents had done this for a child in their care, according to University of Utah Health.

The limits the authors put on their own numbers

The University of Utah Health release is headed "One in four Americans," but the paper is more cautious. It describes a nonprobability sample with quotas for age, gender identity, race and ethnicity, and region, and calls the estimates hypothesis-generating.

The authors say the true share could be higher: people may underreport out of embarrassment, or misremember, since the question asked whether they had ever done it. The survey also did not measure how often it happened or what tactics people used.

The authors do not want doctors to respond with suspicion. "The practical takeaway is not for clinicians to become more suspicious of their patients," said co-first author Andrea Levy of Connecticut State Community College Middlesex. The team instead argues for better communication, so that patients feel heard without having to raise the stakes.

The open question is how often these prescriptions were actually needed, which a survey of this kind cannot say.