Factors affecting the risk of erroneous interpretation of online surgeon rating websites among the general population.
121 Background: Several websites present estimated individual surgeon complication rates for surgeons in both the United States and Great Britain. Though some researchers have raised questions as to the validity and appropriateness of these publicly displayed outcome measures, there remains very little research into how the general public may interpret these data to make health care decisions. Methods: We invited attendees of the 2016 Minnesota State Fair who met entry criteria, (adults > 18 years old, English speakers who were able to use a tablet computer) to complete our survey. Demographic data was presented along with various screen shots from online surgeon rating websites. Patients were then asked to interpret these graphics and report complication rates. Some graphics displayed complications rates for one surgeon alone, while others compared multiple surgeons side-by-side. Results: 392 participants completed the survey from a broad geographic distribution from the upper Midwest (179 unique zip codes). Median age was 49 (Interquartile range 28-61), the female:male ratio was 3:2, 57% had completed a college or graduate degree and 85% were Caucasian vs. 15% ethnic minorities. The majority of participants (76%) were able to correctly estimate complication rates when a single surgeon and his or her complication rates were shown, but when respondents were asked to compare/rank multiple surgeons, respondents overestimated complication rates by 5-7 fold, on average, for the lower ranking surgeons and only 15% of respondents could correctly identify the complication rate of the lowest performing surgeon. College graduates and those with a graduate degree were more likely to correctly estimate complication rates compared to participants with less education (odds ratio 1.98 95% CI 1.04-3.75, p = 0.035). Conclusions: Online surgeon rating websites that compare and rank surgeons may lead the general public to drastically overestimate the risk of postoperative complications. These errors in estimating complication rates appear to be reduced amongst those who are college educated and when viewing single surgeon outcomes in the absence of a comparison.