Could doctors misdiagnose you because they read newspaper?

Patients assume doctors to diagnose diseases using purely their own knowledge and judgement without any bias. But what if we expose the doctors to media information about a disease? For example, if first they read something about a disease in newspaper or internet, and then they examine their patients. Would they misdiagnose similar looking clinical cases because of that?

Researchers designed an interesting experiment to answer this question.

The flamingo is ready to present this research.

Hi and hola para amigos!

The experiment consisted of three phases.

In the first phase, to expose 38 internal medicine residents to disease information on the internet, half of the these doctors were asked to evaluate the accuracy of the Wikipedia entry for Legionnaires’ disease, the other half evaluate the entry for Q fever. Then they returned to their daily clinical work.

In the second phase, which has started six hours later, they were requested to diagnose eight written cases. This activity was presented as if this is an unrelated activity. To show it as an unrelated study, the researchers made some tricks such as the invitation letters came from another institution, the materials had different letterheads and font type, etc.

These eight cases had different diagnoses than the ones they saw six hours ago. But two cases had similar signs and symptoms to Legionnaires’ disease, the other two had similar to Q fever. The remained four cases acted as filler which were neutral to the experiment. The participants were allowed to diagnose these eight cases in 10 minutes. The low amount of time was due to getting them to use nonanalytical reasoning.

In the third phase, they were provided four cases that they diagnosed in the previous phase. These were two filler cases and two similar-looking cases. They were asked to re-evaluate their diagnosis using some reflection techniques that include a structured process such as writing and ranking all possible diagnoses in order of likelihood. 30 minutes were allocated for the third phase to get them to use their analytical reasoning.

That’s the design. Yes, there were more details that shows the rigor of the study but I don’t want the video to make longer. The study has an excellent methodology.

Anyway, let’s jump into the results. Did reading Wikipedia six hours ago affected their diagnostic reasoning?

Well, yes! In the Phase 2, the participants misdiagnosed the cases that are similar to the disease they’ve read on Wikipedia than the cases that they haven’t read six hours ago (0.61 vs 0.29). Also the researchers reported that “mean diagnostic accuracy scores were significantly lower on bias-expected cases than on bias-not-expected cases”.

These results show that if we rely only on our non-analytical reasoning, which is fast and requires minimal effort, we are prone to availability bias, which has been defined as “the tendency to weigh likelihood of things by how easily they are recalled”. It seems this bias “may emerge from exposure to disease information in the media”.

If it is so easy to be captured by this bias, how can we prevent ourselves? The remedy is reflection by using our analytical reasoning. Phase 3 results show that this is effective. The researchers reported that “deliberate reflection restored performance on bias-expected cases to pre-bias levels”. Yes, it is slow and requires more cognitive effort but prevents harmful effects for patients.

The results are dramatic but it is worth keeping in mind the limitation of the study. It was conducted with written cases. As Dr. Satid pointed out, the issue is complex from the lens of situated learning theory. So we may see different results if we carry out this research in a clinical environment.

If you want to read more about the study, you can find the link at the description section of the video:

See you and adios para amigos.

And also, don’t forget the flamingo.

Follow the Flamingo on Twitter: https://twitter.com/MedEdFlamingo