Anecdotes are not evidence

Like anyone else who writes about medical or health topics when reporting and commenting on science, I always get responses with anecdotes about what was written with it either does or does not apply to them. Anecdotes and testimonials always turn up in the comments sections. While they may be important to the individual, anecdotes are of no value when trying to move forward the science and the understanding of a topic and they are simply not that helpful.

For example, on one of my previous posts (that I am too lazy to look up), I had one comment from a runner that his Hoka One One running shoes caused his plantar fasciitis. The very next post was from a runner who swore that his Hoka One One cured his plantar fasciitis! Surely that is enough “evidence” of how useless anecdotes or testimonials are. What should a clinician do when treating a runner with plantar fasciitis in the context of those two anecdotes? Should they use Hoka running shoes to treat plantar fasciitis or should they advise all their runner patients with plantar fasciitis to avoid the Hoka running shoes?

Having said that, unfortunately anecdotes can be powerful convincers as we have a tendency to empathize with an individual personal story and can relate it to themselves. These accounts are subjective and prone to so many forms of error such as wishful thinking, self-deception, confirmation biases and post hoc reasoning (1). I have previously blogged in more detail on that and how ineffective treatments can work (natural history, placebo, self-delusion etc). Because of that, there is no scientific value to anecdotes and they are certainly not any form of evidence that stacks up to scrutiny.

In the medical literature a type of anecdote is the case report, typically this involves the reporting of an unusual or rare condition. What brought this up for me today was this case report in which a runner with plantar fasciitis got better after taking up barefoot running.

Reduction in Plantar Heel Pain and a Return to Sport After a Barefoot Running Intervention in a Female Triathlete With Plantar Fasciitis
Peter Francis, Cassie Oddy Mark Johnson
International Journal of Athletic Therapy and Training
Patient characteristics:
In a 27 year-old female triathlete, magnetic resonance imaging revealed mild thickening and oedema at the calcaneal insertion of the plantar fascia in keeping with a degree of plantar fasciitis.
After 6 weeks of conservative treatment failed to elicit a return to sport, the patient engaged in 6 sessions of barefoot running (15 – 30 minutes) on a soft grass surface, without further conservative treatment.
Comparative outcome:
After two sessions of barefoot running the patient was pain free before, during and after running. This outcome was maintained at the 6 week follow-up period.
This is the first case report to use barefoot running as a treatment strategy for chronic heel pain. Barefoot running has the potential to reduce the load on the plantar fascia and warrants further investigation using a case series.

Where does one start (after the facepalms and eye rolling has finished)? There are any number of reasons that the plantar fasciitis could have got better and it could have easily had nothing to do with the barefoot running. It could easily have been the natural history. It could have easily been the treatment that was done before the “two sessions” of barefoot running. It could have been … etc.

As for this “Barefoot running has the potential to reduce the load on the plantar fascia” – barefoot running does not reduce the load on the plantar fascia.

I and many others could easily publish case reports (anecdotes) in which barefoot running increased the symptoms of plantar fasciitis! See how useless these case reports and anecdotes are. They add no value to the literature.

You could also head over to any of the patient support groups for those with plantar fasciitis in places like Facebook and suggest barefoot running as a treatment for plantar fasciitis and you will get ridiculed. I have previously posted about minimalism as an intervention for plantar fasciitis and the fan boys responded in the comments with their anecdotes that I was wrong!

This also reminded me of another case report (anecdote) about plantar fasciitis clearly up after starting a gluten-free diet. Based on the above, surely it is obvious why no one is taking it seriously. At least the authors in this case were honest enough to subtitle the paper: “Relationship or just coincidence?

Anecdotes can be legitimate, I use them a lot in my lectures and courses to support claims and illustrate the point, but it is always based on other evidence. The purpose here is the anecdote or story is that it is the anchor that I use to hang the evidence on. This facilitates learning and long-term memory in students.

I like this comment on the anecdote: “The problem with anecdotal evidence comes from the capacity humans have to let their personal needs, prior expectations, attitudes, prejudices, and biases unwittingly influence the outcome of their evaluations.” Source.

As always, I go where the evidence takes me until convinced otherwise …. and anecdotes are not evidence and barefoot running is not a panacea for plantar fasciitis.

Francis, P., Oddy, C., & Johnson, M. (2017). Reduction in Plantar Heel Pain and a Return to Sport After a Barefoot Running Intervention in a Female Triathlete With Plantar Fasciitis International Journal of Athletic Therapy and Training, 1-17 DOI: 10.1123/ijatt.2016-0072

Last updated by .


6 Responses to Anecdotes are not evidence

  1. AnonymousCoward March 30, 2017 at 3:32 pm #

    It seems to me that anecdotes and such case studies are an extremely important aspect of science – they contribute inspiration.

    For example, say we are living quite a long time ago. A peasent comes up to us and says “usually when I throw stones in the water, they sink. But one time I threw a stone in the water and it floated.”

    We might think that they are lying, or that they mistook a peice of wood for a stone. But we can give them the benefit of the doubt and follow them to the river where they found the floating rock. And maybe all their rocks will sink, and maybe they’ll pick up some wood – but hey, maybe we’ll discover pumice!

  2. Mark Cucuzzella March 31, 2017 at 11:19 am #

    Craig i am one of your fan boys as you call us. here is another anecdote from 30 years of running marathons.
    My plantar fascia and foot muscles are bulletproof. the evidence is the physical exam of my feet.

    • Craig Payne March 31, 2017 at 6:27 pm #

      So how do you explain all the anecdotes of barefoot making plantar fasciitis worse?

  3. David Cashley March 31, 2017 at 12:05 pm #

    “See how useless these case reports and anecdotes are. They add no value to the literature.”

    So disappointed you made this statement Craig. The case report and the case series mark the beginning of an intervention. A hunch to be explored. An educated guess to be examined. The promise of something interesting and noteworthy that may lie around the next corner or that may turn out to be a blind alley. Without the case study, science is dead. It is the “I wonder” and the “what if” that predates the RCTs. Are you suggesting that we should perform RCTs based solely on hunches? Good luck getting them funded. Case studies are quick cheap pointers that may just lead us in the right direction and as such they are profoundly valuable. Furthermore, given the increasing computer power that we have at our disposal for statistical analysis, it is also true to say that well-designed studies of N=1 can now give us some powerful data that was never the case before. In a profession like ours where research funding is hens teeth the case study may prove to be the saving grace, so lets not throw the baby out with the bath water just yet.

    • Craig Payne March 31, 2017 at 6:28 pm #

      Anecdote: Hoke’s caused my plantar fasciitis
      Anecdote: Hoka’s cured my plantar fasciitis
      –> anecdotes useless to base clinical reasoning and decision making on.

  4. Cassie Oddy May 16, 2017 at 9:27 am #

    Thanks for the interest in our case report guys. The development of testable hypotheses in this area is worthwhile and paves the way for design of future empirical research
    Best wishes

Leave a Reply