Randomized trials are for clinicians who are uncertain as to whether they are right or not—and I am certain I am right.
Sir Ian Chalmers, quoting an unnamed orthopedic surgeon
This chapter opens with an exercise we use to introduce the concept of evidence-based practice to final-year students in human movement sciences/ kinesiology (i.e. non-clinicians). If you are an experienced clinician or an expert on evidence-based practice you may want to skip over this chapter!
The “case” for the students to consider involves Mrs J, a 55-year-old woman. Students are told she presents with persistent knee pain due to osteoarthritis. The students are given the information in the box (below) and are asked to suggest a treatment for Mrs J.
Which evidence carries most weight?
You are asked to advise Mrs J, a 55-year-old woman with knee osteoarthritis, as to whether or not knee arthroscopy is a good idea. You have your own personal opinion, and you obtain the following four pieces of further information. Which of the four options carries the most weight with you? Would you advise that surgery is a good idea?
Dr X, an expert knee surgeon, advises in favor of surgery because “I have done hundreds of these operations and obtained good or excellent results in over 90% of them.” The surgeon offers you and your friend the phone numbers of patients who can provide testimonials. You call a few of these patients and they all vouch for surgery.
A published study of cases done by another surgeon, Dr Y, shows that 75% of patients who have had this type of surgery reported improvements. Overall 75% of patients had an “excellent or good” outcome. Patients were recruited and interviewed two years after the surgery.
A published study examined patients who had presented with knee pain to a specialist in osteoarthritis two years earlier. One group of patients had undergone arthroscopic surgery, the other had not. Patients who had undergone surgery reported playing more golf and tennis than those who had not undergone arthroscopic surgery. The paper concluded that surgery was associated with superior outcomes compared to conservative management.
A physiotherapy student obtained ethics approval to attend doctors’ offices and recruit patients with knee osteoarthritis. The surgeon decided to allocate patients randomly to either “surgery” or “no surgery.” Two years later, the student interviewed the patients again and found that both groups of patients (those who had had surgery and those who had not) had similar levels of pain and function. Both groups had pain scores of around 50 out of 100 where 100 is severe pain.
In our student exercise we call for a vote and every year the first ballot results in option 1 (surgery) receiving about 80% of the votes! We then lead an open-class discussion and emphasize that the quality of the data should carry more weight than ...