The Good and the Bad of Evidence-Based Medicine and the Challenge of Practising in a Time of Increased Connectivity and Artificial Intelligence

As we move well into the twenty-first century, the concept of evidence-based medicine is well established, and surgeons will become more obliged to practise medicine along these lines. Of course they should, we hear commentators say, and it certainly does make sense to have evidence for what we do when we choose to operate on people. The difficulty is deciding on what level of evidence is sufficient—at one extreme, we may think that knowing that something has ‘worked well in our hands’ should be enough to justify what we do, whereas at the other extreme, randomized controlled trials (RCT) or meta-analyses thereof may be considered by some as required to really have sufficient evidence. We all know the flaws of the first extreme, having been taught well by history and proper follow-up that our intuition isn’t always right, but the second extreme also has flaws, in that many conditions aren’t suited to RCTs and even with those that are, statistical methods averaging out results will tend to miss those “outsiders” who may have benefited from surgery when the average result found that surgery was not beneficial. We all know that those paying for surgery—government bodies and private health funds—are constantly looking for ways to reduce costs and declining to pay for procedures with insufficient ‘evidence’ is one way they will always be looking at to achieve this. We are already seeing the impact this is having on arthroscopic knee surgery, not long ago the most common orthopaedic sports medicine procedure performed. It won’t be enough for surgeons to expect funding for a procedure just because they believe it is right, and surgeons will need to be proactive and take the lead in providing the necessary evidence by studying carefully and scientifically what they do. Once funding is removed from a procedure, it will be very hard to get it back.