Randomized Controlled Trial supporting Total Knee Replacement

Photo by Eraxion (stockxpert)

            Total knee replacements are the standard of care when it comes to surgical treatment of end-stage knee osteoarthritis that does not respond to conservative measures. It is one of the most predictable operations in orthopaedics in terms of outcomes. The number of total knee replacements has been increasing from year-to-year. The number of total knee replacements performed annually increased from 31.2 per 100,000 person-years between 1971-1976 to 220.9 between 2005 and 2008.

A short aside to explain that statistic; a person year is equivalent to 1 person during the period of 1 year, so 2 people over the course of 1 year make 2 person-years, or 1 person over the course of 2 years also makes 2 person-years.

            In either case, the number of total knee replacements performed is staggering. In 2012 670,000 total knee replacements were performed for a total cost of over $36.1 billion. Total knee replacements are one of the most thoroughly studied topics in orthopaedics. Searching for ‘total knee arthroplasty’ on pub-med (the largest collection of scholarly publications) produces over 20,000 results. We have good literature reporting on the longevity of total knee replacements, on problems that can occur with them, the impact on patient related pain and function levels. However, up until the October issue of New England Journal of Medicine, we did not have the highest possible quality evidence comparing total knee replacements with non-operative management.

            The highest quality type of study in medicine is a randomized controlled trial. This involves two groups, one for each treatment being compared. Patients are randomized to the groups, and are then followed to monitor for effect. The reason that this is the highest quality type of study is that the randomization ensures that the two comparison groups are the same.

The scientific term for this is minimizing ‘confounders’. A theoretical example of this is a study that looks at cab drivers versus the general population and the rates of lung cancer (lets remember that this is theoretical and not a real study). Since no randomization was performed, it’s possible that the study finds cab drivers have higher rates of lung cancer, however this may be because a higher proportion of cab drivers smoke cigarettes compared to the general population. This would be an example of smoking acting as a confounder in the study.

            The randomized controlled trial from Denmark looking at total knee arthroplasty randomized 100 people to getting a total knee replacement versus getting non-surgical treatment (exercise, diet, insoles and pain medications). These patients were then followed for 1 year. The study found that at 1 year follow up the patients that underwent a total knee replacement had significantly less pain and more improvement in function compared to the non-operative group.

These were not unexpected findings, but still important to see that these findings are confirmed in a high quality RCT. The study did find a couple of interesting things.

            The total knee replacement group had significantly more serious adverse events. These included problems such as infection, blood clots, and significant stiffness. Another interesting finding was that the majority of patients in the non-operatively group chose not to undergo a total knee replacement during the year following the end of the study.

            While the results of this study aren’t surprising, it is an important addition to the literation. It’s also important to take note of how much of orthopaedics doesn’t have the highest quality evidence supporting that decision making. The biggest reason for this is that randomized controlled trials in surgery are logistically difficult to carry out, randomization is hard to enforce (there are ethical problems with telling a patient they will or will not get surgery based on the roll of a dye), and incredibly expensive. 

Continue Reading