It’s a generally accepted “fact” that total knee arthroplasty (TKA) ranks among the most significant modern medical advancements. But the October 22, 2015 NEJM published the first rigorously controlled randomized study that “proves” that “fact” by comparing TKA to nonsurgical management.
One hundred patients with moderate-to-severe knee osteoarthritis were randomly assigned to undergo TKA followed by 12 weeks of rigorous nonsurgical treatment, or the nonsurgical treatment alone. Over a 12-month follow-up period, TKA was superior to nonsurgical treatment in terms of pain relief and functional improvement, but it was also associated with a higher number of serious adverse events, including deep-vein thrombosis and infection.
The study authors concluded that “the benefits and harms of the respective treatments underscore the importance of considering patients’ preferences and values during shared decision making about treatment for moderate-to-severe knee osteoarthritis.” JBJS Deputy Editor Jeffrey Katz, MD concurred with that conclusion in an accompanying editorial: “Treatment decisions should be shared between patients and their clinicians and anchored by the probabilities of pain relief and complications and the importance patients attach to these outcomes,” he wrote. “Each patient must weigh these considerations and make the decision that best suits his or her values.”
This randomised trial article is plagued with various issues, with the fatal flawed analysis of serious adverse events (SAE) in which the authors commented the total-knee-replacement (TKR) group had more SAE than nonsurgical-treatment group.
Much of the issues are easily found by looking at the supplementary data provided: http://www.nejm.org/doi/suppl/10.1056/NEJMoa1505467/suppl_file/nejmoa1505467_appendix.pdf
It is mind boggling so the authors to consider that “baseline characteristics were similar in the two study groups.” The TKR group are 20% more likely living alone (36% vs non TKR gp 16%, p = 0.02) and 14 % more likely to be employed (full/part-time, sick leave, p = 0.02). These characteristics alone can significantly influence (possibly reduced) the difference in KOOS result, and thus the clinically significant effect size between the 2 treatments.
The study SAEs are based on FDA’s definition: Death, Life-threatening, Hospitalization (initial or prolonged), Disability or Permanent Damage, Congenital Anomaly/Birth Defect, Required Intervention to Prevent Permanent Impairment or Damage (Devices), Other Serious (Important Medical Events) which “may jeopardize the patient and may require medical or surgical intervention (treatment) to prevent one of the other outcomes. Examples include allergic brochospasm (a serious problem with breathing) requiring treatment in an emergency room, serious blood dyscrasias (blood disorders) or seizures/convulsions that do not result in hospitalization. The development of drug dependence or drug abuse would also be examples of important medical events.” (http://www.fda.gov/Safety/MedWatch/HowToReport/ucm053087.htm)
In the intention-to-treat analysis, only the TKR group has post index TKR occurence of opposite knee pain resulting in opposite side TKR (considered an SAE) and TKR has remarkably excessive gastrointestinal plus other non-knee SAEs not apparently related to TKR itself *; these issues are not properly addressed or discussed by the authors even though these SAEs are clearly flagged in both the results and conclusion in the abstract!
Outcome measurement at 12 months belies the sustained benefits (> 10 years) of TKR when faced with expected clinical progression & deterioration of osteoarthritis over time; one would expect a more reasonable comparison in KOOS over 2 -5 years. It is also only then can one detect other SAEs like opioid dependence and abuse.
The issue of usage of pain medication at 12 months (one of the clearly defined secondary outcomes) is of interest as well but the result not addressed in the article but can only be found in the supplementary data.
Dr Katz, although threading carefully to maintain good relations with NEJM (a journal known for it’s limited openness in allowing discussion in various controversial studies involving knee arthroscopy), has alas missed an opportunity to bring NEJM readers’ attention to these issues.
*Musculoskeletal SAEs only in the TKR group: pain in the opposite knee leading to TKR (n=3) and back pain leading to spinal fusion.
Skin SAE only in non-TKR gp: 1 X melanoma.
Gastrointestinal SAEs occuring only in TKR group: Carcinoid tumor in the small intestine; abdominal pain with outpouching of the colon; and hiatal hernia.
Other SAEs in non TKR gp: a cerebral thrombosis; coronary thrombosis; hospitalization due to suspicion of cerebral thrombosis caused by neurological symptoms in the legs; and metastatic breast cancer. In TKR gp, this included hospitalization due to depression and anxiety; malignant breast cancer (n=2); atrial fibrillation; leukemia; pneumonia; hospitalization due to retinal detachment;
trauma to the cranium from a fall leading to hospitalization; and Myelomatosis.
In total SAEs on site other than index knee is 5 in non TKR vs 16 in TKR (p = 0.04)