Years ago I received sage advice from a surly senior surgeon: “you can do arthroplasty in a young patient,” he growled, “but you better do it perfect” (sic). Perhaps more threat than advice, it may have marked the point when I approached the young arthritic patient with “trepidation” rather than the simple “caution” that Meehan et al. recommend in the conclusion to their valuable study documenting higher early septic and aseptic failure rates in total knee arthroplasty patients under the age of fifty years. These rates were identified in more than 120,000 such patients in the California Patient Discharge Database. Caution implies taking care to avoid mistakes; trepidation describes the feeling that something untoward is imminent, no matter how much care one might take. So what should the young patient expect: poor results no matter what, or good results provided extraordinary care could be taken?

Assume perfection in the surgical technique and the best imaginable implant for a hypothetical patient. Now consider this thought experiment: if that patient remains active indefinitely, how will the arthroplasty fail? Most surgeons would agree that, barring infection, wear and particle-induced osteolysis ultimately result in loosening. Indeed, that is how we conceptualize arthroplasty in the younger patient and why we try to avoid it out of fear of failure resulting from accelerated wear: not “early” or immediate failure, but premature failure that would lead to revision arthroplasty in a patient with potentially several more decades of life still ahead. Registry data from New Zealand, evaluated by Wainwright et al., indicated that patients younger than fifty years of age at the time of primary total knee arthroplasty had a greater chance of requiring a revision than of dying, those around fifty-eight years had a 50:50 chance of needing a revision, and those older than sixty-two years would normally not outlast the prosthesis. Patients over seventy-seven years of age had a >90% chance of dying before requiring a revision, whereas those around forty-seven years were twice as likely to require a revision as to die.

The single most important surgery in the life of a young patient with knee arthritis is arguably the first revision—it is probably more important than the procedures designed to defer the initial arthroplasty, such as cartilage transplantation, meniscal allograft, or osteotomy. If done well, the first revision may well be as durable as a primary arthroplasty and endure to the patient’s final years. If it is done poorly, then subsequent revisions will likely not achieve that goal either. Accordingly, we might conclude that a durable primary arthroplasty, good for perhaps two decades, coupled with a robust first revision represents a good strategy for getting a young arthritic patient through a lifetime without catastrophic failure.

However, this is not the experience for some patients, as revealed by both Meehan et al. and Aggarwal et al. Tragically, many young arthroplasty patients in those two studies faced their first revision arthroplasty within a year, rather than after enjoying years of service, because of infection or unsatisfactory function. Our hope for decades of service from two serviceable knee reconstructions is further dashed by the data of Aggarwal et al. revealing the number of first revisions that failed prematurely, undoubtedly placing the limb in jeopardy.

We understand why younger, more active patients may “wear out” an arthroplasty sooner than their older counterparts, but why should they be more susceptible to periprosthetic infection? Shouldn’t relative youth protect against infection? Meehan et al. eliminated the usual confounding variable in studies of young arthroplasty patients by excluding inflammatory arthropathy (something that may not have been done in the paper on revision total knee arthroplasty by Aggarwal et al.). Additionally, we expect that the young patient would be less likely to suffer from medical comorbidities such as congestive heart failure, one of the four greatest risk factors for infection in the study by Meehan et al., but it is not clear whether the other three comorbidities (psychosis, obesity, and diabetes mellitus) were found at higher rates in young people and could therefore potentially explain the elevation in infection rate. Meehan et al. have reasonably postulated that posttraumatic arthritis is more common in patients under fifty years of age and we know that arthroplasties performed for that diagnosis are more complex, with compromised results. Apart from that postulate and the exclusion of inflammatory arthropathy, we know nothing regarding the arthritis severity at which arthroplasty was indicated or what surgical procedures may have preceded knee replacement.

Like all other good investigations, these papers incite us to look more closely. From the data, we cannot appreciate which patients under fifty years of age suffered the all-too-frequent cascade of unsuccessful surgeries that can culminate precipitously in arthroplasty. The groups of younger patients in such studies differ from their older counterparts in one conspicuous measure: fewer have primary osteoarthritis with no prior surgery. More of them will have had well-intentioned procedures that inevitably increase the risk of periprosthetic infection. Occasionally, some prior surgery (e.g., the usually innocuous arthroscopy) may even have failed because of undiagnosed, occult infection. The number of prior surgeries can be hypothesized to be an important possible explanation for the higher incidence of periprosthetic infection in the younger patient.

The early infection rate of 1.36% found by Meehan et al. in patients younger than fifty years of age is disappointing but not shocking. The 3.49% rate of aseptic failure within one year of surgery is of greater concern. Most, if not all, of these patients would never have enjoyed satisfactory function or pain relief after arthroplasty, much less years of activity ultimately leading to wear and failure. These early failures would be due to stiffness, instability and patellar complications, unexplained pain, “dissatisfaction,” and the somewhat more specific “chronic regional pain syndrome.” Prior surgical procedures can also increase the risk of these adverse outcomes. Additionally, prolonged pain and repeated disappointment prior to arthroplasty often increase dependence on chronic opioid medication and the risk of depression, which are in turn associated with difficulties after surgery. Noble et al. reported that “satisfaction [after total knee replacement (TKR)] correlated significantly (p < 0.001) with age less than 60, absence of residual symptoms, fulfillment of expectations, and absence of functional impairment. Satisfaction with TKR is primarily determined by patients’ expectations, and not their absolute level of function.”

Patient expectations prior to knee arthroplasty can be modified by preoperative patient education. Knee arthroplasty is a good and reliable procedure. However, although it more easily meets the needs of our senior patients in terms of pain relief and function, there is no healthy adult who would prefer the results of an arthroplasty to the performance of their own normal knee joint. Aggarwal et al. posit “unique demands” that young patients put on a total knee arthroplasty as reasons for failure. This is the conventional model of premature failure. The early failures reported are, however, more likely to result from unfulfilled unique “expectations” than from high activity levels. Unreasonable activity might lead to failure after several years, but earlier failure implies that surgical goals (i.e., expectations) were never met. Older patients might be more easily persuaded to accept an arthroplasty that could be better and thus avoid further surgery. The young patient provides an uncomfortable mirror of our abilities as surgeons: we cannot match the performance of a normal human knee joint, and this must be communicated with painful clarity to all patients, but especially to active individuals under fifty years of age.

Dissatisfaction becomes tragedy when revision is elected because of dissatisfaction with the primary arthroplasty as opposed to a diagnosis that results from a systematic and comprehensive evaluation and that provides clear and limited indications for revision arthroplasty. There is a well-defined, disciplined assessment of the problem arthroplasty that concludes whether further surgery would be expected to help (provided no complications occur). If that disciplined assessment reveals no correctable problem, then patience and later reevaluation are appropriate. Revision surgery is more likely to fail than to succeed without an accurate diagnosis and substantive pathology. Pain alone has never been a reliable indication for revision arthroplasty.