Time to Put Aside the Controversy Between Total Hip Arthroplasty and Hemiarthroplasty

Commentary Should an individual with a displaced femoral neck fracture be treated with a total hip arthroplasty or a hemiarthroplasty? The debate is ongoing, but the details being debated and the basis for the debate vary in different countries. Treatment traditions, such as lower and upper age limits toward other treatment options, and organizational issues, such as the availability of skilled surgeons, will influence the way that we utilize the 2 types of interventions. The evidence base must be read in the light of strict patient selection to the randomized trials. Most often, eligible patients are between the ages of 60 and 75 years and are active and healthy, as is the case in the 4 randomized controlled trials (RCTs) showing better patient-reported outcome and function after total hip arthroplasty compared with hemiarthroplasty1-4. The other 4 RCTs more reflect the ordinary population with hip fracture and do not show any difference between methods in this aspect5-8. Registry studies, such as the current study, are a valuable complement to the RCTs, in particular when the results of RCTs are conflicting. Registry studies also have the benefit of better external validity than RCTs, reflecting outcome in non-selected patients treated by non-selected surgeons. Ravi et al. describe an association between total hip arthroplasty and lower revision, based on health-care databases. Studying the same type of data, Jameson et al.9 found a similar revision rate, but a higher dislocation rate, after total hip arthroplasty compared with hemiarthroplasty. Two studies based on national orthopaedic registries have conflicting results; Moerman et al. found that total hip arthroplasty to treat a fracture was associated with a higher revision rate10, and Hansson et al. found that total hip arthroplasty was associated with fewer revisions11. Registry studies may be flawed with confounding by indication (i.e., factors that lead the clinician to choose one treatment over the other will also affect outcome). When such variables are available, a proper registry study design will adjust for them. For the choice between total hip arthroplasty and hemiarthroplasty, frailty in patients is considered. Frailty is an elusive trait, for which simple proxy variables are lacking. Frailty may influence both the risk of complications and the decision whether to perform a secondary procedure. Patients with more frailty usually undergo a hemiarthroplasty, and, when facing hip complications, they may be advised against undergoing a revision surgical procedure. However, as discussed in the current study, it may be more tempting to revise a hemiarthroplasty (by adding a cup) than to exchange components of a total hip implant inserted in osteoporotic bone. Hence, an outcome defined as a revision surgical procedure might not be the most objective. A prosthesis left unrevised does not equal success. The 3 most common hip complications after any arthroplasty to treat hip fracture are dislocation, infection, and periprosthetic fracture. Other surgical factors besides the type of articulation will influence the risk of each individual complication. Ravi et al. do not report on periprosthetic fracture. One can agree that the stem type and the canal fit, not the articulation, affect the risk for subsequent fracture. Multiple factors, both patient-related and treatment-related, influence the risk of infection. The longer duration of a total hip arthroplasty might imply a higher risk of infection. Therefore, it is regrettable that the current study does not analyze whether their limited set of codes have resulted in underreporting of infections. The aforementioned RCTs showed an infection rate of 2% to 7% (i.e., higher than the current study)1,2,4,7,8. The dislocation rate is also higher in the RCTs: up to 7% after total hip arthroplasty with a lateral approach2, and as high as 18% after total hip arthroplasty with a posterior approach7. Regrettably, it seems that Ravi et al. underestimate the dislocation rate with their figure of 1.7% for fracture-related total hip arthroplasty. As they do not provide information on the surgical approach, it is difficult to assess the magnitude of any underreporting. Ravi et al. have done a praiseworthy effort to compare total hip arthroplasty and hemiarthroplasty in terms of complications and costs as well, not just looking at the outcome of a revision surgical procedure. Still, there is uncertainty with regard to the true incidence of dislocation in particular. In spite of several RCTs and registry studies, there has been, to my knowledge, no convincing evidence supporting either arthroplasty intervention over the other in the general population with hip fracture. In addition, other surgical decisions such as surgical approach and use of bone cement will influence the clinical outcome to a larger extent than the articulation. Perhaps now is the time to put aside the controversy between total hip arthroplasty and hemiarthroplasty, stating that we have 2 well-functioning implant types, as long as we are referring to elderly patients with femoral neck fracture who are no longer able-bodied. Patients with hip fracture may benefit more from us directing our research efforts onto other, more pressing issues.