Background: Controversy still exists as to the optimum management of the posterior cruciate ligament (PCL) in total knee arthroplasty. Surgeons can choose to kinematically substitute the PCL with a posterior-stabilized total knee replacement or alternatively to utilize a cruciate-retaining, also known as minimally stabilized, total knee replacement. Proponents of posterior-stabilized total knee replacement propose that the reported lower survivorship in registries when directly compared with minimally stabilized total knee replacement is due to confounders such as selection bias because of the preferential usage of posterior-stabilized total knee replacement in more complex or severe cases. In this study, we aimed to eliminate these possible confounders by performing an instrumental variable analysis based on surgeon preference to choose either posterior-stabilized or minimally stabilized total knee replacement, rather than the actual prosthesis received.

Methods: Cumulative percent revision, hazard ratio (HR), and revision diagnosis data were obtained from the Australian Orthopaedic Association National Joint Replacement Registry from September 1, 1999, to December 31, 2014, for 2 cohorts of patients, those treated by high-volume surgeons who preferred minimally stabilized replacements and those treated by high-volume surgeons who preferred posterior-stabilized replacements. All patients had a diagnosis of osteoarthritis and underwent fixed-bearing total knee replacement with patellar resurfacing.

Results: At 13 years, the cumulative percent revision was 5.0% (95% confidence interval [CI], 4.0% to 6.2%) for the surgeons who preferred the minimally stabilized replacements compared with 6.0% (95% CI, 4.2% to 8.5%) for the surgeons who preferred the posterior-stabilized replacements. The revision risk for the surgeons who preferred posterior-stabilized replacements was significantly higher for all causes (HR = 1.45 [95% CI, 1.30 to 1.63]; p < 0.001), for loosening or lysis (HR = 1.93 [95% CI, 1.58 to 2.37]; p < 0.001), and for infection (HR = 1.51 [95% CI, 1.25 to 1.82]; p < 0.001). This finding was irrespective of patient age and was evident with cemented fixation and with both cross-linked polyethylene and non-cross-linked polyethylene. However, the higher revision risk was only evident in male patients.

Conclusions: There was a 45% higher risk of revision for the patients of surgeons who preferred a posterior-stabilized total knee replacement compared with the patients of surgeons who preferred a minimally stabilized total knee replacement.