Prostheses for primary total knee replacement in Italy

Background Total knee replacement (TKR) is known as an effective procedure for patients with osteoarthritis (OA). The wide array of prosthesis models on the market makes the issuing of evidencebased guidance on choice, a priority for the equitable and transparent use of resources in any health system.

The assessment of the clinical value of each model, from the viewpoint of a national health system, requires data on:
i) which models are used in practice;
ii) acceptable estimates of performance for those models;
iii) their price.

Aim We aimed to identify the TKR prostheses currently used in Italy, to retrieve evidence of performance for all the models identified from clinical studies and arthroplasty registers, and to construct a cost-effectiveness model to support the choice for healthcare payers.

Methods We first performed a context-specific analysis to identify which prosthesis models are currently in use and their purchase prices. We involved 172 public hospitals and private accredited providers of the Italian national health service. Then we carried out a systematic review of evidence from clinical studies and an analysis of implant performance from arthroplasty registers. We were interested in recipients aged 50 and above that had received primary TKR and had been followed for at least 5 years. Finally, with the purchase prices and the effectiveness estimates identified, we carried out an economic evaluation to compare the different models.
XII

Results Literature in this field was too generic to answer our research questions. Very few studies met our inclusion criteria (4 primary and one registry-based studies) and those that did were poorly reported. For this reason we decided to give more weight to performance estimates from periodic reports of the arthroplasty registers. We received full collaboration by one national and one regional register (the Australian AOAJRR and the Emilia-Romagna RIPO respectively). Unfortunately, heterogeneity of reporting formats and attitudes, and low data-sharing ethos, caused a sizeable loss of data on costs and effectiveness. We were unable to make all the planned comparisons.

According to our calculations the most cost-effective among the prostheses with multiradius femoral design is NEXGEN LPS (ZIMMER), whereas SCORPIO (STRYKER) is the most cost-effective single-radius prosthesis. Little can be said for the remaining models, but our ICERs (incremental cost-effectiveness ratios) provide at least some benchmarking for similar prostheses with sufficient survival data in the future.

Conclusions Despite the limitations and constraints of our study we were able to show that purchase price and effectiveness are not sufficient to inform decision makers. What is required is the linking of the two into ICERs within prosthetic design classes. It is essential however that performance is calculated on recipient populations which are broadly similar in terms of age, gender, indications and whenever possible, BMI.