Complications after the Introduction of Well-Documented Components in Primary Total Hip Arthroplasty

Total hip arthroplasty (THA) is usually a very successful procedure that reduces pain and improves patients’ physical abilities. The results after THA have improved thanks to decades of development in implant designs and surgical techniques. However, there is always a risk for unexpectedly high complication rates when new components are introduced in a new institution. Currently, the most common major complication types in THA are prosthesis dislocation, periprosthetic infection, and periprosthetic femoral fracture (PFF). For dislocation-prone patients, unconventional implants that offer extra stability, such as dual-mobility cups and constrained liners, are increasingly being used in primary THA. In patients at high risk for PFF, cemented stems are preferred.

Hospitals in the European Union are legally obligated to go through tendering processes frequently to increase the efficiency of the use of public funds and to improve the competitiveness of European companies. Thus, the introduction of new THA components is likely a frequent event in European hospitals. At Coxa Hospital for Joint Replacement, new THA implants were introduced after completion of a tendering process in April 2016. These were all well-documented implants with good scientific track records.

The aim of this dissertation was to investigate complications after the introduction of well-documented components at our institution and to study the prevention of dislocation after primary THA. In Study I, the risk for dislocation was assessed between the previously used Pinnacle and the recently introduced Continuum cup systems. In Study II, the prevalence of revision for PFF and the risk for all-cause revision were compared between the previously used cemented Exeter stem and the recently introduced cemented CPT stem, which are both classified as ‘taper-slip stems’. In Study III, two prediction models employing different statistical approaches (traditional logistic regression and elastic net method that utilizes machine learning) were built with early revision for dislocation as the outcome. Studies I-III were based on Coxa’s own database. In Study IV, the survivorship of the dual-mobility cups and constrained liners used in primary THA and the effect of femoral head size on the survivorship of constrained liners were investigated from Finnish Arthroplasty Register data, with a large cohort of conventional primary THA patients as a reference group.

In Study I, a high prevalence of early dislocation (5.1%) was found with the newly introduced Continuum cup when it was used with a neutral liner. Compared with the previously used Pinnacle cup system, the risk for dislocation was higher in the Continuum cup used with a neutral liner (odds ratio 4.8, 95% confidence interval [CI] 1.4-17), but similar in the Continuum cup used with an elevated rim liner (odds ratio 1.2, CI 0.2-7.8). In Study II, PFF was the most common reason for revision, covering 35% of all revisions during the first two postoperative years. The 2-year prevalence of PFF revision was 1.6% (CI 1.0-2.4) for the CPT stem and 1.0% (CI 0.6-1.6) for the Exeter stem. The use of the CPT stem increased the overall risk for revision compared with the use of the Exeter stem (hazard ratio 1.8, CI 1.2-2.7). In Study III, the predictive capability of both prediction models was low, and the discrimination of both models was moderate. In the logistic regression model, femoral fixation method, use of antiepileptic drugs, and the primary reason for operation were the most important predictors of revision for dislocation. In Study IV, the survivorship of the dual-mobility cups and constrained liners used with a large (36 mm) diameter femoral head was comparable with or only slightly lower than conventional THA implants even though the patients who received these unconventional implants were on average severely more morbid, and therefore more vulnerable to complications. In these study groups, only a few revisions for dislocation occurred. However, the constrained liners used with smaller (22-32 mm) femoral heads had a clearly higher revision rate, mostly due to a high rate of revisions for dislocation.

The introduction of new primary THA components led to an increased risk for dislocation and a higher risk for revision in our high-volume academic joint replacement hospital. The higher risk for dislocation was explained by the weaker coverage of the recently introduced Continuum’s neutral liner compared with the previously used Pinnacle neutral liner. Therefore, the routine use of an elevated rim liner with the Continuum cup is recommended. Because the risk for PFF seems to be a concern with taper-slip stems, especially the CPT, the consideration of other stem designs for high-risk patients in primary THA is encouraged. Moreover, when new implants are introduced, the results should be carefully monitored because unexpectedly high complication rates may occur even though the implants are well- documented and have a good scientific track-record. Because dislocation is a quite rare and multifactorial event, the prediction of an early dislocation revision after primary THA is very difficult, even with a large cohort of patients and specific data available. Moreover, the use of sophisticated methods that utilize machine learning may not necessarily offer significant advantage. Our results, along with the recent literature, support the increased use of DMCs for dislocation-prone patients in primary THA. Still, more long-term results are needed before the use of these implants becomes more widespread. Also, it seems that enlarging the femoral head with CLs enhances the survival of the implant, potentially because of the decreased risk for dislocations. Therefore, it is recommended that a 36 mm femoral head is preferred when a CL is used.