Design: Single-institution retrospective cohort with Kaplan-Meier survival analysis
Population: 298 patients (302 hips) undergoing two-stage exchange arthroplasty for periprosthetic joint infection, 2005-2022, mean follow-up 6 years
Primary outcome: Five-year survival free of reinfection, re-revision and reoperation after reimplantation
Effect: 89% reinfection-free at 5 years (95% CI 85.2-93.1); spacer dislocation associated with dislocation after reimplantation HR 6.2 (95% CI 2.3-16.9, P<0.001)

This series covered 298 patients and 302 hips who had two-stage exchange arthroplasty for chronic periprosthetic joint infection using a commercially available, partially intraoperatively fabricated articulating high-dose antibiotic spacer, between August 2005 and May 2022. Mean age 65, mean BMI 32, mean follow-up six years; 40% had already had a PJI-related procedure and 88% met the 2011 MSIS criteria.

The infection control results are reassuring. At five years after reimplantation, survival free of reinfection was 89% (95% CI 85.2-93.1), free of aseptic re-revision 93% (89.3-96.1), free of any re-revision 86% (81.3-90.3) and free of any reoperation 80% (74.7-84.9). McPherson host C (HR 3.7, 95% CI 1.4-9.6) and limb grade 3 (HR 3.3, 1.0-11.1) predicted reinfection.

The finding that should change something is elsewhere. Dislocation, not infection, caused most aseptic re-revisions — 15 of 20. And dislocation of the spacer during the interval was strongly associated with dislocation after reimplantation: hazard ratio 6.2 (95% CI 2.3-16.9, P<0.001). Spacer fracture and spacer dislocation each occurred in about 5% of patients.

That gives a concrete decision rule. A hip whose spacer dislocated is a hip that should be reimplanted with a higher-stability bearing — dual mobility or constrained — rather than a standard bearing chosen on general principles. The event that identifies the risk has already happened, in front of you, before the definitive operation.

Use a higher-stability bearing at reimplantation if the spacer dislocated
Dislocation, not reinfection, is what drives aseptic re-revision in this population
McPherson host C and limb grade 3 are the reinfection predictors worth recording preoperatively
Perispacer fracture in about 5% — most were managed without an operation
89% reinfection-free survival at five years is the figure to quote when consenting for a two-stage exchange

Why it matters

It turns an event already observed during the spacer interval into a bearing decision at the second stage.

The statistics, in plain English

A hazard ratio of 6.2 with an interval from 2.3 to 16.9 is a strong signal with imprecise magnitude — the direction is secure, the size is not. This was a single-institution series using one specific spacer, so absolute rates may not transfer, but the association between spacer dislocation and later dislocation is a mechanical observation that should. The limb grade 3 hazard ratio, with a lower bound of 1.0 and P=0.049, is borderline and should not be leaned on.