When a hip or knee replacement fails, the failure usually comes in one of two familiar forms: the bone around the implant fractures, or bacteria colonize the prosthesis. A new analysis argues that medicine has largely overlooked the most dangerous scenario of all — the two happening at once. In a scoping review published in the Journal of Bone and Joint Infection, researchers led by Noémie Reinert and Mario Morgenstern of the University Hospital of Basel combed five major biomedical databases and distilled more than 4,000 retrieved records down to 32 studies covering 273 patients with so-called infected periprosthetic fractures. Their verdict: this hybrid complication is among the most burdensome in orthopedic surgery, and the evidence base for treating it is so thin and inconsistent that reported success rates should be read as rough sketches rather than reliable benchmarks.
The scale of the underlying problem is growing. Periprosthetic fractures are devastating complications of joint arthroplasty, and as populations age and the number of hip and knee replacements climbs worldwide, their incidence is rising in step. These injuries already carry high rates of instability, nonunion and infection, and mortality figures reach as high as 60 percent. Infection complicates between roughly 5.6 and 16.1 percent of periprosthetic fractures, a striking figure given that even isolated prosthetic joint infection — without any fracture — carries mortality of about 11 percent after infected hip replacement and 13 percent after infected knee replacement, rates the authors note are comparable to those of common cancers. When fracture and infection coexist, clinicians must somehow reconcile two treatment philosophies that pull in opposite directions.
That tension is the conceptual heart of the review. Chronic prosthetic joint infections are typically treated with a two-stage procedure: the implants are removed, an antibiotic-loaded spacer is inserted, and reimplantation follows only after a period without any hardware in the body. Fracture-related infections, by contrast, demand stable fixation, because bone cannot heal without mechanical stability. An infected periprosthetic fracture therefore forces surgeons to choose between eradicating infection through implant removal and preserving the stability the fracture needs — or to improvise some combination of both. To bring order to this therapeutic dilemma, the authors introduced a scenario framework. In the first scenario, a fracture is treated and infection follows; in the second, a fracture occurs in a joint that is already infected, whether through loosening or during the staged treatment of a known prosthetic joint infection.
The diagnostic picture the review paints is sobering. Of the 32 included studies, only eight addressed diagnosis, and their findings consistently showed that no single test can be trusted on its own. Serum inflammatory markers — C-reactive protein, erythrocyte sedimentation rate and white blood cell count — are cheap and fast but nearly useless in this setting, for two opposite, timing-dependent reasons. In the acute phase, the fracture and surgery themselves trigger a systemic inflammatory response that elevates these markers whether or not bacteria are present. One study found positive predictive values of just 29 percent for C-reactive protein, 21 percent for sedimentation rate and 18 percent for white cell count in patients with periprosthetic fractures. In chronic, low-grade biofilm infections caused by low-virulence organisms, the problem flips: markers may remain entirely normal despite active infection.
Synovial fluid analysis emerged as the most informative tool when the joint itself is involved, but even here the thresholds are unsettled. One detailed analysis found excellent sensitivity for white cell counts in the joint fluid, identifying optimal cut-offs of 2,707 cells per microliter and 77 percent polymorphonuclear cells — close to the values used for prosthetic joint infection alone. A different study, however, identified substantially higher thresholds of 4,552 cells per microliter and 79.5 percent, suggesting that fracture-related inflammation pushes the diagnostic boundary upward. Meanwhile, a normal nucleated cell count appears highly reassuring: one study found 100 percent sensitivity, meaning a normal count effectively excludes infection. Intraoperative frozen-section histopathology fared worse than many surgeons might expect. In the setting of recent fracture, trauma-induced inflammatory cell infiltration produced false-positive rates as high as 66.6 percent in one study, and a positive predictive value of only 5.3 percent in another, compared with over 80 percent in non-traumatic revisions.
The authors argue that the way out of this diagnostic fog is not a new single test but a multimodal approach that deliberately borrows from two established frameworks. Most included studies applied diagnostic criteria designed for prosthetic joint infection — chiefly the MSIS/ICM or EBJIS definitions — while the consensus definition for fracture-related infection was rarely used, likely reflecting the historical dominance of arthroplasty research. But because an infected periprosthetic fracture is a hybrid pathology, in which infection may involve the prosthetic joint, the fracture site, or both simultaneously, neither framework alone captures the full disease spectrum. The reviewers recommend applying both definitions in parallel, letting the pattern of joint involvement determine which pathway dominates, with microbiological culture of multiple deep tissue samples remaining the diagnostic cornerstone.
On the treatment side, the review found a strong and consistent preference for staged surgery. Of 24 treatment studies covering 162 cases, 83.3 percent described a two-stage approach, in which infected hardware is removed, the fracture is stabilized — often with antibiotic-laden spacers or coated plates — and reconstruction follows once the infection appears controlled. One-stage exchange appeared in only two studies with 14 cases, and debridement with implant retention was attempted in just a handful of carefully selected patients, all in the scenario where a stable implant was already in place. Even so, three case reports documented failures of debridement that had to be converted to two-stage revision. Most tellingly, 58.3 percent of treatment studies reported that additional unplanned surgeries beyond the original strategy were needed to cure the infection — a measure of how much morbidity this complication imposes even under specialist care.
The outcome numbers, while grim, come with an important caveat the authors emphasize repeatedly. Across nine studies reporting results, mean infection eradication was 76.1 percent and recurrence 22.3 percent, with eradication ranging widely from 58.3 to 91.7 percent. Both figures are slightly worse than those typically achieved for isolated prosthetic joint or fracture-related infections, which generally reach 80 to 90 percent eradication. Bone healing was reported in only four of the nine studies, covering 58 fractures, with a mean healing rate of 91.4 percent. Yet the authors insist these figures are descriptive summaries drawn from small, heterogeneous, predominantly retrospective observational studies — at least half of the treatment studies included three or fewer patients — and should not be treated as definitive benchmarks for what patients can expect.
The review also positions itself against a concurrent systematic review that reached broadly convergent conclusions, reporting eradication rates of 70 to 80 percent across 13 studies and 210 patients. The Basel and Leuven team argues its own work goes further in two respects: it is broader in scope, and it is the first to organize findings by the temporal and causal relationship between fracture and infection rather than by treatment modality alone — a distinction they contend is essential, because the two scenarios carry different biological severity and demand different surgical aggressiveness. In the scenario where infection precedes fracture, a near-universal need for staged revision and complex reconstruction emerged from the literature.
The authors’ conclusion is a call to arms rather than a treatment guide. They urge standardized diagnostic criteria that explicitly validate existing definitions in the fracture setting, uniform outcome reporting that treats fracture healing as an endpoint equal in importance to infection eradication, systematic documentation of timing and healing status, and well-designed prospective multicenter studies large enough to replace anecdote with evidence. Until then, they write, infected periprosthetic fractures should be recognized as a distinct hybrid pathology requiring the integrated application of both prosthetic joint infection and fracture-related infection principles — a worst-case scenario that, for now, medicine manages more by accumulated judgment than by data.
Subject of Research: Diagnosis, treatment and outcomes of infected periprosthetic fractures around the hip and knee
Article Title: The co-occurrence of peri-prosthetic fracture and infection: a scoping review on current diagnostic principles, treatment and outcome
Article References: Reinert, N., Bessems, L., Clauss, M., Maai, N., Onsea, J., Vles, G., Stoffel, K., Metsemakers, W.-J., & Morgenstern, M. (2026). The co-occurrence of peri-prosthetic fracture and infection: a scoping review on current diagnostic principles, treatment and outcome. Journal of Bone and Joint Infection, 11(5), 635-649. https://doi.org/10.5194/jbji-11-635-2026
Image Credits: AI Generated
Keywords: periprosthetic fracture, prosthetic joint infection, fracture-related infection, two-stage revision, scoping review, orthopedic surgery, synovial fluid analysis, inflammatory markers, antibiotic spacer, infection recurrence, arthroplasty complications, bone healing
News Source: Ophelia Keating. (October 14, 2026). When Hip and Knee Replacements Break and Infect: Surgeons Map a Worst-Case Orthopedic Crisis. Scienmag.
