That challenge is central to PJI research, and it is directly reflected in the programme of the upcoming Joint EBJIS–MSIS 2026 Conference in Porto, which features dedicated sessions on evolving definitions and classification frameworks. The field is no longer asking "how do we treat PJI?" but increasingly, "are we reliably identifying it in the first place?"

The Infection That Does Not Announce Itself

A significant portion of PJI cases do not present with the typical clinical features clinicians recognise: fever, wound drainage, sinus tract formation, or sharply elevated inflammatory markers. Instead, many patients present with pain, mechanical symptoms, or radiological loosening - findings that point, logically enough, towards aseptic failure.

The problem is that "aseptic" may not always mean sterile. Low-virulence organisms -particularly coagulase-negative staphylococci and Cutibacterium acnes - can colonise implant surfaces within a protective biofilm that shields bacteria from immune surveillance, conventional antibiotics, and standard microbiological culture. The result is a subclinical infective process that is functionally invisible to routine preoperative work-up.

The prevalence data is significant. Across multiple studies, occult infection has been identified in approximately 7 to 15 percent of revision arthroplasties performed for presumed aseptic loosening.⁻ In one prospective cohort, 12.1 percent of presumed aseptic revisions were found to have PJI on enhanced postoperative analysis, with implant failure occurring in 37.5 percent of those missed cases compared with just 1.1 percent of true aseptic cases. Preoperative inflammatory markers did not reliably distinguish infected from non-infected cases, reinforcing the limitations of conventional work-up.

Criteria That Work Better in Research Than in Practice

The response to diagnostic inconsistency has been the development of multicriteria scoring systems, produced by authoritative bodies including MSIS, EBJIS, ICM, and IDSA over the past decade. These frameworks incorporate serology, synovial fluid analysis, histology, culture, and modern biomarkers into weighted diagnostic scores.

In research settings, where algorithms are applied rigorously, these systems performed well. In routine clinical practise, the picture is less reassuring. A study presenting 277 clinical vignettes to 12 experienced physicians across academic arthroplasty surgeons, community surgeons, and infectious disease specialists, found a mean undecided rate of 23.5 percent and poor interobserver agreement across all groups, with kappa values ranging from 0.49 to 0.63. Mean physician accuracy was 90.8 percent, which sounds reasonable until it is noted that two individual tests - the alpha-defensin laboratory assay and the alpha-defensin lateral-flow test - outperformed the physicians using multiple criteria combined.

The complexity of the scoring systems appears to be part of the problem. Competing definitions, evolving thresholds, and the cognitive burden of reconciling multiple imperfect results in a single clinical encounter create conditions for diagnostic variability that no guideline has fully resolved.

The Grey Zone Is Not Neutral Ground

Even when clinicians apply structured criteria carefully, a meaningful proportion of cases fall into an equivocal space that no framework classifies cleanly. These grey-zone patients carry real clinical risk.

Comparative studies of the EBJIS 2021 and ICM 2018 definitions found inconclusive rates of 7.8 to 14.6 percent under ICM and 4.7 to 8 percent under EBJIS, a difference that reached statistical significance.⁻More importantly, approximately 45 to 55 percent of preoperative inconclusive cases were later confirmed as infected once intraoperative data became available.⁻

Outcomes in this group are especially concerning. Patients who fell into the equivocal category and were managed as aseptic fared worse than those correctly identified as uninfected: in one large multi-centre validation study, the EBJIS "infection likely" group had a subsequent PJI rate of 23 percent, exceeding even the confirmed infection group at 17 percent. The grey zone, in other words, is not a benign diagnostic limbo. It is an area of concentrated clinical risk.

What This Should Change

The combined message from occult infection research, physician behaviour studies, and grey-zone outcome data is consistent: PJI is underdiagnosed not through negligence but because the biology can be subtle, the available diagnostic tools have recognised limitations, and the decision frameworks can be difficult to apply with precision outside controlled research environments.

Several implications follow. Routine intraoperative sampling in all revision arthroplasties, not only those with clinical suspicion, should be standard practice. A multimodal approach - multiple tissue cultures, histopathology, and sonication of explanted components -improves detection in ways that no single test can replicate.⁻ Equivocal preoperative results should prompt further investigation rather than default to an aseptic assumption.

More broadly, the field may benefit from diagnostic frameworks that are both rigorous and practical enough to be applied consistently by surgeons outside specialist centres. That balance has not yet been fully achieved - and its pursuit will remain one of the most important challenges in orthopaedic infection management.