Deep cultures can remain negative in fracture-related infection even after careful sampling. Sample number and location, laboratory processing, recent antibiotic exposure and bacterial distribution can all affect organism recovery, while histology and other confirmatory findings may still establish the diagnosis.
A negative culture can occur in a confirmed FRI
A negative culture can occur in a confirmed FRI. When FRI is suspected and deep tissue and sonication cultures return negative, the diagnosis depends on other findings. Under the FRI consensus definition, infection can be confirmed by a fistula, sinus or wound breakdown communicating with bone or implant, purulent drainage or the presence of pus, microbiological confirmation, or histopathological confirmation (Metsemakers et al., 2018). The FRI criteria also accept more than five polymorphonuclear neutrophils per high power field as confirmatory in chronic or late onset cases (Govaert et al., 2020; McNally et al., 2020). A patient can therefore have confirmed FRI without recovery of an organism. When cultures are negative and no other confirmatory criterion is present, there is suspicion based on suggestive criteria, without confirmation.
Culture-negative FRI is reported at very different frequencies
| Study | Population | Culture-negative FRI |
| Onsea et al., 2022 | Multicentre validation cohort | 41 cases, 8.5% |
| Vanvelk et al., 2023 | 613 FRIs in an international cohort | 47 cases, 8% |
| Corrigan et al., 2026 | 430 confirmed FRIs, all with ≥5 deep tissue samples | 109 cases, 25.3% |
Reported frequencies depend strongly on which patients are included and how FRI is defined. The 8% figure from the international cohort included all FRIs, while 5% of patients without clinical confirmatory criteria had completely negative cultures (Vanvelk et al., 2023). In the recent Oxford cohort, one quarter of confirmed FRIs were culture-negative despite at least five deep tissue specimens and a two week antibiotic-free period before sampling (Corrigan et al., 2026). Onsea et al. and Vanvelk et al. contain overlapping multicentre data, so these figures should not be read as independent prevalence estimates (Onsea et al., 2022; Vanvelk et al., 2023).
Culture has two roles in FRI diagnosis
Culture can itself confirm FRI when phenotypically indistinguishable organisms are recovered from at least two separate deep tissue or implant specimens (Metsemakers et al., 2018). In other patients, infection is already confirmed by a sinus, pus or histology. Culture then provides the microbiological diagnosis.
Identifying the organism is important for antimicrobial treatment, particularly when more than one organism is present. Sampling too few sites can recover one pathogen while missing another. In a study of 513 procedures, increasing the number of deep tissue specimens increased the probability of recovering all clinically relevant organisms (Dudareva et al., 2021). Recent sonication studies also found additional organisms that were absent from the tissue cultures and could alter antimicrobial selection (Zouitni et al., 2026; Abdo et al., 2026).
Five separate deep tissue samples improve organism recovery
Each additional deep tissue sample increases the chance of finding organisms that may be present in some parts of the infected site and absent from others. In a study of 513 procedures, the analysis estimated how often all clinically relevant organisms would still have been recovered if fewer specimens had been taken. An organism was counted as clinically relevant when it was recovered from more than one specimen (Dudareva et al., 2021).
With three samples, all clinically relevant organisms would have been recovered in 84% of procedures.
With four samples, this increased to 92%.
With five samples, it reached 97%.
These are modelled estimates from the available larger sample sets rather than three separate groups of patients. The main gain from five samples was the much lower risk of missing a clinically relevant organism. FRI recommendations therefore use at least five deep tissue samples, collected separately with clean instruments (Govaert et al., 2020).
The samples should represent the suspected infection: the fracture or nonunion, necrotic bone, deep soft tissue and implant bed where appropriate. Each specimen should be taken with a separate instrument and placed in a separate container. Sampling before extensive debridement, suction or cautery is recommended, and the anatomical origin of each specimen should be recorded for later interpretation (Morgenstern et al., 2018; Govaert et al., 2020).
Laboratory processing affects what grows
Once the tissue reaches the laboratory, recovery depends on how it is processed and cultured. In the Dudareva study, each specimen was processed within four hours, homogenised separately, inoculated into aerobic and anaerobic blood culture bottles and incubated for up to ten days (Dudareva et al., 2021). Other FRI studies use different media, processing methods and incubation periods, commonly extending culture to 14 days. Slow growing organisms may require longer incubation, while prolonged culture also increases the opportunity to recover contaminants (Morgenstern et al., 2018). Differences between laboratory protocols therefore contribute to the variation in reported culture sensitivity between studies.
Antibiotic exposure can reduce culture yield
Therapeutic antibiotics can suppress bacterial growth and reduce the chance of recovering viable organisms in culture. Current FRI recommendations therefore advise avoiding therapeutic antibiotics for at least two weeks before planned microbiological sampling where the clinical situation permits (Govaert et al., 2020). The size of this effect in FRI is difficult to quantify from the available clinical studies.
In an international cohort of 613 FRIs, 47 had no positive cultures. Nineteen of these 47 patients, or 40%, had received antibiotics during the two weeks before tissue sampling (Vanvelk et al., 2023). The study describes antibiotic exposure among the culture-negative cases rather than providing a controlled estimate of its effect on culture yield.
Culture negativity also occurs after an antibiotic-free interval. In the Oxford cohort, therapeutic antibiotics had been stopped for at least two weeks and at least five deep tissue samples were obtained from every patient. Cultures were still negative in 109 of 430 confirmed FRIs (Corrigan et al., 2026).
Bacteria are unevenly distributed through an infected fracture
Bone infection is spatial. Bacteria can be concentrated in the nonunion, necrotic bone, deep soft tissue, the implant bone interface or the implant surface, with relatively few organisms in nearby tissue (Morgenstern et al., 2018; Dudareva et al., 2021). A biopsy only contains the organisms present in the piece of tissue that was removed. Culture also needs enough viable bacteria in that specimen to grow.
This makes sampling a surgical responsibility. The surgeon chooses where tissue is taken, how many samples are collected and whether each sample remains separate. Those decisions can determine how completely the microbiology describes the infection.
Sonication adds another sampling compartment
Sonication examines an additional location: the surface of removed fixation material. Tissue culture samples pieces of bone and soft tissue, while sonication releases organisms from the implant surface into fluid for culture. The two methods often recover the same pathogens, while each can also recover organisms missed by the other.
Combining tissue and sonication culture increased sensitivity from 74% to 88% in one cohort (Zouitni et al., 2026). In another 157 confirmed FRIs, pathogen detection increased to 70.1% when both methods were used, compared with 60.5% for tissue culture and 57.3% for sonication alone (Abdo et al., 2026). The same complementary pattern was found in a prospective nonunion study (Trenkwalder et al., 2023). These studies support sonication as an additional sampling compartment when fixation material is removed.
Histology may confirm infection when cultures are negative
Histology looks for the tissue response to infection rather than bacterial growth. In 430 confirmed FRIs sampled with at least five deep tissue cultures, 321 were culture positive and 334 had positive histology. Among the 109 culture-negative cases, 69 had more than five neutrophils per high power field (Corrigan et al., 2026). Histology therefore supplied confirmatory evidence in many patients whose cultures yielded no organism.
Culture identifies viable organisms that grow under the laboratory conditions used; histology identifies the inflammatory response in the sampled tissue.
Serum inflammatory markers are suggestive criteria
CRP, ESR and leukocyte count can support suspicion of FRI, but their sensitivity is limited, particularly in chronic or low-grade infection. Elevated serum inflammatory markers are therefore classified as suggestive criteria in the FRI diagnostic definition (Govaert et al., 2020; McNally et al., 2020).
In one recent cohort, CRP was at least 10 mg/L in 12 of 42 confirmed FRIs with available measurements, ESR was at least 30 mm/h in 17 of 32, and leukocyte count exceeded 10 × 10⁹/L in 3 of 44 (Zouitni et al., 2026). These values come from one cohort, but they show how confirmed FRI can occur without a marked systemic inflammatory signal.
Implications of a negative culture: confirmed versus unconfirmed FRI
In a patient with a confirmatory criterion such as a sinus communicating with the fracture, pus at surgery or confirmatory histology, FRI is already established. Negative cultures then leave the causative organism unidentified. This limits the microbiological information available for targeted antimicrobial treatment.
The situation is different when only suggestive criteria are present. Pain, redness, swelling, nonunion, imaging abnormalities and raised inflammatory markers can increase suspicion, but none establishes FRI by itself. In an international cohort, microbiology confirmed infection in 91% of FRIs that presented without clinical confirmatory criteria; 5% of this subgroup had negative cultures (Vanvelk et al., 2023). A negative culture in this setting leaves the diagnosis without microbiological confirmation, so other confirmatory evidence becomes decisive.
Molecular methods can detect bacterial signals beyond culture
Molecular methods are included in the FRI diagnostic literature, although they sit outside the formal confirmatory criteria. The 2020 FRI diagnostic recommendations found only two small FRI studies evaluating PCR methods, so the evidence available for routine diagnostic use was limited (Govaert et al., 2020).
A 2026 proof-of-concept study used transcriptomic detection of bacterial 16S rRNA in tissue and bone from patients investigated for FRI. The assay detected an S. aureus signal in one confirmed FRI with negative conventional cultures and identified an additional P. aeruginosa signal in another polymicrobial infection (Depypere et al., 2026). The same assay also produced S. aureus signals in four patients classified as uninfected, illustrating the difficulty that comes with detecting very small bacterial signals.
Closing note
The surgeon controls several steps that determine culture yield: where tissue is taken, how many samples are collected, whether separate instruments are used and whether sampling precedes therapeutic antibiotics when feasible. Good sampling gives microbiology its best chance of identifying the pathogens. If cultures are still negative, the clinical findings and histology determine what that result means.
References
Metsemakers WJ et al. Fracture-related infection: a consensus on definition from an international expert group. Injury. 2018;49(3):505–510.
Morgenstern M et al. Diagnostic challenges and future perspectives in fracture-related infection. Injury. 2018;49(Suppl 1):S83–S90.
Govaert GAM et al. Diagnosing fracture-related infection: current concepts and recommendations. J Orthop Trauma. 2020;34(1):8–17.
McNally M et al. Definition and diagnosis of fracture-related infection. EFORT Open Rev. 2020;5(10):614–619.
Dudareva M et al. Providing an evidence base for tissue sampling and culture interpretation in suspected fracture-related infection. J Bone Joint Surg Am. 2021;103(11):977–983.
Onsea J et al. Validation of the diagnostic criteria of the consensus definition of fracture-related infection. Injury. 2022;53(6):1867–1879.
Vanvelk N et al. Diagnosis of fracture-related infection in patients without clinical confirmatory criteria: an international retrospective cohort study. J Bone Jt Infect. 2023;8(2):133–142.
Trenkwalder K et al. The value of sonication in the differential diagnosis of septic and aseptic femoral and tibial shaft nonunion in comparison to conventional tissue culture and histopathology: a prospective multicenter clinical study. J Orthop Traumatol. 2023;24(1):25.
Depypere M et al. Combined analysis of host immune response, biofilm genes, and 16S rRNA detection in fracture-related infection: an observational cohort study. J Bone Jt Infect. 2026;11(2):161–173.
Corrigan RA et al. Clinical outcome after surgery for fracture-related infection is dependent on both microbiology and the host inflammatory response. Pathogens. 2026;15(5):532.
Zouitni A et al. Sonication fluid cultures enhance pathogen identification in fracture-related-infection (FRI). J Bone Jt Infect. 2026;11(4):431–440.
Abdo B et al. Sonication versus peri-implant tissue culture in fracture-related infection: diagnostic performance and therapeutic consequences in a single-center cohort. J Bone Jt Infect. 2026;11(4):555–563.