In implant associated infection, bacteria can persist in biofilm on the implant surface, and sonication allows adherent organisms to be released into fluid for culture. It can recover organisms that tissue culture misses. Laboratories use different ultrasound settings, fluid volumes, equipment, processing methods and CFU thresholds, making sonication difficult to standardise.
Summary
Sonication adds the implant surface as a microbiological sampling site and can recover organisms missed by tissue culture.
Published protocols vary in ultrasound settings, fluid volume, equipment, concentration and culture methods, making results difficult to compare between studies and laboratories.
The meaning of a bacterial count depends on how the sonication fluid was processed and how the result was quantified.
Standardising the full sonication workflow would make results more reproducible and comparable between laboratories.
What sonication adds to microbiological diagnosis
The landmark 2007 prospective study included 331 hip and knee revisions, 79 of which met the study definition of PJI. Components were placed in 400 mL Ringer’s solution, vortexed, sonicated at 40 ± 2 kHz for 5 minutes, vortexed again and cultured (Trampuz et al., 2007). Sonicate fluid culture had a sensitivity of 78.5%, compared with 60.8% for periprosthetic tissue culture, while specificity was 98.8% and 99.2%, respectively. Sonicate fluid cultures were positive in 14 PJI cases in which the corresponding tissue cultures remained negative. Among patients who had received antibiotics within the previous 14 days, sensitivity was 75% for sonication and 45% for tissue culture (Trampuz et al., 2007).
In FRI, the added yield of sonication is particularly relevant in cases with suggestive criteria but no confirmatory criterion. A 2021 retrospective study of 230 retrieved fracture fixation devices included 105 such cases. Sonicate fluid culture was positive in 33, compared with 14 by tissue culture, and coagulase negative staphylococci accounted for more than half of the pathogens detected by sonication in this group (Bellova et al., 2021).
The study had no longitudinal follow up, so it was unclear how many of the additional positive sonication cultures represented infection. The number of tissue samples also varied across the cohort. Only 88 of the 230 cases had at least three tissue cultures obtained. In that subgroup, the differences in sensitivity and specificity between sonication and tissue culture were no longer statistically significant. Overall specificity was lower for sonication than for tissue culture, 73.2% versus 88.6% (Bellova et al., 2021).
Why do sonication protocols vary?
Sonication was adopted by laboratories with different equipment and microbiological workflows, and local protocols developed around the systems already in use. A 2026 scoping review of 146 clinical studies found variation across almost every stage of the procedure. Fluid volumes ranged from 50 to 800 mL, reported sonication times from 1 to 10 minutes, and frequency was omitted in 49.3% of studies. Vortexing, centrifugation and CFU thresholds also varied, and even publications from the same research groups did not always use the same protocol (Maai et al., 2026a).
Why is sonication difficult to standardise?
First, ultrasound affects both biofilm release and bacterial viability. Cavitation helps detach biofilm from the implant surface, but increasing exposure can also damage bacterial cells. In a 2024 experimental study with Staphylococcus epidermidis on titanium, 40 kHz for 5 and 10 minutes removed more biofilm than the other conditions tested, while the 5 minute condition recovered the highest number of viable bacteria. Longer or higher frequency exposure also increased the proportion of dead bacteria (Xian et al., 2024). The experiment used only Staphylococcus epidermidis, so it does not establish an optimal setting for other organisms.
Second, ultrasound exposure also depends on temperature, container properties and the sonication apparatus. In a 2009 in vitro study, all three affected bacterial recovery, with additional differences between bacterial species. Considerable variation was also observed between sonication apparatuses, and the authors recommended calibration of individual systems before clinical use (Monsen et al., 2009).
Third, the fluid volume used during sonication depends partly on the implant and the container because the implant has to be fully submerged. A 2026 scoping review of 146 clinical studies found reported volumes ranging from 50 to 800 mL (Maai et al., 2026a). A separate 2026 in vitro study tested two titanium plate sizes in different containers, using either fixed fluid volumes or the volume required for complete submersion. For the smaller plate, bacterial recovery differed between small and medium containers when a fixed volume was used, but that difference disappeared when the volume was adjusted for submersion. For the larger plate, CFU/mL was lower in the largest container, which required 270 mL, than in the smaller containers (Maai et al., 2026b). Because CFU is reported per millilitre, the same number of bacteria released from an implant will give a lower CFU/mL when suspended in a larger fluid volume. Quantitative thresholds therefore have to be interpreted together with the sonication volume used.
Microbiological processing changes the result
After sonication, the fluid is processed for culture. Laboratories may concentrate the fluid, culture different volumes, use different culture systems and apply different thresholds for a positive result.
A 2024 retrospective study at the Mayo Clinic examined 1,448 sonicate fluid cultures from hip and knee components after the laboratory had introduced a concentration step. The earlier protocol had plated 0.5 mL of unconcentrated fluid. The newer method concentrated the sonicate fluid 100 fold and plated 0.1 mL, so the CFU cut off was reassessed for the new protocol (Alvarez Otero et al., 2024).
Considering any growth positive produced a sensitivity of 75.0% and specificity of 78.2%. At a threshold of at least 20 CFU/10 mL, sensitivity was 55.3% while specificity reached 99.8% (Alvarez Otero et al., 2024). Within a defined protocol, the chosen threshold shifts the balance between sensitivity and specificity. The diagnostic performance of a CFU threshold depends on the processing method for which it was validated.
EBJIS uses more than 50 CFU/mL as confirmatory evidence for a non concentrated technique and suggests 200 CFU/mL when a concentration step is used. For other protocol variations, it recommends using cut offs validated for those methods (McNally et al., 2021).
Sonication in FRI
A 2018 systematic review found five eligible studies of sonication fluid culture in FRI. Reference standards, sampling methods and laboratory procedures varied, and the available evidence was insufficient to establish sonication as superior to tissue culture. The authors called for standardised laboratory protocols and uniform diagnostic criteria (Onsea et al., 2018).
A 2023 prospective multicentre study of femoral and tibial shaft nonunion used a defined sonication and membrane filtration protocol and identified a threshold of at least 13.6 CFU/10 mL for distinguishing septic from aseptic nonunion (Trenkwalder et al., 2023). A 2025 analysis of a subset of the same cohort, restricted to presumed aseptic nonunion, identified 11.1 CFU/10 mL for distinguishing low grade infection from aseptic nonunion (Trenkwalder et al., 2025). The 2025 paper is a further analysis of part of the 2023 cohort. These studies show that quantitative thresholds can be derived when the FRI population and laboratory method are clearly defined. Transferability requires validation of the same protocol and threshold in independent cohorts and laboratories.
What would a useful sonication standard need to define?
The protocol needs to specify each step from receipt of the explanted implant to culture interpretation. This includes transport and processing delay, sonication device, frequency and duration, container, fluid composition and volume, temperature, vortexing, concentration, cultured volume, culture conditions and quantitative interpretation.
Frequency and duration can be specified directly. Fluid volume requires a defined rule because the volume needed for immersion depends on the implant and container. That rule then needs validation.
A 2026 scoping review of 146 clinical studies found that basic method details were frequently missing, including container characteristics, fluid volume, sonication settings, processing delay and storage conditions (Maai et al., 2026a). Those details are required for another laboratory to reproduce the procedure.
Testing reproducibility between laboratories
Accredited clinical laboratories have to validate or revalidate changes to established methods. Sonication currently has no external quality assurance programme in which laboratories process the same standardised material and compare their results (Maai et al., 2026a).
Such a programme could distribute standardised samples to multiple laboratories and compare the qualitative and quantitative results obtained. Differences between laboratories would then become visible even when the same protocol is being used.
Closing Note
The available studies are too heterogeneous to compare complete protocols or identify an optimal method (Maai et al., 2026a). Sonication is useful because it samples bacteria from the implant surface. A defined laboratory workflow would make it possible to validate CFU thresholds and compare results across laboratories.
References
Alvarez Otero J, Karau MJ, Greenwood-Quaintance KE, Abdel MP, Mandrekar J, Patel R. Evaluation of Sonicate Fluid Culture Cutoff Points for Periprosthetic Joint Infection Diagnosis. Open Forum Infectious Diseases. 2024;11(5):ofae159. doi:10.1093/ofid/ofae159.
Bellova P, Knop-Hammad V, Königshausen M, Schildhauer TA, Gessmann J, Baecker H. Sonication in the diagnosis of fracture-related infections (FRI): a retrospective study on 230 retrieved implants. Journal of Orthopaedic Surgery and Research. 2021;16:310. doi:10.1186/s13018-021-02460-z.
Maai N, Frank FA, Clauss M, Sonostandard Research Group, Kuehl RA, Morgenstern M, et al. Currently applied sonication protocols for the detection of implant-associated infections in orthopaedics: a scoping review. Journal of Clinical Microbiology. 2026;64(8):e00495-26. doi:10.1128/jcm.00495-26.
Maai N, Reinert N, Kuehl RA, Moriarty TF, Clauss M, Morgenstern M, Chittò M. Sonication of orthopaedic implants: Impact of implant size, container dimensions, and sonication time on bacterial recovery. Journal of Microbiological Methods. 2026;248:107641. doi:10.1016/j.mimet.2026.107641.
McNally M, Sousa R, Wouthuyzen-Bakker M, Chen AF, Soriano A, Vogely HC, et al. The EBJIS definition of periprosthetic joint infection: a practical guide for clinicians. Bone & Joint Journal. 2021;103-B(1):18–25. doi:10.1302/0301-620X.103B1.BJJ-2020-1381.R1.
Monsen T, Lövgren E, Widerström M, Wallinder L. In Vitro Effect of Ultrasound on Bacteria and Suggested Protocol for Sonication and Diagnosis of Prosthetic Infections. Journal of Clinical Microbiology. 2009;47(8):2496–2501. doi:10.1128/JCM.02316-08.
Onsea J, Depypere M, Govaert GAM, Kuehl R, Vandendriessche T, Morgenstern M, et al. Accuracy of Tissue and Sonication Fluid Sampling for the Diagnosis of Fracture-Related Infection: A Systematic Review and Critical Appraisal. Journal of Bone and Joint Infection. 2018;3(4):173–181. doi:10.7150/jbji.27840.
Trampuz A, Piper KE, Jacobson MJ, Hanssen AD, Unni KK, Osmon DR, et al. Sonication of Removed Hip and Knee Prostheses for Diagnosis of Infection. New England Journal of Medicine. 2007;357(7):654–663. doi:10.1056/NEJMoa061588.
Trenkwalder K, Erichsen S, Weisemann F, Augat P, Militz M, von Rüden C, 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. Journal of Orthopaedics and Traumatology. 2023;24:25. doi:10.1186/s10195-023-00708-4.
Trenkwalder K, Erichsen S, Weisemann F, Augat P, SAND Research Group, Hackl S. Membrane Filtration of Sonication Fluid: A Promising Adjunctive Method for the Diagnosis of Low-Grade Infection in Presumed Aseptic Nonunion. Journal of Orthopaedic Research. 2025;43(6):1203–1211. doi:10.1002/jor.26076.
Xian C, Liu Y, Zhou L, Ding T, Chen J, Wang T, et al. Optimal ultrasonic treatment frequency and duration parameters were used to detect the pathogenic bacteria of orthopedic implant-associated infection by ultrasonic oscillation. Journal of Infection and Chemotherapy. 2024;30(12):1237–1243. doi:10.1016/j.jiac.2024.05.013.