Digital Health Applications to Establish a Remote Diagnosis of Orthopedic Knee Disorders: Scoping Review.

BACKGROUND: Knee pain is highly prevalent worldwide, and this number is expected to rise in the future. The COVID-19 outbreak, in combination with the aging population, rising health care costs, and the need to make health care more accessible worldwide, has led to an increasing demand for digital health care applications to deliver care for patients with musculoskeletal conditions. Digital health and other forms of telemedicine can add value in optimizing health care for patients and health care providers. This might reduce health care costs and make health care more accessible while maintaining a high level of quality. Although expectations are high, there is currently no overview comparing digital health applications with face-to-face contact in clinical trials to establish a primary knee diagnosis in orthopedic surgery. OBJECTIVE: This study aimed to investigate the currently available digital health and telemedicine applications to establish a primary knee diagnosis in orthopedic surgery in the general population in comparison with imaging or face-to-face contact between patients and physicians. METHODS: A scoping review was conducted using the PubMed and Embase databases according to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) statement. The inclusion criteria were studies reporting methods to determine a primary knee diagnosis in orthopedic surgery using digital health or telemedicine. On April 28 and 29, 2021, searches were conducted in PubMed (MEDLINE) and Embase. Data charting was conducted using a predefined form and included details on general study information, study population, type of application, comparator, analyses, and key findings. A risk-of-bias analysis was not deemed relevant considering the scoping review design of the study. RESULTS: After screening 5639 articles, 7 (0.12%) were included. In total, 2 categories to determine a primary diagnosis were identified: screening studies (4/7, 57%) and decision support studies (3/7, 43%). There was great heterogeneity in the included studies in algorithms used, disorders, input parameters, and outcome measurements. No more than 25 knee disorders were included in the studies. The included studies showed a relatively high sensitivity (67%-91%). The accuracy of the different studies was generally lower, with a specificity of 27% to 48% for decision support studies and 73% to 96% for screening studies. CONCLUSIONS: This scoping review shows that there are a limited number of available applications to establish a remote diagnosis of knee disorders in orthopedic surgery. To date, there is limited evidence that digital health applications can assist patients or orthopedic surgeons in establishing the primary diagnosis of knee disorders. Future research should aim to integrate multiple sources of information and a standardized study design with close collaboration among clinicians, data scientists, data managers, lawyers, and service users to create reliable and secure databases.

Developmental dysplasia of the hip in children with Down syndrome: comparison of clinical and radiological examinations in a local cohort.

Guidelines for children with Down syndrome (DS) suggest to perform an annual hip screening to enable early detection of developmental dysplasia of the hip (DDH). How to perform this screening is not described. Delayed detection can result in disabling osteoarthritis of the hip. Therefore, we determined the association between clinical history, physical, and radiological examination in diagnosing DDH in children with DS. Referral centers for children with DS were interviewed to explore variety of hip examination throughout the Netherlands. Clinical features of 96 outclinic children were retrospectively collected. Clinical history was taken, physical examination was performed, and X-ray of the hip was analyzed. All the referral centers performed physical examination and clinical history; however, 20% performed X-ray. Following physical examination according to Galeazzi test 26.9% and to limited abduction 10.8% of the outclinic-studied children were at risk for DDH. Radiological examination showed moderate or severe abnormal deviating migration rate of 14.6% resp. 11.5% in the right and left hip. However, no association between clinical history, physical examination, and radiological examination was found.Conclusion: Clinical history and physical examination are insufficient to timely detect DDH in children with Down syndrome. Thereby regular radiological examination of the hip is advised. What is Known: • Developmental dysplasia of the hip (DDH) in people with Down syndrome (DS) develops during childhood. • Guidelines for medical support of children with DS suggest an annual hip screening to enable early detection of hip damaging. How to perform this annual screening is not described. What is New: • This study shows no association between clinical history, physical and radiological examination of the hip. • We recommend regular radiological examination of the hip in children with DS in order to identify DDH early up to 16 years of age.

Optimization of the empirical antibiotic choice during the treatment of acute prosthetic joint infections: a retrospective analysis of 91 patients.

Background and purpose - The preferred treatment of an acute prosthetic joint infection (PJI) is debridement, antibiotics, irrigation and retention of the prosthesis (DAIR). The antibiotic treatment consists of an empirical and targeted phase. In the empirical phase, intravenous antibiotics are started after surgery before micro-organisms are determined in microbiological cultures. Which empirical antibiotic is used differs between hospitals, partly reflecting geographic differences in susceptibility spectrums. We investigated whether flucloxacillin should remain the antibiotic of choice in our hospital for empiric treatment of acute PJI with DAIR. Patients and methods - We retrospectively analyzed 91 patients treated for PJI with DAIR between 2012 and 2016. The susceptibility of micro-organisms was determined in multiple cultures of periprosthetic tissue and synovial fluid for 3 antibiotics: amoxicillin/clavulanic acid, cefazolin, and flucloxacillin. Results - Positive microbiological cultures from 68 patients were analyzed. Staphylococcus aureus was the predominant pathogen, cultured in half of the patients. In one-third of patients more than 1 micro-organism was found. On a patient level, the data showed that 65% were responsive to flucloxacillin, 76% to amoxicillin/clavulanic acid, and 79% to cefazolin. Interpretation - Flucloxacillin appeared to be a suboptimal choice in our patient population treated with DAIR. We therefore changed our practice to cefazolin as the preferred antibiotic in the empirical treatment of acute PJI with DAIR.

Antibiotic prophylaxis is not indicated prior to dental procedures for prevention of periprosthetic joint infections.

Background and purpose - To minimize the risk of hematogenous periprosthetic joint infection (HPJI), international and Dutch guidelines recommended antibiotic prophylaxis prior to dental procedures. Unclear definitions and contradictory recommendations in these guidelines have led to unnecessary antibiotic prescriptions. To formulate new guidelines, a joint committee of the Dutch Orthopaedic and Dental Societies conducted a systematic literature review to answer the following question: can antibiotic prophylaxis be recommended for patients (with joint prostheses) undergoing dental procedures in order to prevent dental HPJI? Methods - The Medline, Embase, and Cochrane databases were searched for randomized controlled trials (RCTs), reviews, and observational studies up to July 2015. Studies were included if they involved patients with joint implants undergoing dental procedures, and either considered HPJI as an outcome measure or described a correlation between HPJI and prophylactic antibiotics. A guideline was formulated using the GRADE method and AGREE II guidelines. Results - 9 studies were included in this systematic review. All were rated "very low quality of evidence". Additional literature was therefore consulted to address clinical questions that provide further insight into pathophysiology and risk factors. The 9 studies did not provide evidence that use of antibiotic prophylaxis reduces the incidence of dental HPJI, and the additional literature supported the conclusion that antibiotic prophylaxis should be discouraged in dental procedures. Interpretation - Prophylactic antibiotics in order to prevent dental HPJI should not be prescribed to patients with a normal or an impaired immune system function. Patients are recommended to maintain good oral hygiene and visit the dentist regularly.

Gentamicin serum concentrations in patients with gentamicin-PMMA beads for infected hip joints: a prospective observational cohort study.

BACKGROUND: Gentamicin-polymethylmethacrylate (PMMA) beads release gentamicin gradually, and high concentrations develop only locally. It is unclear how frequent and in which patients gentamicin serum concentrations are measurable and possibly lead to toxicity. The aim of this study was to investigate the measurability of gentamicin serum concentrations after the implantation of gentamicin-PMMA beads and to assess the nephrotoxicity of these beads. METHODS: In this observational cohort study, gentamicin and creatinine concentrations were measured in 34 serum samples of 23 patients with implanted gentamicin-PMMA beads for infected hip joints with our regular immunoassay (lower limit of quantitation 0.4 mg/L). Samples were also analyzed with an adjusted immunoassay with a lower limit of quantitation of 0.05 mg/L. RESULTS: Gentamicin serum concentrations were >0.4 mg/L in 9 of 34 (26%) of all the samples measured (both after the first implantation and change of beads) and in 5 of 23 patients (22%) after the first implantation of gentamicin-PMMA beads. Gentamicin serum concentrations were >0.05 mg/L in 31 samples (91%). Nephrotoxicity (defined as increase in serum creatinine >44 µmole/L and/or a relative increase >25%) occurred more frequently in patients with measurable gentamicin serum concentrations than in those without measurable gentamicin serum levels (57% versus 43%, P = 0.02). Both nephrotoxicity and gentamicin serum concentration could not be associated with the number of implanted gentamicin-PMMA beads. CONCLUSIONS: Gentamicin serum concentrations >0.4 mg/L can be measured after the implantation of gentamicin-PMMA beads in certain patients with infected hip joints. Furthermore, elevated (>0.4 mg/L) gentamicin serum concentrations are associated with nephrotoxicity in patients with gentamicin-PMMA beads for infected hip joints.

Preoperative and perioperative risk factors, and risk score development for prosthetic joint infection due to Staphylococcus aureus: a multinational matched case-control study.

OBJECTIVES: We aim to identify the preoperative and perioperative risk factors associated with post-surgical Staphylococcus aureus prosthetic joint infections (PJI) and to develop and validate risk-scoring systems, to allow a better identification of high-risk patients for more efficient targeted interventions. METHODS: We performed a multicenter matched case-control study of patients who underwent a primary hip and knee arthroplasty from 2014 to 2016. Two multivariable models by logistic regression were performed, one for the preoperative and one for perioperative variables; predictive scores also were developed and validated in an external cohort. RESULTS: In total, 130 cases and 386 controls were included. The variables independently associated with S. aureus-PJI in the preoperative period were (adjusted OR; 95% CI): body mass index >30 kg/m2 (3.0; 1.9 to 4.8), resident in a long-term care facility (2.8; 1.05 to 7.5), fracture as reason for arthroplasty (2.7; 1.4 to 5.03), skin disorders (2.5; 0.9 to 7.04), previous surgery in the index joint (2.4; 1.3 to 4.4), male sex (1.9; 1.2 to 2.9) and American Society of Anesthesiologists index score 3 to 4 (1.8; 1.2 to 2.9). The area under the receiver operating characteristic curve was 0.73 (95% CI 0.68 to 0.78). In perioperative model, the risk factors were the previous ones plus surgical antibiotic prophylaxis administered out of the first 60 minutes before incision (5.9; 2.1 to 16.2), wound drainage for >72 hours after arthroplasty (4.5; 1.9 to 19.4) and use of metal bearing material versus ceramic (1.9; 1.1 to 3.3). The area under the receiver operating characteristic curve was 0.78 (95% CI 0.72 to 0.83). The predictive scores developed were validated in the external cohort. DISCUSSION: Predictive scores for S. aureus-PJI were developed and validated; this information would be useful for implementation of specific preventive measures.

Rates and Predictors of Treatment Failure in Staphylococcus aureus Prosthetic Joint Infections According to Different Management Strategies: A Multinational Cohort Study-The ARTHR-IS Study Group.

INTRODUCTION: Guidelines have improved the management of prosthetic joint infections (PJI). However, it is necessary to reassess the incidence and risk factors for treatment failure (TF) of Staphylococcus aureus PJI (SA-PJI) including functional loss, which has so far been neglected as an outcome. METHODS: A retrospective cohort study of SA-PJI was performed in 19 European hospitals between 2014 and 2016. The outcome variable was TF, including related mortality, clinical failure and functional loss both after the initial surgical procedure and after all procedures at 18 months. Predictors of TF were identified by logistic regression. Landmark analysis was used to avoid immortal time bias with rifampicin when debridement, antibiotics and implant retention (DAIR) was performed. RESULTS: One hundred twenty cases of SA-PJI were included. TF rates after the first and all surgical procedures performed were 32.8% and 24.2%, respectively. After all procedures, functional loss was 6.0% for DAIR and 17.2% for prosthesis removal. Variables independently associated with TF for the first procedure were Charlson ≥ 2, haemoglobin < 10 g/dL, bacteraemia, polymicrobial infection and additional debridement(s). For DAIR, TF was also associated with a body mass index (BMI) > 30 kg/m2 and delay of DAIR, while rifampicin use was protective. For all procedures, the variables associated with TF were haemoglobin < 10 g/dL, hip fracture and additional joint surgery not related to persistent infection. CONCLUSIONS: TF remains common in SA-PJI. Functional loss accounted for a substantial proportion of treatment failures, particularly after prosthesis removal. Use of rifampicin after DAIR was associated with a protective effect. Among the risk factors identified, anaemia and obesity have not frequently been reported in previous studies. TRIAL REGISTRATION: This study is registered at clinicaltrials.gov, registration no. NCT03826108.

Staphylococcus aureus is one of the most virulent bacteria and frequently causes prosthetic joint infections.Knowledge of the treatment of this type of infection has advanced in recent years, and treatment guidelines have led to improved management. Typically, the successful treatment of these infections has been determined by clinical cure, that is, the symptoms of infection have disappeared, but has not taken into account loss of function (such as significant difficulties walking), which is critical for the patient's quality of life. Our aim in this study was to evaluate the success of current management strategies for S. aureus prosthetic joint infection, including recovery of functionality, and the factors that predict why some of these infections are not cured, to identify areas for improvement.In a multinational cohort of 128 patients with S. aureus prosthetic joint infection, rates of treatment failure were found to be high, with significant rates of loss of function, especially when the prosthesis needed to be removed. Loss of function was less frequent when the infection was initially treated with surgical cleaning without removal of the prosthesis, even when this procedure failed at first. We found that anaemia and obesity were associated with lower treatment success, and that the probability of treatment success increased when surgical cleaning without prosthesis removal was performed early, and when the antibiotic rifampicin was used in combination with another antibiotic.

Effectiveness and cost-effectiveness of dynamic bracing versus standard care alone in patients suffering from osteoporotic vertebral compression fractures: protocol for a multicentre, two-armed, parallel-group randomised controlled trial with 12 months of follow-up.

INTRODUCTION: Patients with osteoporosis may suffer from a fracture after minimal trauma. Osteoporotic vertebral compression fractures (OVCFs) are among the most common fractures, often leading to substantial pain. There is a need for evidence-based conservative treatment to aid in the management of OVCFs. The objective of this randomised controlled trial (RCT) is to evaluate the effectiveness and cost-effectiveness of dynamic bracing in addition to standard care for improving quality of life (QoL) in patients suffering from an OVCF. METHODS AND ANALYSIS: Ninety-eight postmenopausal women from two academic and four community hospitals with a recent symptomatic thoracolumbar OVCF will be randomised into either the standard care or dynamic bracing group. In the dynamic bracing group, the Spinova Osteo orthosis will be used in addition to standard care. Standard care comprises pain control with analgesics, physical therapy and osteoporosis medication. The primary outcome parameter is QoL 1 year after inclusion, as measured by the Quality of Life Questionnaire of the European Foundation for Osteoporosis (QUALEFFO-41). Secondary outcome parameters are pain, pain medication used, functional disability, sagittal spinal alignment, recurrence rate of OVCFs and physical activity in daily life. A trial-based economic evaluation consisting of both cost-effectiveness analysis and cost-utility analysis will be performed based on empirical data obtained in the RCT. A process evaluation will assess the feasibility of dynamic bracing. All outcomes will be assessed at baseline, 6 weeks, 3 months, 6 months, 9 months and 12 months. ETHICS AND DISSEMINATION: Ethical approval has been granted by the Medical Ethics Committee, University Hospital Maastricht and Maastricht University (METC azM/UM) (NL74552.068.20/METC 20-055). Patients will be included only after verification of eligibility and obtaining written informed consent. Results will be disseminated via the Dutch National Osteoporosis Patient Society and via publications and conferences. TRIAL REGISTRATION NUMBER: NL8746.

Antimicrobial efficacy of gentamicin-loaded acrylic bone cements with fusidic acid or clindamycin added.

The increasing gentamicin resistance among bacteria in septic joint arthroplasty has stimulated interest in adding a second antibiotic into gentamicin-loaded bone cement. A first aim of this in vitro study is to investigate whether addition of fusidic acid or clindamycin to gentamicin-loaded bone cement has an additional antimicrobial effect against a collection of 38 clinical isolates, including 16 gentamicin-resistant strains. A modified Kirby-Bauer test, involving measurement of the inhibition zone around antibiotic-loaded bone cement discs on agar plates, was used to investigate whether adding a second antibiotic has an additional antimicrobial effect. Second, a selected number of strains was used to study their survival in an interfacial gap made in the different bone cements to mimic the gap between bone and cement as existing near a prosthesis. Gentamicin-loaded bone cement had an antimicrobial activity against 58% of the 38 bacterial strains included in this study, while 68% of the strains were affected by bone cement loaded with a combination of gentamicin and clindamycin. Bone cement loaded with the combination of gentamicin and fusidic acid had antimicrobial activity against 87% of the bacterial strains. In the prosthesis-related gap model, there was a clear trend toward less bacterial survival for gentamicin-loaded bone cement after adding clindamycin or fusidic acid. Addition of clindamycin or fusidic acid into gentamicin-loaded bone cement yields an additional antimicrobial effect. The combination gentamicin and fusidic acid was effective against a higher number of clinical isolates than the combination of gentamicin with clindamycin, including gentamicin-resistant strains.

Outcomes in chevron osteotomy for Hallux Valgus in a large cohort.

Clinical and radiological related outcomes have been reported for Chevron osteotomy as correction for mild to moderate hallux valgus, but only for relatively small patient series. Moreover, evaluation of the patient's point of view has mostly been conducted by means of more physician-based outcome measures. The goal of this study was to evaluate the effect of the Chevron osteotomy for hallux valgus on patients' daily lives using the Foot and Ankle Outcome Score (FAOS) as a validated and a hallux valgus specific patient reported outcome measure (PROM). Secondary outcome measures were radiological correction, complication rate, and re-operations. All 438 Chevron procedures (336 patients), at two surgical hospital sites in the period between January 2010 and October 2014, were retrospectively evaluated with a follow-up of at least 6 months. Patients were invited to fill in a cross-sectional online FAOS. For the FAOS, a total response of 60% was achieved. The FAOS ranged between 71 and 88 with a follow-up of on average 36 months. Patients with an undercorrection of their hallux valgus (11.6% of the procedures) scored significantly lower on three subscales of the FAOS (range between 61 and 77 versus 72-84). Patients who had a reoperation (12.6% of the procedures) also scored significantly lower on four subscales: 58-100 versus 73-89. Postoperative radiological measurements improved significantly with a mean difference of 6.1 (5.9; 6.4) degrees for the intermetatarsal angle and 13.7 (13.0; 14.5) degrees for the hallux valgus angle. In this large study cohort, Chevron osteotomy for hallux valgus offers good PROM scores on FAOS. These scores were significantly lower in patients with radiological undercorrection or with a reoperation. Results of the FAOS appear to modulate with physician based outcomes and therapeutic incidents. Improvement of outcome may therefore well be possible by increased attention on these surgical details.

The influence of cyclic loading on gentamicin release from acrylic bone cements.

Antibiotic-loaded acrylic bone cement is widely used in total joint replacement to reduce infections. Walking results in cyclic loading, which has been suggested to stimulate antibiotic release. The goal of this study is to compare antibiotic release from cyclically loaded bone cement with the release from unloaded bone cement. Two models of the frontal aspect of a femoral stem were cemented with CMW 1 Radiopaque G, Palacos R-G and Palamed G. Both were immersed in water, and the gentamicin concentration in the water was monitored. One model was cyclically loaded at 5 Hz during immersion achieving physiological stresses in the bone cement mantle. After 10.8 x 10(6) cycles, initial release of gentamicin from Palamed G was increased significantly for loaded over unloaded, but not from CMW 1 Radiopaque G and Palacos R-G.

The influence of ultrasound on the release of gentamicin from antibiotic-loaded acrylic beads and bone cements.

Gentamicin-loaded acrylic beads are loosely placed in infected bone cavities, whereas gentamicin-loaded acrylic bone cement is used as a mechanical filler in bone to anchor prosthetic components. Both drug delivery systems are used to decrease infection rates by gentamicin release. The objective of this study is to investigate the effects of pulsed ultrasound on gentamicin release from both materials. Gentamicin release from gentamicin-loaded beads (Septopal) and from three commercially-available brands of gentamicin-loaded bone cement (CMW 1, Palacos R-G, and Palamed G) was measured after 18 h of exposure in PBS to an ultrasonic field of 46.5 kHz in a 1:3 duty cycle with an average acoustic intensity of 167 mW/cm(2). Samples not exposed to ultrasound were used as controls. Pulsed ultrasound significantly enhanced gentamicin release from gentamicin-loaded beads, whereas gentamicin release from the gentamicin-loaded bone cements was not significantly enhanced. Mercury intrusion porosimetry revealed an increased distribution of pores between 0.1 and 0.01 microm in beads after gentamicin release, while in bone cements no increase in the number of pores was found. Increased gentamicin release in beads due to ultrasound may be explained by micro-streaming in a porous structure, whereas the absence of changes in pore structure after gentamicin release in bone cement is concurrent with the lack of an enhanced release of the antibiotic by ultrasound. As an effective treatment of infections requires high local concentrations of antibiotic, increased gentamicin release due to ultrasound may be of clinical significance, especially since ultrasound has been demonstrated to increase bacterial killing by antibiotics.

Increased release of gentamicin from acrylic bone cements under influence of low-frequency ultrasound.

The release profile of antibiotics from antibiotic-loaded bone cement, used to prevent infections in total joint arthroplasty, is neither ideal nor complete. Ultrasound has been used to allow drugs to cross otherwise impermeable barriers. The aim of this study was to establish a possible effect of ultrasound on antibiotic release from bone cements. Samples were made of three commercially available gentamicin-loaded bone cements. Part of the samples was allowed to release gentamicin for 3 weeks before insonation. An insonation device produced an ultrasound field with a time average acoustic intensity of 167 mW/cm2 at a frequency of 46.5 kHz. The samples were exposed to the ultrasound field or not exposed to it as a control. The amount of gentamicin released was measured by fluorescence polarization immunoassay. There was a limited increase of gentamicin release with application of ultrasound in fresh samples but not in the samples that had been allowed to release gentamicin. For fresh samples, a linear regression model showed that this ultrasound effect was statistically significant. The mechanism behind these observations is not clear, but it is suggested that microstreaming or localized temperature rises may be involved.

The release of gentamicin from acrylic bone cements in a simulated prosthesis-related interfacial gap.

Gentamicin is added to polymethylmethacrylate bone cements in orthopedics as a measure against infection in total joint arthroplasties. Numerous studies have been published on gentamicin release from bone cements, but none have been able to estimate the local concentrations in the prosthesis-related interfacial gap, present after implantation. The aim of this study was to develop a method allowing determination of antibiotic release in such a gap. Two-hundred-micrometer-wide gaps with a volume of 6 microl and a surface area of 0.6 cm2 were created by inserting stainless-steel strips in gentamicin-loaded bone cement plugs prior to polymerization. After hardening, the gap surface was exposed to 6 microl or 10 ml of phosphate-buffered saline. Within 2 h, gentamicin concentrations in the gaps reached around 4000 microg/ml for 4 different CMW and Palamed cements and 2500 microg/ml for Palacos R. Concentrations measured in the larger volume were several hundred times lower than in the gaps. This simulated prosthesis-related interfacial gap model offers new insights in the clinical efficacy of antibiotic-loaded bone cements. It is demonstrated that concentrations up to 1000-fold the antibiotic resistance levels for most bacterial strains causing implant infection can be achieved in a realistic in vitro model.

Pseudomonas aeruginosa biofilm formation and slime excretion on antibiotic-loaded bone cement.

BACKGROUND: Infection is an infrequent but serious complication of prosthetic joint surgery. These infections will usually not clear until the implant is removed and re-implantation has a high failure rate, especially when Pseudomonas aeruginosa is involved. MATERIAL AND METHODS: We examined Pseudomonas aeruginosa biofilm formation on plain and gentamicin-loaded bone cement with confocal scanning laser microscopy (CSLM). Two different stains were applied in order to visualize and quantify the distribution of bacterial cells and extracellular polymeric substances (slime) from the bone cement surface to the top of the biofilm. Staining with LIVE/DEAD viability stain differentiated between live and dead bacteria within the biofilm, and slime production was evaluated after staining with Calcofluor white. RESULTS: CSLM showed that the biofilm was a nonuniform structure of variable thickness, with differences in local bacterial cell and slime densities. Incorporation of gentamicin in bone cement resulted in a 44% reduction in bacterial viability, while the slime density increased significantly. In addition, conventional plate counting showed the development of small-colony variants on gentamicin-loaded bone cement with a decreased sensitivity for gentamicin (MIC: 8 m/L), as compared with normal-sized colonies taken from plain and gentamicin-loaded bone cement (MIC: 3 m/L). The enhanced slime production on antibiotic-loaded bone cement, together with the formation of small-colony variants, resulted in decreased susceptibility to antibiotics--probably concomitant with the onset of persistent and relapsing infections. INTERPRETATION: In the clinical situation, our findings help to explain the frequent re-implantation failure of joint replacements infected with P. aeruginosa when the procedure has been performed using antibiotic-loaded bone cement.