The incidence of prosthetic joint infections (PJIs) has shown increasing trend worldwide which can be attributed to several factors including higher number of arthroplasties and the ageing population [1]. Staphylococcus aureus and coagulase-negative Staphylococci (CNS) are the most frequent causative agents, however, further Gram-positive and Gram-negative bacteria as well as fungal pathogens can also be involved, potentially resulting in polymicrobial infections [2, 3].

Both Gram-positive and Gram-negative microorganisms have the ability to form biofilms on the surface of prosthetic materials, particularly Staphylococci and Pseudomonas aeruginosa [4, 5]. This should always be considered in the treatment of PJIs, hence rifampicin is frequently used with other antibiotics in such cases. Combination is essential as monotherapy can rapidly lead to rifampicin resistance [6, 7]. However, resistance may develop even when rifampicin is used in combination, especially in the presence of certain predisposing factors including postoperative drainage, fistula, open wound and abscess [8]. The aim of the treatment is to eradicate pathogens, provide bactericidal as well as anti-biofilm activity and reduce the risk of resistance development [9].

Difficult-to-treat (DTT) pathogens have the potential to develop resistance against antibiotics and produce biofilm resulting in difficult eradication. Microbiological diagnosis can also be challenging for some species, eg. Cutibacterium acnes and Finegoldia spp. A study found that as high as 58% of the patients had PJI caused by DTT organisms with less favourable prognosis and the need of prolonged treatment [3].

The purpose of this study was to investigate the effect of rifampicin resistance and patient-related factors on recovery rates in patients with PJI undergoing DAIR procedure (Debridement, Antibiotics and Implant Retention). Co-morbidities and other medical conditions, orthopaedic factors including revision before DAIR procedure, mobile element exchange, preoperative score systems as well as antibiotic treatment regimes were also reviewed.

DAIR procedure is most efficient in the early postoperative period (< 6 weeks) and in acute haematogenous infections. The nonmobile elements of the prosthesis are left in situ during surgery. Debridement includes exploration of the joint, removal of necrotic and infected tissues, the use of high volume washing fluid and the exchange of mobile components. The surgical procedure is followed by prolonged antibiotic treatment [13]. The aim of the procedure is to save the prosthesis by eradicating the infection and removing the biofilm. It is recommended to perform open surgery as arthroscopic synovectomy is unsatisfactory [14, 15]. However, DAIR procedure cannot be used for the treatment of late PJIs as compact and mature biofilm can only eradicated by the exchange of prosthesis [16]. There are different recommendations regarding antibiotic route and duration, however, all regimes include prolonged intravenous treatment followed by oral antibiotics. The duration of intravenous antibiotic therapy ranges from 2 to 6 weeks and the total duration of treatment is 12 weeks in most of the cases, however it can be as long as 24 weeks if knee joint is involved. Moreover, treatment regime is also determined by the type and outcome of surgical procedure(s). Rifampicin has essential role in the management of infections with biofilm formation [17]. The dose of rifampicin is 300–600 mg twice daily per os. Liver function tests should be reviewed before starting treatment and followed up on a regular basis. Also, drug interactions must be carefully assessed when antibiotic therapy is being planned.

Postoperative wound discharge can be a sign as well as a predisposing factor of PJI. Ongoing discharge may act as a basis of infection which can spread to deeper tissues and reach the surface of implant [18]. Infections caused by high-virulence organisms (eg. S. aureus and beta-haemolytic Streptococcus spp.) can develop after 7–10 days. It is essential to recognise early PJIs as soon as possible because biofilm is unmature at this stage and therefore it can be eradicated more effectively [16, 17, 19]. Low-grade infections develop six weeks to two years after surgery and are generally caused by low-virulence organisms, eg. S. epidermidis and C. acnes. Most of the PJIs are recognised in this period. Biofilms are mature at this stage therefore implants cannot be saved, ie. all components must be carefully removed [19, 20]. Acute haematogenous infections may develop any time after surgery [18]. Symptoms develop suddenly without prodromal sings resulting in decreased mobility. Pathogens spread from a different source of infection and reach the implant with the bloodstream. Most frequent sources are skin and soft tissue infections, urinary tract infections, respiratory and gastrointestinal foci [21, 22].

The aim of our study was to determine recovery rates in patients with PJI undergoing DAIR procedure and to investigate the effect of rifampicin resistance and selected patient-related factors on clinical outcome. The overall recovery rate in the rifampicin-sensitive group was 72.3%, whereas it was 76.9% if rifampicin-resistant organism was isolated. This is unexpected and may be due to the low number of resistant isolates in the study population. Various recovery rates have been reported in the literature ranging from 54.2 to 91%, respectively [23–25]. In a previous study, we found recovery rate of 92.5% in the rifampicin-sensitive and 60.0% in the resistant group among patients undergoing two-stage revision due to PJI [26]. This may suggest that rifampicin-resistance has higher impact on recovery rates among patients with two-stage revision as compared to DAIR procedure, however, further investigations are required to confirm these findings.

The prevalence of prosthetic joint infections after knee arthroplasty was found higher among males in a study but the difference was not significant [27]. In our cohort, 59.6% were males in the sensitive and 61.5% in the resistant group. The recovery rate was 72.3% for males and 76.7% for females. We found no statistical evidence of the negative impact of sex on recovery. The vast majority of our patients were in the ≥ 60 years age group: the overall mean age was 68.4 years (standard deviation (SD) = 15.8 years). No relation between age and the prevalence of PJI was found in previous studies [27, 28]. In our study, we may assume that age has a presumable negative impact on recovery rate both in the sensitive and the resistant group.

It has been shown that higher BMI results in higher risk of PJIs [29]. In our cohort, there was only a small difference of BMI between the sensitive and resistant group: 30.1 kg/m² and 30.6 kg/m², respectively. Neither clinically relevant nor statistically significant effect of BMI on recovery rates was observed in either the sensitive or the resistant group. Diabetes mellitus has been confirmed as a significant predisposing factor for PJIs [16, 27, 28]. The prevalence of DM was found to be 5.0% in a study on patients with THR and was associated with higher rates of both surgical and non-surgical site infections [30]. In our cohort, we found higher prevalence: 16.4% of the patients had Type 1 or Type 2 DM. Despite this, there was no significant difference in the recovery rate of diabetic and non-diabetic population, although analysis was limited due to the low number of patients in the rifampicin resistant group. However, our data may suggest a negative impact of diabetes on clinical outcome.

Several medical conditions have been demonstrated to increase the risk of development of PJI, hence it is important to consider them in the management of PJIs. Factors increasing the risk of PJI after implanting primary endoprosthesis include diabetes mellitus, urinary tract infection (UTI), high ASA score and immunosuppression [31, 32]. Another study found that obesity, COPD, excessive ethanol consumption, depression and malignancies can also be predisposing factors [33]. Uncontrolled DM, severe obesity (BMI > 40 kg/m²), liver failure, renal insufficiency, smoking, drug abuse, previous prolonged hospitalisation, malnutrition, severe acquired immunodeficiency, posttraumatic arthrosis and inflammatory arthropathy also represent risk factors for periprosthetic infections [34]. The prevalence of PJI was also shown higher among patients receiving intraarticular steroid injection [35].

63.8% of the patients had hypertension in the rifampicin-sensitive and 69.2% in the resistant group. From a different perspective, 62.0% of the patients who recovered had hypertension whereas the prevalence was 82.4% among patients with treatment failure. The prevalence was 58.8% and 60.0% in recovered patients in the sensitive and resistant group, however, we found 76.9% and 100% of prevalence in patients who did not recover. Of note, there were only three patients in the last group. Although a previous study demonstrated that the effect of hypertension is statistically not significant [36], we found 1.3 times lower recovery rates among patients with hypertension. It has been observed that PJIs develop more frequently in patients belonging to ASA group III and IV [37]. In the rifampicin sensitive group, we found 70.9% recovery in patients with ASA II score and 80.0% with ASA III. Recovery rates for ASA II and III patients in the resistant group were 80.0% and 50.0%, respectively. We could not confirm the negative impact of higher ASA scores on recovery rates in the two groups.

Risk scores, such as KLIC and CRIME80 are used to assess the probability of therapeutic failure in patients with PJI [38]. Recovery rates were compared among patients with CRIME80 score ≥ 3 (19 cases). Average score of recovered patients was found higher in the resistant group. Nineteen patients had higher than 4.5 KLIC score. There was no significant different between the average score of recovered patients in the sensitive and the resistant group. 92.31% of the patients had KLIC score in the 2–3.5 range in the resistant group. Although the same range was predominant in the sensitive group, distribution appeared more balanced.

It has previously been demonstrated that the injury of knee or hip joint capsule, previous surgery and trauma significantly increase the risk of development of PJI [39]. We examined whether revision prior to DAIR had impact on recovery rates. Patients who had previous revision showed 66.7%recovery in the rifampicin sensitive and 80.0% in the resistant group, whereas the rates were 74.3% and 75.0% without revision. From a different approach, patients without prior revision had slightly higher recovery rates in the sensitive group, however, recovery rates were moderately higher among patients with previous revision in the resistant group.

Pathogenic spectrum has been changing worldwide. Moreover, there is a general increase in antibiotic resistance rates [1, 40] and in the incidence of polymicrobial infections [41]. The basis of the treatment of PJIs is the removal of biofilm, however, this is only achievable in the first 3–4 weeks of infection. After this period, the implant may have to be removed resulting in a significantly prolonged treatment and healing [42]. The biofilm grows continuously on the implant surface before becoming fully mature after 4–6 weeks. Postoperative exploration, effective washing, exchange of mobile components and retaining fix elements can be satisfactory for unmature biofilms. However, for mature biofilms, the procedure must include the complete removal of the implant [4]. Also, diffusion rate of rifampicin depends on the organism as well as the age of biofilm.

In our study, Staphylococcus spp. was predominant in the rifampicin-sensitive group (18–18 S. aureus and CNS isolates) followed by S. agalactiae, S. dysgalactiae and 6 C. acnes, however, Enterobacterales (mostly E. coli), P. aeruginosa and E. faecalis were the most frequent isolates in the resistant group. We observed no development of rifampicin resistance in our cohort and none of the patients had received previous rifampicin treatment. A study in 2017 found that coagulase-negative Staphylococci are the most frequent causative agents of PJIs (30–43%) followed by S. aureus (12–23%). Streptococci are the second most frequent pathogens causing PJIs, particularly S. agalactiae, S. pyogenes and S. dysgalactiae. Brucella and Candida species have also been detected, however, no pathogen has been identified in 11% of the cases [43]. Streptococcal PJIs mostly develop as a result of haematogenic infection. A study found that 25% of the specimens grew Streptococcus sp. in relation to PJIs and blood cultures were also positive in 22% of the cases. 22% of the infections were polymicrobial, mostly involving S. aureus [44].

PJIs caused by S. aureus are characterised by strong inflammatory symptoms soon after implantation indicating early postoperative infection. However, coagulase-negative Staphylococci generally cause late and potentially chronic infection with mature biofilm and less characteristic symptoms. MRSA infection itself is a risk factor resulting in more difficult eradication of PJIs. One study concluded that in case of PJIs caused by low-virulence organisms (eg. coagulase-negative Staphylococci, Cutibaterium and Acinetobacter species) DAIR procedure can be appropriate and result in full recovery, however, this finding should be treated with caution. The incidence of PJIs caused by CNS is worldwide increasing [45]. In our cohort, 33.3% of the patients in the sensitive group had infection caused by coagulase-negative Staphylococci, higher than indicated in the literature, however, no infections due to CNS was found in the resistant group. The prevalence of Enterococci in orthopaedic infections is also on an upward trend [46]. In our cohort, 7.5% of the patients had enterococcal PJI.

Before drawing final conclusions, a few limitations of this study need to be considered. Relatively few patients were included, making statistical analysis and determining significance challenging in certain cases. Our study cohort included patients with PJI affecting different joints, which could also interfere with our results. However, identical therapeutic procedures were performed in all cases, therefore in our opinion, this factor might have only moderate effect on the final conclusions. Also, the surgical technique of DAIR was not exactly the same for all patients as mobile parts were not always exchanged, therefore this variable was also evaluated in our study.

The majority of the patients had PJI caused by a rifampicin-sensitive organism, however, in our study, we could not find enough evidence to confirm that rifampicin resistance is associated with significantly lower recovery rates. Also, the isolation of rifampicin-sensitive organism was not more frequent among recovered patients and only Gram-positive organisms were observed in the treatment failure group. No development of rifampicin resistance was observed in the study population. The most frequent isolates were Staphylococci in the sensitive and Gram-negative rods in the resistant group.

We found that rifampicin resistance, BMI and sex had no statistically significant impact on recovery rates, although increasing age and diabetes may well have a negative clinical impact on clinical outcome. In conclusion, recognition of microbiological and patient-related factors may help estimate the risk and reduce treatment failure rates after DAIR performed in patients with PJI. Further investigations with larger patient cohort are needed to confirm the effect of rifampicin resistance on recovery rates in patients with PJI undergoing DAIR procedure.