23 N. Benito and M. Franco contributed equally to this manuscript
N. Benito
Correspondence
Corresponding author. N. de Benito, Unit of Infectious Diseases, Hospital de la Santa Creu i Sant Pau—Institut d'Investigació Biomèdica Sant Pau, Sant Antoni Maria Claret, 167, 08025 Barcelona, Spain.
23 N. Benito and M. Franco contributed equally to this manuscript
Affiliations
Infectious Diseases Unit, Hospital de la Santa Creu i Sant Pau—Institut d'Investigació Biomèdica Sant Pau, Barcelona, SpainDepartment of Medicine, Universitat Autònoma de Barcelona, SpainSpanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, Spain
23 N. Benito and M. Franco contributed equally to this manuscript
M. Franco
Footnotes
23 N. Benito and M. Franco contributed equally to this manuscript
Affiliations
Infectious Diseases Unit, Hospital de la Santa Creu i Sant Pau—Institut d'Investigació Biomèdica Sant Pau, Barcelona, SpainDepartment of Medicine, Universitat Autònoma de Barcelona, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitari Bellvitge, Barcelona, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Clínic Universitari, Barcelona, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitari Vall d'Hebron, Barcelona, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitario Ramón y Cajal, Madrid, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitario Valdecilla, Santander, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitario Virgen Macarena, Sevilla, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitario Cruces, Bilbao, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitario Virgen del Rocío, Sevilla, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainInfectious Diseases Unit, Department of Infectious Diseases, Hospital Universitario Son Espases, Palma de Mallorca, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitari Bellvitge, Barcelona, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Clínic Universitari, Barcelona, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitari Vall d'Hebron, Barcelona, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Clinical Microbiology, Hospital de la Santa Creu i Sant Pau—Institut d'Investigació Biomèdica Sant Pau, Barcelona, Spain
Spanish Network for the Research in Infectious Diseases (REIPI RD12/0015), Instituto de Salud Carlos III, Madrid, SpainDepartment of Infectious Diseases, Hospital Universitari Bellvitge, Barcelona, Spain
It is important to know the spectrum of the microbial aetiology of prosthetic joint infections (PJIs) to guide empiric treatment and establish antimicrobial prophylaxis in joint replacements. There are no available data based on large contemporary patient cohorts. We sought to characterize the causative pathogens of PJIs and to evaluate trends in the microbial aetiology. We hypothesized that the frequency of antimicrobial-resistant organisms in PJIs has increased in the recent years. We performed a cohort study in 19 hospitals in Spain, from 2003 to 2012. For each 2-year period (2003–2004 to 2011–2012), the incidence of microorganisms causing PJIs and multidrug-resistant bacteria was assessed. Temporal trends over the study period were evaluated. We included 2524 consecutive adult patients with a diagnosis of PJI. A microbiological diagnosis was obtained for 2288 cases (90.6%). Staphylococci were the most common cause of infection (1492, 65.2%). However, a statistically significant rising linear trend was observed for the proportion of infections caused by Gram-negative bacilli, mainly due to the increase in the last 2-year period (25% in 2003–2004, 33.3% in 2011–2012; p 0.024 for trend). No particular species contributed disproportionally to this overall increase. The percentage of multidrug-resistant bacteria PJIs increased from 9.3% in 2003–2004 to 15.8% in 2011–2012 (p 0.008), mainly because of the significant rise in multidrug-resistant Gram-negative bacilli (from 5.3% in 2003–2004 to 8.2% in 2011–2012; p 0.032). The observed trends have important implications for the management of PJIs and prophylaxis in joint replacements.
Prosthetic joint replacement is one of the most useful medical advances of recent decades and is increasingly performed worldwide. Although prosthetic joint infection (PJI) occurs in a small proportion of patients (1%–3%), it is a devastating complication and the absolute number of such infections is expected to rise in the coming years [
In clinical practice, knowledge of the microbiological spectrum of PJIs is of paramount importance. First, this information is essential for guiding empiric antibiotic therapy, particularly in early postoperative infections [
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
]. Patients who receive effective initial treatment before microbiological results are confirmed are less likely to experience treatment failure, according to a recent study [
]. Second, the surgical antimicrobial prophylaxis chosen for joint replacement should cover the most common pathogens causing surgical site infections [
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
Microbiological aetiology, epidemiology, and clinical profile of prosthetic joint infections: are current antibiotic prophylaxis guidelines effective?.
Organism profile in periprosthetic joint infection: pathogens differ at two arthroplasty infection referral centers in Europe and in the United States.
]. The threat of infections caused by multidrug-resistant organisms is increasing worldwide, yet little is known about their possible role in PJIs. There are no available data based on large contemporary patient cohorts to address these questions.
Our aim was to characterize the pathogens causing PJIs and to evaluate trends in microbial aetiology in a large cohort of patients from 2003 to 2012. We hypothesized that the frequency of antimicrobial-resistant organisms in PJIs has increased in recent years.
Methods
Setting, study design and patients
An ambidirectional observational study was performed in 19 hospitals in Spain, within the framework of the Spanish Network for Research in Infectious Diseases (REIPI) (www.reipi.org). The REIPI Group for the Study of Prosthetic Joint Infections is a multicentre collaborative research group of infectious disease specialists and microbiologists nationwide with long-term experience of managing orthopaedic infections.
Consecutive patients older than 16 years with PJIs diagnosed from 2003 to 2012 were included in the current study. Only episodes of infection diagnosed for the first time during the study period were included; reactivation of infection prior to this period was excluded.
Data collection
This cohort study was ambidirectional, with both prospective and retrospective data collection. First, data were obtained from the REIPI cohort, which prospectively enrolled consecutive patients with PJIs from 2003 to 2006. Characteristics of this cohort have been previously described [
]. Except for queries on critical variables, additional information was not requested for cases from the REIPI prospective cohort. Second, we also retrospectively collected data from patients who developed PJIs from 2007 through 2012 from the REIPI and other hospitals meeting the participation criteria. The participation criteria included: (1) centres with access to orthopaedic surgery, (2) identification procedures to ensure that all consecutive cases diagnosed at the centre were included and that ascertainment bias was minimized; (3) availability of and/or access to most of the requested data for resolving queries. A standard case report form specifically developed for this study was used to collect data at all sites. Most participating centres have electronic databases with prospectively collected information on patients with PJI; data were obtained from the databases and, when necessary, from the patient's medical records at each participating hospital. Completed case report forms were sent to the coordinating centre for data entry or were entered directly into the common electronic database by site investigators. The Hospital de la Santa Creu i Sant Pau (Barcelona, Spain) was the coordinating centre for the current study. The Institutional Review Board of the Hospital de la Santa Creu i Sant Pau approved the study before data collection. All case report forms were reviewed at the coordinating centre. The process of gathering data, reviewing the case report forms, and sending and resolving queries was carried out by the coordinating centre between January 2013 and December 2014.
Clinical data and definitions
We collected information on patient demographics and pre-existing conditions, arthroplasty characteristics, classification of the PJI and microbiological diagnosis. Definitions were established for all variables to ensure standardized data collection. For every patient, the following data were recorded: age and gender; co-morbidities and immunosuppressive therapy; the American Society of Anesthesiologists (ASA) score for the patient before the surgical procedure closest to diagnosis of infection (usually the implant of the arthroplasty); previous exposure to antibiotics (≥7 days) or hospitalization in the previous 90 days (≥2 days); receipt of haemodialysis, intravenous therapy, wound care or specialized nursing care at home in the 30 days before the last surgical procedure or onset of haematogenous PJI; residence in a nursing home or long-term care facility. We collected the following information about the arthroplasty: the reason for and date of implantation, site, time from admission to implantation, primary or revision arthroplasty, cemented versus uncemented arthroplasty and use of antibiotics in bone cement. Date of diagnosis, classification of the PJI, type and number of cultured samples and their results were also recorded.
Cefazolin – in some centres, cefuroxime – was the surgical antibiotic prophylaxis used. Vancomycin or teicoplanin was used in penicillin-allergic patients. We used the Charlson co-morbidity score to quantify baseline co-morbidities [
]. Infections were classified as follows: (1) Early post-interventional PJI: Infection that manifests within 1 month following an invasive procedure (surgery or arthrocentesis); (2) Chronic PJI: Infection with symptoms that persist for >3 weeks and are outside the early post-interventional period; (3) Acute haematogenous PJI: Infection with symptoms lasting ≤3 weeks after an uneventful postoperative period [
]. Culture specimens were collected and processed at each participating institution, following Spanish guidelines for the microbiological diagnosis of bone and joint infections [
Diagnóstico microbiológico de las infecciones osteoarticulares.
in: Cercenado E. Cantón R. Procedimientos en Microbiol. Clínica. Recomendaciones de la Sociedad Española Enfermedades Infecciosas y Microbiol. Clínica. 2009
]. Identification and susceptibility testing of isolates was performed in the clinical microbiology laboratory at each centre, using standard microbiological techniques and in accordance with methods approved by the Clinical and Laboratory Standards Institute [
Diagnóstico microbiológico de las infecciones osteoarticulares.
in: Cercenado E. Cantón R. Procedimientos en Microbiol. Clínica. Recomendaciones de la Sociedad Española Enfermedades Infecciosas y Microbiol. Clínica. 2009
]. The microbial aetiology of PJI was established when two or more intraoperative cultures or a combination of preoperative aspiration and intraoperative cultures yield the same organism (indistinguishable based on common laboratory tests including genus and species identification or common antibiogram) [
]. When the diagnostic criteria for PJI were met (according to the previous definition), a virulent microorganism (e.g. Staphylococcus aureus) isolated in a single specimen of a tissue biopsy or synovial fluid was also considered as the causative organism[
]; since coagulase-negative staphylococci are frequently not identified to species level (and inter-centre differences were observed), we did not consider polymicrobial infection when different coagulase-negative staphylococci (species or antibiograms) were isolated simultaneously. Multidrug-resistant organisms were defined according to Magiorakos et al. (acquired non-susceptibility to at least one agent in three or more specified antimicrobial categories) [
Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance.
Continuous variables were summarized as medians and interquartile ranges and categorical variables as percentages of the total sample for that variable. We determined the overall and biennial percentages of culture-positive PJIs and the proportion of polymicrobial infections. We then calculated the overall and biennial percentages of culture-positive PJIs accounted for by the most common organisms and groups of organisms, and the proportion of multidrug-resistant bacteria. Overall percentages for organisms isolated in PJIs were estimated with a 95% CI. The Mantel–Haenszel χ2 test for trend was used to determine whether there was a statistically significant linear trend in the proportion of infections due to the most frequent microorganisms and multidrug-resistant bacteria over the study period. Trends for quantitative variables were examined using the Spearman rank correlation test. A significant level of p <0.05 was used for all statistical tests. Data were analysed using IBM® SPSS®, version 22.0.
Results
A total of 2524 episodes of PJIs were diagnosed in adults from 2003 through 2012 in 19 participating hospitals located in 8 of 17 administrative Spanish regions (‘autonomous communities’); all except one were university hospitals. Seventeen hospitals had more than 500 beds, and two between 400 and 500 beds.
Patient characteristics
Patient characteristics are outlined in Table 1. The proportion of males increased from 37.5% in 2003–2004 to 42.9% in 2011–2012 (p 0.048 for trend). The median age of the patients at the time of diagnosis of infection increased from 72 (interquartile range 13) in 2003–2004 to 75 (interquartile range 15) in 2011–2012 (p 0.001). The percentage of patients with a Charlson score ≥2 increased significantly during the study period, from 38.2% in 2003–2004 to 41.4% in 2011–2012 (p 0.046). Rising trends were found for the percentage of patients receiving chronic immunosuppressive treatments, with neurological disorders and malignant neoplasms, increasing from 3.9%, 8% and 7.4%, respectively, in 2003–2004 to 7.4%, 11.6% and 11.5% in 2011–2012 (p 0.013, p 0.005 and p 0.007, respectively). For other underlying conditions, there were no significant trends over the period. The vast majority of patients underwent hip or knee arthroplasty, each of which accounted for almost half the cases. Primary arthroplasties accounted for 77.2% of infected prosthetic joints, and degenerative joint disease was the most common reason for joint replacement.
Table 1Characteristics of patients with a prosthetic joint infection diagnosed from 2003 through 2012
A total of 1309 (52.7%) episodes of PJI were classified as chronic PJI, 888 (35.7%) as early post-interventional PJI, and 288 (11.6%) as acute haematogenous infections.
Microbial aetiology
A microbiological diagnosis was obtained in 2288 cases (90.6%): 300 cases in 2003–2004 (89.3%), 344 in 2005–2006 (92.2%), 426 (92%) in 2007–2008, 560 (90%) in 2009–2010, and 658 (90.1%) in 2011–2012. The proportion of cases with a microbiological diagnosis did not vary significantly during the study period (p 0.691). Information about patients receiving antibiotics in the 14 days before reliable microbiological sampling was obtained was available for 2027 episodes. A total of 497 patients (24.5%) had received antimicrobial treatment, 15.7% (78) of whom had a culture-negative PJI; 6.6% (101) of 1530 patients without antibiotics had a culture-negative PJI (p <0.001). In all, 17.4% of all episodes (399) were polymicrobial. There were no significant trends over time in the proportion of polymicrobial PJIs (p 0.303).
Table 2 lists the causative microorganisms of PJIs during the study period. Aerobic Gram-positive cocci were the most common group of organisms causing PJIs, followed by aerobic Gram-negative bacilli. The biennial proportion of these microorganisms, however, changed during the study period (Fig. 1a). A statistically significant rising linear trend was observed for PJIs caused by aerobic Gram-negative bacilli (p 0.024), mainly due to the increase in the last 2-year period (25% in 2003–2004, 33.3% in 2011–2012), whereas aerobic Gram-positive cocci decreased from 80.3% in 2003–2004 to 74.3% in 2011–2012 (p 0.020). Even though fungi were not commonly involved in PJIs, the proportion of infections due to them doubled from 0.7% in 2003–2004 to 1.5% in 2011–2013 (p 0.049). There were no significant trends over time in the proportion of aerobic Gram-positive bacilli and anaerobic bacteria.
Table 2Microbiology results for culture-positive prosthetic joint infections
Microorganism or microorganism group
Total no. (%; 95% CI) of culture-positive infections (n = 2288)
a More than one aerobic Gram-positive coccus was isolated in 232 out of 1777 (13.1%) episodes of prosthetic joint infection where these organisms were identified.
b More than one species of coagulase-negative staphylococci was identified in 81 out of 905 (9%) episodes of prosthetic joint infection where these microorganisms were involved.
c Staphylococcus haemolyticus 10, S. simulans 5, S. saccharolyticus 4, S. schleiferi 4, S. cohnii 3, S. intermedius 3, S. lentus 1, S. saprophyticus 1.
bb Prevotella species 5 (P. bivia 2, P. corporis 1, P. melaninogenica 1, P. buccae 1), Parabacteroides distasonis 1, Porphyromonas asaccharolytica 1, Fusobacterium sp. 1.
Unless stated otherwise, data are number (%) of patients with indicated characteristic.
a More than one aerobic Gram-positive coccus was isolated in 232 out of 1777 (13.1%) episodes of prosthetic joint infection where these organisms were identified.
b More than one species of coagulase-negative staphylococci was identified in 81 out of 905 (9%) episodes of prosthetic joint infection where these microorganisms were involved.
c Staphylococcus haemolyticus 10, S. simulans 5, S. saccharolyticus 4, S. schleiferi 4, S. cohnii 3, S. intermedius 3, S. lentus 1, S. saprophyticus 1.
d Two species of viridans streptococci were involved in four prosthetic joint infection cases.
e Streptococcus mitis 18, S. oralis 7, S. sanguis 5, S. parasanguis 2.
f Streptococcus anginosus 13, S. intermedius 6, S. constellatus 5.
g Streptococcus bovis group 3, S. salivarius 2, nutritionally variant (deficient) streptococci 1.
h Enterococcus faecalis and E. faecium were involved in one episode of prosthetic joint infection.
i Enterococcus gallinarum 2, E. hirae 1, E. durans 1, E. casseliflavus 1, E. avium 1.
s Other Gram-negative bacilli include: Pasteurella multocida 3, Haemophilus spp. 2, Campylobacter fetus 1.
t Corynebacterium diphtheriae 6, C. jeikeium 4, C. aquaticum 1, C. ulcerans 1.
u More than one anaerobic bacterium was involved in eight cases of prosthetic joint infection.
v Two species of anaerobic Gram-positive bacilli were identified in one prosthetic joint infection.
w Propionibacterium acnes 83, P. avidum 6, Propionibacterium without identification to species level 22.
x Clostridium perfringens 3, C. absonum 1, C. ramosum 1, C. septicum 1, C. sphenoides 1.
y Finegoldia magna 5, Parvimonas micra 5, Peptostreptococcus anaerobius 3, Peptococcus niger 4, Peptostreptococcus not identified to species level 15.
z Two patients had more than one species of aerobic Gram-negative bacilli (3 species in one case and 2 species in the other one).
aa Bacteroides fragilis 12, B. stercoris 2, B. thetaiotamicron 1, Bacteroides sp. 1.
bb Prevotella species 5 (P. bivia 2, P. corporis 1, P. melaninogenica 1, P. buccae 1), Parabacteroides distasonis 1, Porphyromonas asaccharolytica 1, Fusobacterium sp. 1.
cc Veillonella sp.
dd Mycobacterium tuberculosis 5, M. fortuitum 4.
ee Candida albicans 16, C. parapsilosis 6, C. glabrata 2, C. tropicalis 1, C. famata 1, Candida sp. 1.
Fig. 1Trends in the microbial aetiology of prosthetic joint infections (a): distribution of aerobic Gram-positive cocci (b) and Gram-negative bacilli (c). p values indicate p for trend from 2003–2004 to 2011–2012.
The majority of cases of PJIs were caused by staphylococci (1492, 65.2%; 95% CI 63.2%–67.2%), followed by Enterobacteriaceae (466, 20.4%; 95% CI 18.7%–22%). The proportion of staphylococci decreased over the study period (67% in 2003–2004, 62.3% in 2011–2012; p 0.093) and the percentage of Enterobacteriaceae increased (17.7% in 2003–2004, 24.2% in 2011–2012; p 0.070), although these changes were not statistically significant. The following six species, in decreasing order, were involved in more than 80% of all PJIs: Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli, Pseudomonas aeruginosa, Enterococcus faecalis and Propionibacterium acnes (however, it should be remembered that, in many cases, coagulase-negative staphylococci were not identified to species level). Among the aerobic Gram-positive cocci and Gram-negative bacilli, the specific microbial aetiology did not change significantly overall during the study period and no particular organism contributed disproportionately to the decline in Gram-positive cocci or the increase in Gram-negative bacilli as causes of PJI (Fig. 1b,c). The only significant trends over time were the proportion of infections caused by Morganella morganii, increasing from 0.7% in 2003–2004 to 3.3% in 2011–2012 (p 0.012), and the declining percentage of Salmonella species, from 0.7% in 2003–2004 to 0% in 2011–2012 (p 0.041). There were no significant trends for other isolated microorganisms.
The four most frequent species involved in each of the classification groups were: (1) Chronic PJI: S. epidermidis (391, 33.2%), S. aureus (236, 20%), coagulase-negative staphylococci not identified at the species level (197, 16.7%) and Propionibacterium acnes (61, 5.2%); (2) Early post-interventional PJI: S. aureus 299 (35.6%), S. epidermidis 130 (15.5%), E. coli 129 (15.4%) and P. aeruginosa 128 (15.3%); (3) Acute haematogenous infections: S. aureus 104 (39.2%), E. coli 33 (12.5%), Streptococcus agalactiae 29 (10.9%) and viridans group streptococci 12 (4.5%).
Multidrug-resistant PJIs
Multidrug-resistant bacteria (following the specified definition) were involved in 321 PJIs (14%, 95% CI 12.6%–15.5%) during the study period, including 180 methicillin-resistant S. aureus (MRSA) (7.9%, 95% CI 6.8%–9.1%) and 146 multidrug-resistant Gram-negative bacilli (6.3%, 95% CI 5.4%–7.4%); MRSA and multidrug-resistant Gram-negative bacilli were simultaneously involved in five PJIs. The proportion of PJIs caused by multidrug-resistant bacteria increased from 9.3% in 2003–2004 to 15.8% in 2011–2012 (p 0.008) (Fig. 2). The percentage of MRSA increased from 4.7% in 2003–2004 to 9.5% in 2009–2010, but decreased to 7.6% in 2011–2012 (p 0.183). The following four species accounted for almost 90% of all PJIs due to multidrug-resistant Gram-negative bacilli: E. coli (74, 3.2%, 95% CI 2.5%–4%), P. aeruginosa (25, 1.1%, 95% CI 0.6%–1.5%), Klebsiella species (17, 0.7%, 95% CI 0.4%–1.1%) and Proteus species (15, 0.7%, 95% CI 0.3%–1%). Multidrug-resistant Gram-negative bacilli increased from 5.3% in 2003–2004 to 8.2% in 2011–2012 (p 0.032); specifically, there was an increase over time in the proportion of multidrug-resistant E. coli (the proportion doubled from 2% in 2003–2004 to 4.3% in 2011–2012; p 0.061), Klebsiella pneumoniae (0% in 2003–2004 to 1.1% in 2011–2012; p 0.051), P.–aeruginosa (0.7% in 2003–2004 to 1.8% in 2011–2012; p 0.044), and M. morganii (0% in 2003–2004 to 0.8% in 2011–2012; p 0.025). Forty-two of 146 multidrug-resistant Gram-negative bacilli were extended-spectrum β-lactamases producing Enterobacteriaceae; in the global series, the proportion (1.9%, 95% CI 1.3%–2.5%) increased from 0.7% in 2003–2004 to 2.6% in 2011–2012 (p 0.117). We found nine AmpC β-lactamase-producing Enterobacteriaceae and three carbapenemase-producing Gram-negative bacilli. Other resistance mechanisms were involved in multidrug-resistant aerobic Gram-negative bacilli leading to non-susceptibility to at least one agent in three or more antimicrobial categories [
Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance.
]. Concerning resistance to specific antibiotics, the most relevant was the notable increase in non-susceptibility to ciprofloxacin among Gram-negative bacilli (112 isolates, 17.7%), ranging in the global series from 3% in 2003–2004 to 6.4% in 2011–2012 (p <0.039).
Fig. 2Trends in the multidrug-resistant organisms involved in prosthetic joint infections. GNB indicates Gram-negative bacilli. p values indicate p for trend from 2003–2004 to 2011–2012.
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
Microbiological aetiology, epidemiology, and clinical profile of prosthetic joint infections: are current antibiotic prophylaxis guidelines effective?.
]. However, during the study period, we found a significant linear increase in the proportion of PJIs caused by aerobic Gram-negative bacilli, mainly due to the increase in the last 2-year period; no particular species contributed disproportionately to this overall increase. There was a significant increase in the percentage of fungi as causative pathogens, although few cases were isolated. Antimicrobial resistance also emerged among organisms isolated during the study period, with a significant rise in the proportion of multidrug-resistant Gram-negative bacilli.
To make appropriate empirical antimicrobial decisions, the common microbiological causes of PJIs should be known. This is mainly of interest for patients with acute PJIs treated with debridement and implant retention without a microbiological diagnosis before surgical intervention. In patients with all hardware removed, a delay before appropriate treatment should be less crucial. Ours is the largest study providing information about causative pathogens of PJIs. As in previous series, we found that aerobic Gram-positive cocci are implicated in most of these infections, which are largely driven by Staphylococcus infection [
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
Microbiological aetiology, epidemiology, and clinical profile of prosthetic joint infections: are current antibiotic prophylaxis guidelines effective?.
]. Recent studies in different geographical areas have reported higher frequency of these pathogens in surgical site infections following arthroplasties, ranging from 17% to 36% [
Activity of commonly used antimicrobial prophylaxis regimens against pathogens causing coronary artery bypass graft and arthroplasty surgical site infections in the United States, 2006–2009.
Microbiological aetiology, epidemiology, and clinical profile of prosthetic joint infections: are current antibiotic prophylaxis guidelines effective?.
]. In the present report, aerobic Gram-negative bacilli, mainly Enterobacteriaceae, were isolated in 28% of infections. Of interest, our study is the first to demonstrate an increase in the proportion of PJIs due to Gram-negative organisms. Several studies that have characterized the microbial aetiology of healthcare-associated infections like bloodstream infections over the past decade have reported similar consistent increases in infections due to Gram-negative organisms [
Predominance of Gram-negative bacilli and increasing antimicrobial resistance in nosocomial bloodstream infections at a university hospital in southern Taiwan, 1996–2003.
Our observed rise of multidrug-resistant organisms as a cause of PJIs, mainly attributable to the increase of antimicrobial-resistant Gram-negative infections, is of particular concern. Over the last decade, the implication of multidrug-resistant Gram-negative bacilli in healthcare-associated infections has steadily increased and is a public health issue of growing importance in Europe and worldwide [
]. The notable (almost 18%) and increasing resistance to quinolones found in the present study is of greatest concern because ciprofloxacin is considered a cornerstone in the treatment of PJIs caused by Gram-negative bacilli [
]. Antimicrobial therapy of PJIs will therefore become increasingly complex, and the importance of an individualized and accurate aetiological diagnosis will increase in the future. On the other hand, the percentage of MRSA is now stabilizing or decreasing in most European countries and the USA [
], and we also found a smaller proportion of infections caused by MRSA in the last 2 years of the study period.
Explanations for the observed microbial trends remain unclear. Possible explanations include differences in patient population over time; we observed that patients with PJIs were increasingly older and more often had complex underlying diseases. Other possibilities, such as changes in surgical procedures or microbiological diagnosis, seem unlikely causes of the observed trends.
Antimicrobial prophylaxis in surgery is one of the most effective measures for preventing surgical site infections [
]. However, concerns arise about the appropriateness of currently recommended regimens for joint replacement surgery, as recent studies indicate that they may have inadequate activity against a substantial proportion of bacteria involved in surgical site infections, ranging from 54% to 68% [
Microbiological aetiology, epidemiology, and clinical profile of prosthetic joint infections: are current antibiotic prophylaxis guidelines effective?.
Activity of commonly used antimicrobial prophylaxis regimens against pathogens causing coronary artery bypass graft and arthroplasty surgical site infections in the United States, 2006–2009.
]. The shifts in microbial aetiology observed in our investigation argue in favour of reassessing antimicrobial prophylaxis strategies. However, the need to minimize PJI risk must be weighed against the potentially undesirable consequences of expanding the spectrum of antimicrobial prophylaxis regimens. We suggest that identifying risk factors for PJIs caused by antimicrobial-resistant pathogens would help identify high-risk patients that may benefit from broader-spectrum prophylaxis regimens.
The limitations of our study are first, those implicit to a retrospective observational design, although our main strength is that it includes data from a large number of PJI cases from many hospitals over a long period of time, which would be very difficult to execute with another type of design. Second, the study assesses microbial aetiology and trends in our country. Given differences in patient characteristics, patient care, and hospital and health system factors, our results may not be generalizable to other countries. Nonetheless, recent findings from other studies suggest that the microbiological trends observed could be similar in different geographical areas worldwide. We used a standardized definition of multidrug-resistant microorganism [
Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance.
], which is another strength of our study and enables comparisons of results with other centres.
In summary, ours is the largest study reporting information about the microbial aetiology of PJIs. Most infections are caused by staphylococci, although the rate of infection by Gram-negative bacilli and fungi increased from 2003 through 2012, as did the proportion of multidrug-resistant infections, mainly due to the increase of resistant Gram-negative bacilli. These results suggest that empiric and specific antimicrobial therapy of PJIs could become more challenging. Reassessing antimicrobial prophylaxis strategies and other preventive measures for patients undergoing joint replacement could be required. Identifying risk factors for antimicrobial-resistant infections may help prevent them and improve treatment for these patients.
Funding
This work was supported by Plan Nacional de I+D+i and Instituto de Salud Carlos III , Subdirección General de Redes y Centros de Investigación Cooperativa , Ministerio de Economía y Competitividad , Spanish Network for Research in Infectious Diseases ( REIPI RD12/0015 )—co-financed by European Development Regional Fund ‘A way to achieve Europe’ ERDF.
Transparency Declaration
The funder had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had access to all data in the study and had final responsibility for the decision to submit for publication.
All authors report no conflicts of interest relevant to this article.
Contribution to Authorship
NB conceived the study. NB and MF searched the literature, designed the study, analysed the data, interpreted the results, and drafted the report. AR, AS, DR, LS, GF, MDT, LG, ES, AB, MR, JE, JMB, JM, AJ, CD, AR, BS, GE, LM, CP and IM collected the data and revised critically the report for important intellectual content. PC and JA supervised the study, interpreted the results and revised critically the report. NB and MF contributed equally. All authors give final approval of the version to be published.
Members of the Group for the Study of Prosthetic Joint Infections (in alphabetical order)
Hospital Universitari Arnau de Vilanova, Lleida: Fernando Barcenilla, Ferran Pérez-Villar, Laura Prats-Gispert. Hospital de Basurto, Bilbao: Ramón Cisterna, Sofía Ibarra, Íñigo López, Juan M. Santamaría. Hospital Universitari de Bellvitge, Barcelona: Javier Cabo, Dolores García, Jaime Lora-Tamayo (currently, Hospital 12 de Octubre), Oscar Murillo, Salvador Pedrero. Hospital el Bierzo, León: Susana Álvarez-Parrondo, Rafael Muedra-Font, Carmen Raya-Fernández, Cristina Rodríguez-Alonso. Hospital Universitario Central de Asturias, Oviedo: Alfonso Moreno. Hospital Universitario de Burgos, Burgos: María Aránzazu Blanco-Martínez-de-Morentin, Rebeca Cabo-Magadan. Hospital Clínic Universitari, Barcelona: Andreu Combalia, Sebastián García, Juan C. Martínez-Pastor, Eduard Tornero. Hospital Universitario Cruces, Bilbao: Josu Merino-Pérez, José Miguel Montejo. Hospital del Mar, Barcelona: Albert Alier, Juan P. Horcajada, Virginia Plasencia, Luis Puig. IIS-Fundación Jiménez Díaz: Álvaro Auñon, Antonio Blanco, Joaquín García-Cañete, Enrique Sandoval. Hospital Universitario Marqués de Valdecilla, Santander: Michel Fakkas-Fernández, Carlos Garcés-Zarzalejo, Concepción Fariñas-Alvarez, María del Carmen Fariñas, Luis Martinez-Martinez, Carlos Salas-Venero. Hospital Universitario Ramón y Cajal, Madrid: Javier Cobo, Patricia Ruiz-Carbajosa. Hospital de la Santa Creu i Sant Pau, Barcelona: Marcos Jordán, Xavier Crusi. Hospital Universitario Son Espases, Palma de Mallorca: Carmen Marinescu, Francisco Montaner, Antonio Ramírez. Hospital Universitari Vall d’Hebron, Barcelona: Pablo S. Corona, Maily Lung. Hospital Universitario Virgen Macarena, Sevilla: Miguel Ángel Muniain-Ezcurra, Cecilia Peñas-Espinar, Ana Isabel Suárez. Hospital Universitario Virgen del Rocío, Sevilla: Rocío Álvarez, José-Antonio Cordero, Macarena López-Pliego, Julián Palomino, Andrés Puente.
References
Peel T.N.
Buising K.L.
Choong P.F.M.
Diagnosis and management of prosthetic joint infection.
Guiding empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic joint infection managed by debridement, irrigation and prosthesis retention.
Microbiological aetiology, epidemiology, and clinical profile of prosthetic joint infections: are current antibiotic prophylaxis guidelines effective?.
Organism profile in periprosthetic joint infection: pathogens differ at two arthroplasty infection referral centers in Europe and in the United States.
Diagnóstico microbiológico de las infecciones osteoarticulares.
in: Cercenado E. Cantón R. Procedimientos en Microbiol. Clínica. Recomendaciones de la Sociedad Española Enfermedades Infecciosas y Microbiol. Clínica. 2009 (Available from:)
Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance.
Activity of commonly used antimicrobial prophylaxis regimens against pathogens causing coronary artery bypass graft and arthroplasty surgical site infections in the United States, 2006–2009.
Predominance of Gram-negative bacilli and increasing antimicrobial resistance in nosocomial bloodstream infections at a university hospital in southern Taiwan, 1996–2003.
30128-8&id=gr1.jpg){kind=link}
30128-8&id=gr2.jpg){kind=link}