Infection Control & Hospital Epidemiology http://journals.cambridge.org/ICE Additional services for Infection
Control & Hospital Epidemiology:
Email alerts: Click here Subscriptions: Click here Commercial reprints: Click here Terms of use : Click here
Deep Infection after Total Hip Replacement: A Method for National Incidence Surveillance Viktor Lindgren, Max Gordon, Per Wretenberg, Johan Kärrholm and Göran Garellick Infection Control & Hospital Epidemiology / Volume 35 / Issue 12 / December 2014, pp 1491 - 1496 DOI: 10.1086/678600, Published online: 16 January 2015
Link to this article: http://journals.cambridge.org/abstract_S0195941700093887 How to cite this article: Viktor Lindgren, Max Gordon, Per Wretenberg, Johan Kärrholm and Göran Garellick (2014). Deep Infection after Total Hip Replacement: A Method for National Incidence Surveillance. Infection Control & Hospital Epidemiology, 35, pp 1491-1496 doi:10.1086/678600 Request Permissions : Click here
Downloaded from http://journals.cambridge.org/ICE, IP address: 141.209.100.60 on 03 Nov 2015
infection control and hospital epidemiology
december 2014, vol. 35, no. 12
original article
Deep Infection after Total Hip Replacement: A Method for National Incidence Surveillance Viktor Lindgren, MD;1,2 Max Gordon, MD;2,3 Per Wretenberg, MD, PhD;1 Johan Ka¨rrholm, MD, PhD;2,4 Go¨ran Garellick, MD, PhD2,4
objective. The purpose of this study was to estimate the incidence of deep periprosthetic joint infections (PJIs) after primary total hip replacement (THR) in Sweden prior to the introduction of a national initiative to reduce these infections. design.
Prospective open cohort study with 2 years follow-up of each subject.
setting. All THR-performing clinics in Sweden. methods. All patients registered for a primary THR in the Swedish Hip Arthroplasty Register between July 1, 2005, and December 31, 2008, were selected for the study (45,531 patients with 49,219 THRs) and were matched with the Swedish Prescribed Drug Register. All patients with a minimum of 4 weeks of continuous outpatient antibiotic treatment within 2 years after their primary THR (1,989 patients with 2,219 THRs) were selected for a medical records review, and the number of cases with PJI was determined. results. The cumulative incidence of PJI within 2 years after primary THR was 0.9% (95% confidence interval, 0.85–1.02; n p 443), and 405 of these had been reoperated. The incidence rate of PJI for the first 3 months was 5 per 10,000 THR-weeks and, thereafter, 0.3 per 10,000 THR-weeks. Staphylococcus aureus and coagulase-negative staphylococci were the most common bacteria isolated. conclusions. This study describes a new method of national postoperative infection surveillance. The incidence is similar to previous smaller reports and is useful for monitoring changes over time to evaluate the national initiative to reduce infections. Infect Control Hosp Epidemiol 2014;35(12):1491-1496
The third most common complication following total hip replacement (THR) is periprosthetic joint infection (PJI).1 Commonly, it leads to great suffering in the patient, requires large resources of the healthcare system, and is associated with high costs to society, especially if affecting patients of working age.2-4 If diagnosed early and treated aggressively with debridement and irrigation, it may be possible to retain the implant. If discovered late, 1- or 2-stage exchange or resection is often needed.5,6 Regardless of surgical intervention, patients become dependent on antibiotics over a long period of time.7,8 In the literature, the incidence of deep infection following primary THR ranges between 0.2% and 1.6%, depending on follow-up time and the methodology of the study, but until now no study has had a complete national study base.9,10 The diagnosis of postoperative periprosthetic infection can be defined as early (diagnosed within 3 months), delayed (diagnosed between 3 and 24 months), or late (diagnosed after more than 24 months after the primary procedure). The early and delayed infections are often considered to be caused by peri- or intraoperative contamination. Whether the in-
fection is diagnosed before or after 3 months may depend on the aggressiveness of the bacteria.7 Preventive measures might therefore decrease the early and delayed infection rate, but the effectiveness of this work is difficult to evaluate because the overall incidence of infection is low and because of the large number of factors involved.11,12 In Sweden, a nationwide initiative called Prosthesis Related Infections Should Be Stopped (PRISS) was started in 2008 in order to reduce the incidence of PJI following both knee and hip replacement surgery. By 2013, all operating units in Sweden had adopted the pre-, peri-, and postoperative recommendations. The aim of this study was to find the cumulative incidence and incidence rate of early and delayed postoperative deep PJIs following primary THR, a baseline incidence of PJI after primary THR to later evaluate the PRISS initiative.
registries The Swedish Hip Arthroplasty Register (SHAR) was started in 1979, and all private and public orthopedic units in Sweden
Affiliations: 1. Department of Molecular Medicine and Surgery, Section of Orthopaedics, Karolinska Institutet, Stockholm, Sweden; 2. Swedish Hip Arthroplasty Register, Registercentrum Va¨stra Go¨taland, Gothenburg, Sweden; 3. Department of Clinical Sciences, Danderyd Hospital, Karolinska Institutet, Stockholm, Sweden; 4. Department of Orthopaedics, Institute of Clinical Sciences at Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden. Received April 28, 2014; accepted August 1, 2014; electronically published October 24, 2014. 䉷 2014 by The Society for Healthcare Epidemiology of America. All rights reserved. 0899-823X/2014/3512-0008$15.00. DOI: 10.1086/678600
1492
infection control and hospital epidemiology
december 2014, vol. 35, no. 12
voluntarily participate. A total of 98% of all primary THRs in Sweden are reported to the register.13 Apart from information regarding the operating unit, the register holds information including the patient’s personal registration number, age, sex, side, diagnosis, and a number of operative details regarding surgical approach, implant, fixation method, and so on. The SHAR is connected to the Swedish Death Register, thereby also including survival time of implant and patient. The Swedish Prescribed Drug Register (SPDR) was introduced on July 1, 2005. This is a register to which all pharmacies in Sweden are obligated to report and automatically includes all outpatient prescribed drugs dispensed in Sweden. The SPDR includes information regarding the drug, amount prescribed, amount dispensed, date of prescribing and dispensing, instruction from prescribing doctor, level of care and speciality of prescribing doctor, and so on. The SPDR is based on the personal registration number and can therefore be connected to other national healthcare and quality registries in order to evaluate, for example, postoperative infections.14 The SPDR is updated with new prescriptions every month and annually includes about 100 million new prescriptions.
methods The study was given ethical approval by the Regional Ethical Committee in Gothenburg on October 11, 2010 (reference 553-10). All operations between July 1, 2005, and December 31, 2008, reported to the Swedish Hip Arthroplasty Register for a primary THR were included in the study (n p 49,219). All diagnoses, bilateral hip replacements, and types of implants regardless of the method of fixation were included. By using patients’ personal registration numbers, cohorts were then matched with the SPDR for all dispensed antibiotic prescriptions between July 1, 2005, and December 31, 2010. All types of antibiotics with Anatomical Therapeutic Chemical codes J01, J04, and P01 were included (Figure 1). Because current recommendations in Sweden for treatment of PJIs all include long-term antibiotic treatment, and because uncomplicated wound infections were excluded, we limited the search by including only the dispensed amount of antibiotics, suggesting a continuous outpatient medication for at least 4 weeks (Table 1).15 We also limited the observation time to 2 years after the primary THR for each patient to include only early and delayed PJIs. Dispensed antibiotics for which instructions from the prescribing doctor specifically indicated treatment for an infection (eg, urinary tract infection, pneumonia) other than a PJI were excluded. By matching the 2 registries, we found that 1,989 patients with 2,217 THRs had been prescribed and dispensed more than 4 weeks of antibiotic treatment within the first 2 years after the primary THR. A questionnaire for each of the 2,217 THRs, including a list of dispensed antibiotics, was sent to a physician at the primary operating unit (76 different units) to complete and return. In the questionnaire, the receiving physician verified
figure 1. Study flowchart. SHAR, Swedish Hip Arthroplasty Register; SPDR, Swedish Prescribed Drug Register; THR, total hip replacement.
whether the patient had been treated for a deep PJI after the primary THR. If yes, supplementary information on date of diagnosis and how the diagnosis was set—including clinic presentation, laboratory markers, cultures, and radiology findings–was to be filled in as well as whether the patient was reoperated. Finally, the established infecting microorganism, if known, was to be specified. In this study, the diagnosis of deep PJI was established when a patient met 1 or more of the following objective criteria (adapted from the definition established by the Workgroup of the Musculoskeletal Infection Society):16 (1) open sinus tract to the joint; (2) 2 or more positive perioperative cultures of the same pathogen; (3) when the patient met 2 or more of the following criteria: (a) systematic infection and pus in the artificial joint; (b) C-reactive protein greater than 10 or erythrocyte sedimentation rate greater than 30; (c) 1 positive culture from joint fluid aspirate. We excluded superficial infections, PJI after revision surgery, and infections prior to THR (most often THR after failed infected osteosynthesis). Of the 2,217 questionnaires sent out, 2,191 (99%) were returned, and all orthopedic clinics contributed. Four of the patients had incorrectly been registered in the SHAR as primary THR when in fact they had a revision THR and were therefore excluded.
hip arthroplasty: deep infection surveillance
1493
table 1. Tablets/Doses Required for 4 Weeks of Continuous Treatment Antibiotic
ATC code
Daily dose
4-week treatment
Amoxicillin Amoxicillin and clavulanic acid Azitromycin Cefadroxil Cefalexin Ceftriaxon Ceftributen Cefuroxim Ciprofloxacin Ciprofloxacin Ciprofloxacin Daptomycin Ertapenem Erytromycin Fenoximetylpenicillin Flukloxacillin Flukloxacillin Fusidic acid Klaritromycin Klindamycin Levofloxacin Linezolid Lorakarbef Metronidazole Moxifloxacin Norfloxacin Rifampicin Roxitromycin Sulfametoxazol and trimetoprim Teikoplanin Telitromycin Vancomycin
J01CA04 J01CR02 J01FA10 J01DB05 J01DB01 J01DD04 J01DD14 J01DC02 J01MA02 J01MA02 J01MA02 J01XX09 J01DH03 J01FA01 J01CE02 J01CF05 J01CF05 J01XC01 J01FA09 J01FF01 J01MA12 J01XX08 J01DC08 P01AB01 J01MA14 J01MA06 J04AB02 J01FA06 J01EE01 J01XA02 J01FA15 J01XA01
750 mg 1#3 875/125 mg 1#2 250 mg 1#1 1 g 1#2 3 g 1#2 2 g 1#1 400 mg 1#1 250 mg 1#2 250 mg 1#2 500 mg 1#2 750 mg 1#2 400 mg 1#1 1 g 1#2 250 mg 2#2 1 g 2#3 500 mg 2#3 750 mg 2#3 250 mg 2#3 250 mg 1#2 300 mg 1#2 500 mg 1#1 600 mg 1#2 200 mg 1#2 400 mg 1#3 400 mg 1#1 400 mg 1#2 600 mg 1#1 150 mg 1#2 160 mg/800 mg 1#2 400 mg 1#1 400 mg 2#1 1 g 1#2
84 56 28 56 56 28 28 56 56 56 56 28 28 112 168 168 168 168 56 56 28 56 56 84 28 56 28 56 56 28 56 56
note.
ATC, Anatomical Therapeutic Chemical.
statistics Incidence was described as both the 2-year cumulative incidence and the incidence density rate. Each individual primary THR was studied separately (thereby including patients with bilateral THR). The cumulative incidence was calculated by dividing the total number of identified cases within 2 years after the primary THR by the total number of THRs in patients still alive and unrevised (except for infection) at 2 years after the primary operation. The incidence density rate was described by the number of diagnosed infections divided by the number of THRs at risk within 2 periods of time: before 3 months and at 3–24 months. Incidence density rates were expressed per 10,000 THR-weeks, where a THR-week is a unit of follow-up equal to 1 THR followed for 1 week. THRs in patients that died within each time period were censored, and once a THR was infected, it was excluded for further follow-up time. Confidence intervals (CIs) were calculated on the basis of the binomial distribution with the Wilson score interval. The cohort was also divided into half-years on the basis of the operation date for each THR, and cumulative
incidence calculated by each period and trends were tested with the binomial proportion trend test. Time to diagnosis was described as a median. If the date of diagnosis was missing, the first reoperation date was chosen as the date of diagnosis. We used IBM SPSS Statistics software (ver. 21) for Windows.
results A total of 443 THRs were treated for a deep PJI, representing an overall cumulative incidence of 0.9% (95% CI, 0.85–1.02). When dividing the 3.5-year study period into half-years, an increasing incidence was found (Figure 2). Ninety-one percent of the deep PJIs were reoperated (n p 405) as part of the treatment, and 90% (n p 398) were diagnosed according to the specified objective criteria (Table 2). A total of 425 patients had a specified date of diagnosis or reoperation date, and the median time to diagnosis was 24 days (Figure 3). The incidence density rate within the first 3 postoperative months was 5 cases per 10,000 THR-weeks (95% CI, 4.70–5.83), and for the remaining 3–24 months it
1494
infection control and hospital epidemiology
december 2014, vol. 35, no. 12
figure 2. Cumulative incidence (with 95% confidence interval) of deep periprosthetic joint infection after total hip replacement specified for each half-year of the study period.
decreased to 0.3 cases per 10,000 THR-weeks (95% CI, 0.21– 0.31). The cumulative incidence of the individual clinics ranged from 0% to 4.2% (median, 0.8%). In 393 of the 443 deep PJIs, a known pathogen was specified. Coagulase-negative staphylococci and Staphylococcus aureus were alone or in combination with other microorganisms responsible in 70% of the cases (Table 3).
discussion This study presents a unique model to establish an incidence of deep PJI following primary THR by combining health registries with a medical records review. The criteria employed are well defined, and repeated measurements during the follow-up period are possible. As previously mentioned, earlier studies have determined the PJI incidence after primary THR to be between 0.2% and 1.6%. The diagnostic criteria as well as the population studied and the follow-up time are factors that are all likely to significantly influence the result. The incidence of PJI is not always easy to determine in clinical practice, since the predictive power of a single test is low. Many observations taken together—such as clinical presentation, laboratory findings in blood and in joint aspiration, radiographic findings, histology, and, most importantly, cultures—are often used and needed to establish a diagnosis of PJI.17 Cultures indisputably have a central role in the diagnosis
but have considerable risks of being both false positive (contamination) and false negative because of preoperative antibiotic treatment, biofilm formation, poor managed samples, and a lacking standard for cultivation.16,18 The different definitions of a PJI and the historical absence of a standard makes comparisons difficult.19 The American Musculoskeletal Infection Society introduced a new standard in 2011,16 the practical usefulness of which has been subject to debate.20 It accepts the fact that a PJI can be present even though the criteria of diagnosis have not been fulfilled. Therefore, we included all patients where the doctor reported a deep PJI even though the criteria were not fulfilled, since the opposite would definitely lead to an underreporting of PJI incidence. From a patient point of view, it might not always be of importance how the diagnosis has been set but rather the clinical consequences of an infection and its treatment. Although 81 different doctors reviewed the medical records retrospectively, introducing a variability of data quality, we found that in most of the deep PJI cases, there was still enough information in the questionnaire to establish the diagnosis consistent with the adapted specified criteria from the Workgroup of the Musculoskeletal Infection Society (Table 2). The incidence of infections increased during the period of observation, which has been suggested previously.21 Whether this is a result of deterioration of hygiene and prophylactic standards, increasing patient comorbidity, or increasing resistance and virulence of the microorganisms is, however, not answered by this study. Nevertheless, it emphasizes the need for actions such as the Swedish PRISS initiative to reduce postoperative infection rates by improving all aspects of pre-, peri-, and postoperative care. Future studies are planned to determine the effect of this initiative in reducing the national infection burden. The advantages of this study include the nationwide coverage of all THR-performing clinics, the large study population, the long follow-up, and the possibility of an objective assessment of the clinical course after the surgical and medical treatment. Compared with previous studies of national PJI incidence that are based solely on administrative data based on a payment system (Medicare),10,22 our study identified infected patients by their treatment and not by coding of a diagnosis. Data based on International Classification of Diseases codes are imprecise in following specific PJIs because
table 2. Distribution of Diagnosis Criteria Diagnosis criteria
No.
%
1: Open sinus tract to the prosthesis 2: 2 or more positive perioperative cultivations 3a ⫹ 3b: systemic infection and pus in joint ⫹ elevated CRP/ESR 3a ⫹ 3c: systemic infection and pus in joint ⫹ cultivation from joint aspirate 3b ⫹ 3c: elevated CRP/ESR ⫹ cultivation from joint aspirate Clinical diagnosis/information lacking in questionnaire Total
64 263 23 19 29 45 443
14.4 59.4 5.2 4.3 6.5 10.2 100
note.
CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.
hip arthroplasty: deep infection surveillance
figure 3. Histogram illustrating time from primary total hip replacement to diagnosis of deep periprosthetic infection. Each bar represents 1 week.
the codes indicating infection lack accuracy and are not side or joint specific. They can therefore result in overestimation of infections or underestimation, depending on coding routines.22 We believe that the method used in this study therefore results in a more specific and hopefully more accurate estimation of the national incidence of deep PJIs in Sweden. Our study has limitations. Because only outpatient antibiotic treatment is recorded in the SPDR, the patients who died before discharge or those not receiving or complying with the antibiotic treatment would not be included in the questionnaire group. Furthermore, patients never returning to the primary operating clinic would not be recorded as a PJI, even though they might have consumed large doses of antibiotics. However, this group of patients is not expected to be large, especially since we excluded superficial infections. Although the register data are collected prospectively, the diagnosis of PJI was set retrospectively, which is also a possible weakness in this study. With these concerns in mind, we can conclude that the incidence found in this study is probably a slight underestimation of the true incidence. table 3. Distribution of Infecting Microorganism Microorganism
No.
%
Polymicrobiala Coagulase-negative Staphylococcus Staphylococcus aureus Streptococci Enterococcus faecalis Miscellaneous Unknown Total
109 109 89 32 16 38 50 443
24.6 24.6 20.1 7.2 3.6 8.6 11.3 100
a
All including either S. aureus and/or coagulase-negative Staphylococcus.
1495
Many countries have introduced surgical site infection surveillance systems to monitor postoperative infections after specific surgical procedures, often including hip replacement surgery. Perhaps these registries better monitor infection rates, since they can prospectively include more information regarding clinical presentation. However, in practice, it is difficult to ascertain complete registration from all clinics. Because Sweden lacks such a register, we combined information in the existing healthcare and quality registers with a patient’s personal registration number to find information that indicates an infection. Instead of inventing new registries, a model for automated matching of registries in order to find possible postoperative infections is an attractive alternative.23 The SPDR introduces the possibility of an unbiased report of infection incidence, since every dispensed prescription is recorded automatically. The SPDR seems to be sensitive in identifying patients with a possible postoperative infection but is not sufficient to monitor postoperative infections alone, since the population observed have other infections requiring antibiotic treatment. Future studies are planned to investigate how to automatically combine register data in order to increase specificity and maintain a high sensitivity without the need for review of medical records. The microbial spectra with a majority of infections caused by coagulase-negative staphylococci and S. aureus are consistent with previous studies.24,25 The problem with increasing incidence of methicillin-resistant S. aureus (MRSA) infections reported internationally does not seem to have been a clinical problem in Sweden in 2005–2008, since no MRSA infections were reported. In conclusion, this study demonstrates a method where existing registries can be combined in order to identify possible cases of infection and verify them by a medical records review. We believe that this method is close to the optimum way to track all infections in the postoperative period after THR. We found that the incidence of postoperative deep infection following THR was 0.9% between 2005 and 2008. As future changes are anticipated, new data can be collected and analyzed in order to compare with the preexisting situation.
acknowledgments We wish to thank Szilard Nemes for statistical advice and calculations and the Swedish Hip Arthroplasty Register and the Swedish National Board of Health and Welfare for contributing their databases for this research. ¨ F, Sweden. Financial support. Patient Insurance LO Potential conflicts of interest. All authors report no conflicts of interest relevant to this article. All authors submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest, and the conflicts that the editors consider relevant to this article are disclosed here. Address correspondence to Viktor Lindgren, MD, Department of Orthopaedics, Karolinska University Hospital, Solna, SE-171 76 Stockholm, Sweden ([email protected]).
1496
infection control and hospital epidemiology
december 2014, vol. 35, no. 12
references 1. Bozic KJ, Kurtz SM, Lau E, Ong K, Vail TP, Berry DJ. The epidemiology of revision total hip arthroplasty in the United States. J Bone Joint Surg Am 2009;91(1):128–133. 2. Klouche S, Sariali E, Mamoudy P. Total hip arthroplasty revision due to infection: a cost analysis approach. Orthop Traumatol Surg Res 2010;96(2):124–132. 3. Kurtz SM, Lau E, Watson H, Schmier JK, Parvizi J. Economic burden of periprosthetic joint infection in the United States. J Arthroplasty 2012;27(8 suppl):61–65.e1. 4. de Lissovoy G, Fraeman K, Hutchins V, Murphy D, Song D, Vaughn BB. Surgical site infection: incidence and impact on hospital utilization and treatment costs. Am J Infect Control 2009; 37(5):387–397. 5. Koyonos L, Zmistowski B, Della Valle CJ, Parvizi J. Infection control rate of irrigation and debridement for periprosthetic joint infection. Clin Orthop 2011;469(11):3043–3048. 6. Crockarell JR, Hanssen AD, Osmon DR, Morrey BF. Treatment of infection with debridement and retention of the components following hip arthroplasty. J Bone Joint Surg Am 1998;80(9): 1306–1313. 7. Zimmerli W, Ochsner PE. Management of infection associated with prosthetic joints. Infection 2003;31(2):99–108. 8. Giulieri SG, Graber P, Ochsner PE, Zimmerli W. Management of infection associated with total hip arthroplasty according to a treatment algorithm. Infection 2004;32(4):222–228. 9. Urquhart DM, Hanna FS, Brennan SL, et al. Incidence and risk factors for deep surgical site infection after primary total hip arthroplasty: a systematic review. J Arthroplasty 2010;25(8): 1216–1222.e3. 10. Ong KL, Kurtz SM, Lau E, Bozic KJ, Berry DJ, Parvizi J. Prosthetic joint infection risk after total hip arthroplasty in the Medicare population. J Arthroplasty 2009;24(6 suppl):105–109. 11. Adeli B, Parvizi J. Strategies for the prevention of periprosthetic joint infection. J Bone Joint Surg Br 2012;94(11 suppl A):42–46. 12. Jamsen E, Furnes O, Engesaeter LB, et al. Prevention of deep infection in joint replacement surgery. Acta Orthop 2010;81(6): 660–666. 13. Garellick G, Karrholm J, Rogmark C, Herberts P. Swedish Hip Arthroplasty: Annual Report 2010. Gothenburg: Swedish Hip Arthroplasty Register, 2010. http://www.shpr.se/Libraries/ Documents/AnnualReport-2010-2-eng.sflb.ashx. 14. Stridh Ekman G, Ringback Weitoft G, Nyren O, Dickman PW,
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Ericsson O, Struwe J. National surveillance of surgical-site infection through register-based analysis of antibiotic use after inguinal hernia repair. Br J Surg 2010;97(11):1722–1729. Christensson B, Johansson L, Nilsdotter A˚, et al. Va˚rdprogram fo¨r Led-Och Skelettinfektioner. 2008. http://infektion.net/sites /default/files/pdf/Vardprogram_led_och_skelett_2008.pdf. Accessed February 4, 2014. Parvizi J, Zmistowski B, Berbari EF, et al. New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society. Clin Orthop 2011;469(11):2992– 2994. Bernard L, Lubbeke A, Stern R, et al. Value of preoperative investigations in diagnosing prosthetic joint infection: retrospective cohort study and literature review. Scand J Infect Dis 2004;36(6/7):410–416. Malekzadeh D, Osmon DR, Lahr BD, Hanssen AD, Berbari EF. Prior use of antimicrobial therapy is a risk factor for culturenegative prosthetic joint infection. Clin Orthop 2010;468(8): 2039–2045. Parvizi J, Jacovides C, Zmistowski B, Jung KA. Definition of periprosthetic joint infection: is there a consensus? Clin Orthop 2011;469(11):3022–3030. Oussedik S, Gould K, Stockley I, Haddad FS. Defining periprosthetic infection: do we have a workable gold standard? J Bone Joint Surg Br 2012;94(11):1455–1456. Kurtz SM, Lau E, Schmier J, Ong KL, Zhao K, Parvizi J. Infection burden for hip and knee arthroplasty in the United States. J Arthroplasty 2008;23(7):984–991. Phillips CB, Barrett JA, Losina E, et al. Incidence rates of dislocation, pulmonary embolism, and deep infection during the first six months after elective total hip replacement. J Bone Joint Surg Am 2003;85A(1):20–26. Bolon MK, Hooper D, Stevenson KB, et al. Improved surveillance for surgical site infections after orthopedic implantation procedures: extending applications for automated data. Clin Infect Dis 2009;48(9):1223–1229. Pulido L, Ghanem E, Joshi A, Purtill JJ, Parvizi J. Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop 2008;466(7):1710–1715. Phillips JE, Crane TP, Noy M, Elliott TS, Grimer RJ. The incidence of deep prosthetic infections in a specialist orthopaedic hospital: a 15-year prospective survey. J Bone Joint Surg Br 2006; 88(7):943–948.
Control & Hospital Epidemiology:
Email alerts: Click here Subscriptions: Click here Commercial reprints: Click here Terms of use : Click here
Deep Infection after Total Hip Replacement: A Method for National Incidence Surveillance Viktor Lindgren, Max Gordon, Per Wretenberg, Johan Kärrholm and Göran Garellick Infection Control & Hospital Epidemiology / Volume 35 / Issue 12 / December 2014, pp 1491 - 1496 DOI: 10.1086/678600, Published online: 16 January 2015
Link to this article: http://journals.cambridge.org/abstract_S0195941700093887 How to cite this article: Viktor Lindgren, Max Gordon, Per Wretenberg, Johan Kärrholm and Göran Garellick (2014). Deep Infection after Total Hip Replacement: A Method for National Incidence Surveillance. Infection Control & Hospital Epidemiology, 35, pp 1491-1496 doi:10.1086/678600 Request Permissions : Click here
Downloaded from http://journals.cambridge.org/ICE, IP address: 141.209.100.60 on 03 Nov 2015
infection control and hospital epidemiology
december 2014, vol. 35, no. 12
original article
Deep Infection after Total Hip Replacement: A Method for National Incidence Surveillance Viktor Lindgren, MD;1,2 Max Gordon, MD;2,3 Per Wretenberg, MD, PhD;1 Johan Ka¨rrholm, MD, PhD;2,4 Go¨ran Garellick, MD, PhD2,4
objective. The purpose of this study was to estimate the incidence of deep periprosthetic joint infections (PJIs) after primary total hip replacement (THR) in Sweden prior to the introduction of a national initiative to reduce these infections. design.
Prospective open cohort study with 2 years follow-up of each subject.
setting. All THR-performing clinics in Sweden. methods. All patients registered for a primary THR in the Swedish Hip Arthroplasty Register between July 1, 2005, and December 31, 2008, were selected for the study (45,531 patients with 49,219 THRs) and were matched with the Swedish Prescribed Drug Register. All patients with a minimum of 4 weeks of continuous outpatient antibiotic treatment within 2 years after their primary THR (1,989 patients with 2,219 THRs) were selected for a medical records review, and the number of cases with PJI was determined. results. The cumulative incidence of PJI within 2 years after primary THR was 0.9% (95% confidence interval, 0.85–1.02; n p 443), and 405 of these had been reoperated. The incidence rate of PJI for the first 3 months was 5 per 10,000 THR-weeks and, thereafter, 0.3 per 10,000 THR-weeks. Staphylococcus aureus and coagulase-negative staphylococci were the most common bacteria isolated. conclusions. This study describes a new method of national postoperative infection surveillance. The incidence is similar to previous smaller reports and is useful for monitoring changes over time to evaluate the national initiative to reduce infections. Infect Control Hosp Epidemiol 2014;35(12):1491-1496
The third most common complication following total hip replacement (THR) is periprosthetic joint infection (PJI).1 Commonly, it leads to great suffering in the patient, requires large resources of the healthcare system, and is associated with high costs to society, especially if affecting patients of working age.2-4 If diagnosed early and treated aggressively with debridement and irrigation, it may be possible to retain the implant. If discovered late, 1- or 2-stage exchange or resection is often needed.5,6 Regardless of surgical intervention, patients become dependent on antibiotics over a long period of time.7,8 In the literature, the incidence of deep infection following primary THR ranges between 0.2% and 1.6%, depending on follow-up time and the methodology of the study, but until now no study has had a complete national study base.9,10 The diagnosis of postoperative periprosthetic infection can be defined as early (diagnosed within 3 months), delayed (diagnosed between 3 and 24 months), or late (diagnosed after more than 24 months after the primary procedure). The early and delayed infections are often considered to be caused by peri- or intraoperative contamination. Whether the in-
fection is diagnosed before or after 3 months may depend on the aggressiveness of the bacteria.7 Preventive measures might therefore decrease the early and delayed infection rate, but the effectiveness of this work is difficult to evaluate because the overall incidence of infection is low and because of the large number of factors involved.11,12 In Sweden, a nationwide initiative called Prosthesis Related Infections Should Be Stopped (PRISS) was started in 2008 in order to reduce the incidence of PJI following both knee and hip replacement surgery. By 2013, all operating units in Sweden had adopted the pre-, peri-, and postoperative recommendations. The aim of this study was to find the cumulative incidence and incidence rate of early and delayed postoperative deep PJIs following primary THR, a baseline incidence of PJI after primary THR to later evaluate the PRISS initiative.
registries The Swedish Hip Arthroplasty Register (SHAR) was started in 1979, and all private and public orthopedic units in Sweden
Affiliations: 1. Department of Molecular Medicine and Surgery, Section of Orthopaedics, Karolinska Institutet, Stockholm, Sweden; 2. Swedish Hip Arthroplasty Register, Registercentrum Va¨stra Go¨taland, Gothenburg, Sweden; 3. Department of Clinical Sciences, Danderyd Hospital, Karolinska Institutet, Stockholm, Sweden; 4. Department of Orthopaedics, Institute of Clinical Sciences at Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden. Received April 28, 2014; accepted August 1, 2014; electronically published October 24, 2014. 䉷 2014 by The Society for Healthcare Epidemiology of America. All rights reserved. 0899-823X/2014/3512-0008$15.00. DOI: 10.1086/678600
1492
infection control and hospital epidemiology
december 2014, vol. 35, no. 12
voluntarily participate. A total of 98% of all primary THRs in Sweden are reported to the register.13 Apart from information regarding the operating unit, the register holds information including the patient’s personal registration number, age, sex, side, diagnosis, and a number of operative details regarding surgical approach, implant, fixation method, and so on. The SHAR is connected to the Swedish Death Register, thereby also including survival time of implant and patient. The Swedish Prescribed Drug Register (SPDR) was introduced on July 1, 2005. This is a register to which all pharmacies in Sweden are obligated to report and automatically includes all outpatient prescribed drugs dispensed in Sweden. The SPDR includes information regarding the drug, amount prescribed, amount dispensed, date of prescribing and dispensing, instruction from prescribing doctor, level of care and speciality of prescribing doctor, and so on. The SPDR is based on the personal registration number and can therefore be connected to other national healthcare and quality registries in order to evaluate, for example, postoperative infections.14 The SPDR is updated with new prescriptions every month and annually includes about 100 million new prescriptions.
methods The study was given ethical approval by the Regional Ethical Committee in Gothenburg on October 11, 2010 (reference 553-10). All operations between July 1, 2005, and December 31, 2008, reported to the Swedish Hip Arthroplasty Register for a primary THR were included in the study (n p 49,219). All diagnoses, bilateral hip replacements, and types of implants regardless of the method of fixation were included. By using patients’ personal registration numbers, cohorts were then matched with the SPDR for all dispensed antibiotic prescriptions between July 1, 2005, and December 31, 2010. All types of antibiotics with Anatomical Therapeutic Chemical codes J01, J04, and P01 were included (Figure 1). Because current recommendations in Sweden for treatment of PJIs all include long-term antibiotic treatment, and because uncomplicated wound infections were excluded, we limited the search by including only the dispensed amount of antibiotics, suggesting a continuous outpatient medication for at least 4 weeks (Table 1).15 We also limited the observation time to 2 years after the primary THR for each patient to include only early and delayed PJIs. Dispensed antibiotics for which instructions from the prescribing doctor specifically indicated treatment for an infection (eg, urinary tract infection, pneumonia) other than a PJI were excluded. By matching the 2 registries, we found that 1,989 patients with 2,217 THRs had been prescribed and dispensed more than 4 weeks of antibiotic treatment within the first 2 years after the primary THR. A questionnaire for each of the 2,217 THRs, including a list of dispensed antibiotics, was sent to a physician at the primary operating unit (76 different units) to complete and return. In the questionnaire, the receiving physician verified
figure 1. Study flowchart. SHAR, Swedish Hip Arthroplasty Register; SPDR, Swedish Prescribed Drug Register; THR, total hip replacement.
whether the patient had been treated for a deep PJI after the primary THR. If yes, supplementary information on date of diagnosis and how the diagnosis was set—including clinic presentation, laboratory markers, cultures, and radiology findings–was to be filled in as well as whether the patient was reoperated. Finally, the established infecting microorganism, if known, was to be specified. In this study, the diagnosis of deep PJI was established when a patient met 1 or more of the following objective criteria (adapted from the definition established by the Workgroup of the Musculoskeletal Infection Society):16 (1) open sinus tract to the joint; (2) 2 or more positive perioperative cultures of the same pathogen; (3) when the patient met 2 or more of the following criteria: (a) systematic infection and pus in the artificial joint; (b) C-reactive protein greater than 10 or erythrocyte sedimentation rate greater than 30; (c) 1 positive culture from joint fluid aspirate. We excluded superficial infections, PJI after revision surgery, and infections prior to THR (most often THR after failed infected osteosynthesis). Of the 2,217 questionnaires sent out, 2,191 (99%) were returned, and all orthopedic clinics contributed. Four of the patients had incorrectly been registered in the SHAR as primary THR when in fact they had a revision THR and were therefore excluded.
hip arthroplasty: deep infection surveillance
1493
table 1. Tablets/Doses Required for 4 Weeks of Continuous Treatment Antibiotic
ATC code
Daily dose
4-week treatment
Amoxicillin Amoxicillin and clavulanic acid Azitromycin Cefadroxil Cefalexin Ceftriaxon Ceftributen Cefuroxim Ciprofloxacin Ciprofloxacin Ciprofloxacin Daptomycin Ertapenem Erytromycin Fenoximetylpenicillin Flukloxacillin Flukloxacillin Fusidic acid Klaritromycin Klindamycin Levofloxacin Linezolid Lorakarbef Metronidazole Moxifloxacin Norfloxacin Rifampicin Roxitromycin Sulfametoxazol and trimetoprim Teikoplanin Telitromycin Vancomycin
J01CA04 J01CR02 J01FA10 J01DB05 J01DB01 J01DD04 J01DD14 J01DC02 J01MA02 J01MA02 J01MA02 J01XX09 J01DH03 J01FA01 J01CE02 J01CF05 J01CF05 J01XC01 J01FA09 J01FF01 J01MA12 J01XX08 J01DC08 P01AB01 J01MA14 J01MA06 J04AB02 J01FA06 J01EE01 J01XA02 J01FA15 J01XA01
750 mg 1#3 875/125 mg 1#2 250 mg 1#1 1 g 1#2 3 g 1#2 2 g 1#1 400 mg 1#1 250 mg 1#2 250 mg 1#2 500 mg 1#2 750 mg 1#2 400 mg 1#1 1 g 1#2 250 mg 2#2 1 g 2#3 500 mg 2#3 750 mg 2#3 250 mg 2#3 250 mg 1#2 300 mg 1#2 500 mg 1#1 600 mg 1#2 200 mg 1#2 400 mg 1#3 400 mg 1#1 400 mg 1#2 600 mg 1#1 150 mg 1#2 160 mg/800 mg 1#2 400 mg 1#1 400 mg 2#1 1 g 1#2
84 56 28 56 56 28 28 56 56 56 56 28 28 112 168 168 168 168 56 56 28 56 56 84 28 56 28 56 56 28 56 56
note.
ATC, Anatomical Therapeutic Chemical.
statistics Incidence was described as both the 2-year cumulative incidence and the incidence density rate. Each individual primary THR was studied separately (thereby including patients with bilateral THR). The cumulative incidence was calculated by dividing the total number of identified cases within 2 years after the primary THR by the total number of THRs in patients still alive and unrevised (except for infection) at 2 years after the primary operation. The incidence density rate was described by the number of diagnosed infections divided by the number of THRs at risk within 2 periods of time: before 3 months and at 3–24 months. Incidence density rates were expressed per 10,000 THR-weeks, where a THR-week is a unit of follow-up equal to 1 THR followed for 1 week. THRs in patients that died within each time period were censored, and once a THR was infected, it was excluded for further follow-up time. Confidence intervals (CIs) were calculated on the basis of the binomial distribution with the Wilson score interval. The cohort was also divided into half-years on the basis of the operation date for each THR, and cumulative
incidence calculated by each period and trends were tested with the binomial proportion trend test. Time to diagnosis was described as a median. If the date of diagnosis was missing, the first reoperation date was chosen as the date of diagnosis. We used IBM SPSS Statistics software (ver. 21) for Windows.
results A total of 443 THRs were treated for a deep PJI, representing an overall cumulative incidence of 0.9% (95% CI, 0.85–1.02). When dividing the 3.5-year study period into half-years, an increasing incidence was found (Figure 2). Ninety-one percent of the deep PJIs were reoperated (n p 405) as part of the treatment, and 90% (n p 398) were diagnosed according to the specified objective criteria (Table 2). A total of 425 patients had a specified date of diagnosis or reoperation date, and the median time to diagnosis was 24 days (Figure 3). The incidence density rate within the first 3 postoperative months was 5 cases per 10,000 THR-weeks (95% CI, 4.70–5.83), and for the remaining 3–24 months it
1494
infection control and hospital epidemiology
december 2014, vol. 35, no. 12
figure 2. Cumulative incidence (with 95% confidence interval) of deep periprosthetic joint infection after total hip replacement specified for each half-year of the study period.
decreased to 0.3 cases per 10,000 THR-weeks (95% CI, 0.21– 0.31). The cumulative incidence of the individual clinics ranged from 0% to 4.2% (median, 0.8%). In 393 of the 443 deep PJIs, a known pathogen was specified. Coagulase-negative staphylococci and Staphylococcus aureus were alone or in combination with other microorganisms responsible in 70% of the cases (Table 3).
discussion This study presents a unique model to establish an incidence of deep PJI following primary THR by combining health registries with a medical records review. The criteria employed are well defined, and repeated measurements during the follow-up period are possible. As previously mentioned, earlier studies have determined the PJI incidence after primary THR to be between 0.2% and 1.6%. The diagnostic criteria as well as the population studied and the follow-up time are factors that are all likely to significantly influence the result. The incidence of PJI is not always easy to determine in clinical practice, since the predictive power of a single test is low. Many observations taken together—such as clinical presentation, laboratory findings in blood and in joint aspiration, radiographic findings, histology, and, most importantly, cultures—are often used and needed to establish a diagnosis of PJI.17 Cultures indisputably have a central role in the diagnosis
but have considerable risks of being both false positive (contamination) and false negative because of preoperative antibiotic treatment, biofilm formation, poor managed samples, and a lacking standard for cultivation.16,18 The different definitions of a PJI and the historical absence of a standard makes comparisons difficult.19 The American Musculoskeletal Infection Society introduced a new standard in 2011,16 the practical usefulness of which has been subject to debate.20 It accepts the fact that a PJI can be present even though the criteria of diagnosis have not been fulfilled. Therefore, we included all patients where the doctor reported a deep PJI even though the criteria were not fulfilled, since the opposite would definitely lead to an underreporting of PJI incidence. From a patient point of view, it might not always be of importance how the diagnosis has been set but rather the clinical consequences of an infection and its treatment. Although 81 different doctors reviewed the medical records retrospectively, introducing a variability of data quality, we found that in most of the deep PJI cases, there was still enough information in the questionnaire to establish the diagnosis consistent with the adapted specified criteria from the Workgroup of the Musculoskeletal Infection Society (Table 2). The incidence of infections increased during the period of observation, which has been suggested previously.21 Whether this is a result of deterioration of hygiene and prophylactic standards, increasing patient comorbidity, or increasing resistance and virulence of the microorganisms is, however, not answered by this study. Nevertheless, it emphasizes the need for actions such as the Swedish PRISS initiative to reduce postoperative infection rates by improving all aspects of pre-, peri-, and postoperative care. Future studies are planned to determine the effect of this initiative in reducing the national infection burden. The advantages of this study include the nationwide coverage of all THR-performing clinics, the large study population, the long follow-up, and the possibility of an objective assessment of the clinical course after the surgical and medical treatment. Compared with previous studies of national PJI incidence that are based solely on administrative data based on a payment system (Medicare),10,22 our study identified infected patients by their treatment and not by coding of a diagnosis. Data based on International Classification of Diseases codes are imprecise in following specific PJIs because
table 2. Distribution of Diagnosis Criteria Diagnosis criteria
No.
%
1: Open sinus tract to the prosthesis 2: 2 or more positive perioperative cultivations 3a ⫹ 3b: systemic infection and pus in joint ⫹ elevated CRP/ESR 3a ⫹ 3c: systemic infection and pus in joint ⫹ cultivation from joint aspirate 3b ⫹ 3c: elevated CRP/ESR ⫹ cultivation from joint aspirate Clinical diagnosis/information lacking in questionnaire Total
64 263 23 19 29 45 443
14.4 59.4 5.2 4.3 6.5 10.2 100
note.
CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.
hip arthroplasty: deep infection surveillance
figure 3. Histogram illustrating time from primary total hip replacement to diagnosis of deep periprosthetic infection. Each bar represents 1 week.
the codes indicating infection lack accuracy and are not side or joint specific. They can therefore result in overestimation of infections or underestimation, depending on coding routines.22 We believe that the method used in this study therefore results in a more specific and hopefully more accurate estimation of the national incidence of deep PJIs in Sweden. Our study has limitations. Because only outpatient antibiotic treatment is recorded in the SPDR, the patients who died before discharge or those not receiving or complying with the antibiotic treatment would not be included in the questionnaire group. Furthermore, patients never returning to the primary operating clinic would not be recorded as a PJI, even though they might have consumed large doses of antibiotics. However, this group of patients is not expected to be large, especially since we excluded superficial infections. Although the register data are collected prospectively, the diagnosis of PJI was set retrospectively, which is also a possible weakness in this study. With these concerns in mind, we can conclude that the incidence found in this study is probably a slight underestimation of the true incidence. table 3. Distribution of Infecting Microorganism Microorganism
No.
%
Polymicrobiala Coagulase-negative Staphylococcus Staphylococcus aureus Streptococci Enterococcus faecalis Miscellaneous Unknown Total
109 109 89 32 16 38 50 443
24.6 24.6 20.1 7.2 3.6 8.6 11.3 100
a
All including either S. aureus and/or coagulase-negative Staphylococcus.
1495
Many countries have introduced surgical site infection surveillance systems to monitor postoperative infections after specific surgical procedures, often including hip replacement surgery. Perhaps these registries better monitor infection rates, since they can prospectively include more information regarding clinical presentation. However, in practice, it is difficult to ascertain complete registration from all clinics. Because Sweden lacks such a register, we combined information in the existing healthcare and quality registers with a patient’s personal registration number to find information that indicates an infection. Instead of inventing new registries, a model for automated matching of registries in order to find possible postoperative infections is an attractive alternative.23 The SPDR introduces the possibility of an unbiased report of infection incidence, since every dispensed prescription is recorded automatically. The SPDR seems to be sensitive in identifying patients with a possible postoperative infection but is not sufficient to monitor postoperative infections alone, since the population observed have other infections requiring antibiotic treatment. Future studies are planned to investigate how to automatically combine register data in order to increase specificity and maintain a high sensitivity without the need for review of medical records. The microbial spectra with a majority of infections caused by coagulase-negative staphylococci and S. aureus are consistent with previous studies.24,25 The problem with increasing incidence of methicillin-resistant S. aureus (MRSA) infections reported internationally does not seem to have been a clinical problem in Sweden in 2005–2008, since no MRSA infections were reported. In conclusion, this study demonstrates a method where existing registries can be combined in order to identify possible cases of infection and verify them by a medical records review. We believe that this method is close to the optimum way to track all infections in the postoperative period after THR. We found that the incidence of postoperative deep infection following THR was 0.9% between 2005 and 2008. As future changes are anticipated, new data can be collected and analyzed in order to compare with the preexisting situation.
acknowledgments We wish to thank Szilard Nemes for statistical advice and calculations and the Swedish Hip Arthroplasty Register and the Swedish National Board of Health and Welfare for contributing their databases for this research. ¨ F, Sweden. Financial support. Patient Insurance LO Potential conflicts of interest. All authors report no conflicts of interest relevant to this article. All authors submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest, and the conflicts that the editors consider relevant to this article are disclosed here. Address correspondence to Viktor Lindgren, MD, Department of Orthopaedics, Karolinska University Hospital, Solna, SE-171 76 Stockholm, Sweden ([email protected]).
1496
infection control and hospital epidemiology
december 2014, vol. 35, no. 12
references 1. Bozic KJ, Kurtz SM, Lau E, Ong K, Vail TP, Berry DJ. The epidemiology of revision total hip arthroplasty in the United States. J Bone Joint Surg Am 2009;91(1):128–133. 2. Klouche S, Sariali E, Mamoudy P. Total hip arthroplasty revision due to infection: a cost analysis approach. Orthop Traumatol Surg Res 2010;96(2):124–132. 3. Kurtz SM, Lau E, Watson H, Schmier JK, Parvizi J. Economic burden of periprosthetic joint infection in the United States. J Arthroplasty 2012;27(8 suppl):61–65.e1. 4. de Lissovoy G, Fraeman K, Hutchins V, Murphy D, Song D, Vaughn BB. Surgical site infection: incidence and impact on hospital utilization and treatment costs. Am J Infect Control 2009; 37(5):387–397. 5. Koyonos L, Zmistowski B, Della Valle CJ, Parvizi J. Infection control rate of irrigation and debridement for periprosthetic joint infection. Clin Orthop 2011;469(11):3043–3048. 6. Crockarell JR, Hanssen AD, Osmon DR, Morrey BF. Treatment of infection with debridement and retention of the components following hip arthroplasty. J Bone Joint Surg Am 1998;80(9): 1306–1313. 7. Zimmerli W, Ochsner PE. Management of infection associated with prosthetic joints. Infection 2003;31(2):99–108. 8. Giulieri SG, Graber P, Ochsner PE, Zimmerli W. Management of infection associated with total hip arthroplasty according to a treatment algorithm. Infection 2004;32(4):222–228. 9. Urquhart DM, Hanna FS, Brennan SL, et al. Incidence and risk factors for deep surgical site infection after primary total hip arthroplasty: a systematic review. J Arthroplasty 2010;25(8): 1216–1222.e3. 10. Ong KL, Kurtz SM, Lau E, Bozic KJ, Berry DJ, Parvizi J. Prosthetic joint infection risk after total hip arthroplasty in the Medicare population. J Arthroplasty 2009;24(6 suppl):105–109. 11. Adeli B, Parvizi J. Strategies for the prevention of periprosthetic joint infection. J Bone Joint Surg Br 2012;94(11 suppl A):42–46. 12. Jamsen E, Furnes O, Engesaeter LB, et al. Prevention of deep infection in joint replacement surgery. Acta Orthop 2010;81(6): 660–666. 13. Garellick G, Karrholm J, Rogmark C, Herberts P. Swedish Hip Arthroplasty: Annual Report 2010. Gothenburg: Swedish Hip Arthroplasty Register, 2010. http://www.shpr.se/Libraries/ Documents/AnnualReport-2010-2-eng.sflb.ashx. 14. Stridh Ekman G, Ringback Weitoft G, Nyren O, Dickman PW,
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Ericsson O, Struwe J. National surveillance of surgical-site infection through register-based analysis of antibiotic use after inguinal hernia repair. Br J Surg 2010;97(11):1722–1729. Christensson B, Johansson L, Nilsdotter A˚, et al. Va˚rdprogram fo¨r Led-Och Skelettinfektioner. 2008. http://infektion.net/sites /default/files/pdf/Vardprogram_led_och_skelett_2008.pdf. Accessed February 4, 2014. Parvizi J, Zmistowski B, Berbari EF, et al. New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society. Clin Orthop 2011;469(11):2992– 2994. Bernard L, Lubbeke A, Stern R, et al. Value of preoperative investigations in diagnosing prosthetic joint infection: retrospective cohort study and literature review. Scand J Infect Dis 2004;36(6/7):410–416. Malekzadeh D, Osmon DR, Lahr BD, Hanssen AD, Berbari EF. Prior use of antimicrobial therapy is a risk factor for culturenegative prosthetic joint infection. Clin Orthop 2010;468(8): 2039–2045. Parvizi J, Jacovides C, Zmistowski B, Jung KA. Definition of periprosthetic joint infection: is there a consensus? Clin Orthop 2011;469(11):3022–3030. Oussedik S, Gould K, Stockley I, Haddad FS. Defining periprosthetic infection: do we have a workable gold standard? J Bone Joint Surg Br 2012;94(11):1455–1456. Kurtz SM, Lau E, Schmier J, Ong KL, Zhao K, Parvizi J. Infection burden for hip and knee arthroplasty in the United States. J Arthroplasty 2008;23(7):984–991. Phillips CB, Barrett JA, Losina E, et al. Incidence rates of dislocation, pulmonary embolism, and deep infection during the first six months after elective total hip replacement. J Bone Joint Surg Am 2003;85A(1):20–26. Bolon MK, Hooper D, Stevenson KB, et al. Improved surveillance for surgical site infections after orthopedic implantation procedures: extending applications for automated data. Clin Infect Dis 2009;48(9):1223–1229. Pulido L, Ghanem E, Joshi A, Purtill JJ, Parvizi J. Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop 2008;466(7):1710–1715. Phillips JE, Crane TP, Noy M, Elliott TS, Grimer RJ. The incidence of deep prosthetic infections in a specialist orthopaedic hospital: a 15-year prospective survey. J Bone Joint Surg Br 2006; 88(7):943–948.
