Abstract
Background
In patients undergoing total joint arthroplasty (TJA), increasing attention has been directed recently toward identifying specific patient-related risk factors that may predispose patients to periprosthetic joint infection (PJI). Currently, it is unclear whether having a history of a treated native septic arthritis is a risk factor for PJI after TJA in the same joint. Previous studies have reported contradictory evidence and results varied between a substantially higher rates of PJIs to very low or no reported PJIs.
Questions/purposes
(1) What is the risk of PJI in patients who received TJA and had a history of treated same-joint native joint septic arthritis and (2) What are the associated risk factors for these patients developing PJI?
Methods
This was a multicenter retrospective analysis of patients who received primary THA or TKA between January 2000 and December 2016 and who had a history of treated native joint septic arthritis in the same joint. Patients were included in the study only if they were considered to have resolved their joint infection based on a preoperative evaluation that included: (1) the absence of clinical symptoms and signs of active infection or local joint inflammation, (2) recent plain radiographs showing only advanced degenerative changes without evidence of active osteolysis or bone infection, (3) preoperative laboratory investigations for infection, including erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and total leukocyte counts within normal ranges. Patients were reviewed for the occurrence of postoperative PJI. The final cohort included 62 patients who had a mean followup of 4.4 years (range, 3 months–17 years) from the time of TJA. A total of 21 patients (34%) had less than 2 years of followup, including six (10%) mortalities. In total, eight patients (13%) died during the study period, none of which were due to PJI. Patient characteristics, time interval from treatment of septic arthritis to TJA, and Charlson comorbidity index adjusted for age were collected. We used a Kaplan-Meier analysis to estimate the overall survivorship among all TJAs as well as those who underwent THA versus TKA, and we performed a statistical comparison using the Mantel-Cox log-rank test. We performed a Cox regression hazard ratio (HR) survival analysis to identify risk factors for PJI. The PJI odds ratios (OR) for patients who underwent TJA within 2 years of septic arthritis were calculated as an additional temporal analysis.
Results
In patients with a history of treated same-joint native septic arthritis, the proportion of PJI was five of 62 patients (8%). The Kaplan-Meier analysis demonstrated an overall survivorship free from PJI of 92% at 14.5 ± 1.14 years (95% confidence interval [CI] = 12.3–16.8 years). All PJI cases occurred only in patients who underwent TKA, which when analyzed separately, yielded a survivorship of 85% at 10.5 ± 0.9 years (95% CI = 8.7–12.3 years) versus 100% in patients who underwent THA (p = 0.068). Mean time to PJI occurrence was 10 months (range, 2–20 months). After controlling for relevant confounding variables, such as age, sex, affected joint and comorbidities, we found smoking (HR, 8.06; 95% CI, 1.33–48.67; p = 0.023) to be associated with increased risk for PJI development.
Conclusion
Patients with history of native joint infections are at higher risk of PJI, especially smokers. Despite our limitations, this study suggests careful assessment of several other factors in these patients, including allowing a minimum interval of 2 years from the time of resolving native joint septic arthritis to TJA. Patients who are undergoing TKA seem to be more prone to the PJI risk and may benefit from more aggressive planning. In addition, medical optimization of comorbidities that may confer additional risk, such as diabetes, become exceptionally important in these patients.
Level of Evidence
Level III, therapeutic study.
Introduction
Periprosthetic joint infection (PJI) remains a devastating complication after primary total joint arthroplasty (TJA). Recent initiatives have focused on identifying risk factors that may predispose to higher likelihood of PJI. Currently, it is unclear whether having a history of treatment for native septic arthritis is a risk factor for (PJI) after arthroplasty in the same joint. Evidence shows that both PJIs and native joint septic arthritis share multiple predisposing risk factors, including older age; having certain comorbidities, such as diabetes mellitus or immunosuppression; and a history of previous joint disease [5]. However, a clear link between a history of a resolved native joint septic arthritis as a direct risk factor for PJI when primary TJA is performed in the same joint has not clearly been established. Previous studies have suggested contradictory evidence and have had results that varied between reporting substantially higher PJI rates [14] to very low or no reported PJI rates [2, 4, 12]. However, most of these studies mainly aimed at reporting on the outcomes of TJA as a treatment option for patients with persistent or quiescent native joint septic arthritis versus investigating previous native joint septic arthritis as an independent risk factor for developing PJI. In addition, their patient populations were heterogeneous and at times, not limited to having a history of native joint septic arthritis but rather, any infection about the knee including osteomyelitis [14]. Therefore, better evidence exploring this unique relationship is yet to be provided.
Due to the paucity of evidence, we conducted this study and asked: (1) What is the risk of PJI in patients who received TJA and had a history of treated same-joint native joint septic arthritis? (2) What are the associated risk factors for these patients developing PJI?
Patients and Methods
Study Design and Setting
After receiving institutional review board (IRB) approval, in this multicenter retrospective study we evaluated the records of patients who underwent primary hip or knee TJA between January 2000 and December 2016 who also had a prior history of treated hip or knee native joint septic arthritis in the same joint.
Participants/Study Subjects
Patients were included in the study only if they were considered to have resolved their joint infection during routine preoperative evaluation, which included: (1) the absence of clinical symptoms and signs of active infection or local joint inflammation, (2) recent plain radiographs showing only advanced degenerative changes without evidence of active osteolysis or bone infection, (3) preoperative laboratory investigations for infection, including erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and total leukocyte counts within normal ranges.
Demographics, Description of Study Population
The final cohort included 62 patients who had a mean age of 57 years (range, 16–87 years), mean BMI of 29 kg/m2 (range, 18–46 kg/m2), and a mean followup of 4.4 years (range, 3 months–17 years) from the time of TJA. A total of 21 patients (34%) had less than 2 years of followup including six mortalities (10%). In total eight patients (13%) died during the study period, none of which were due to PJI. There were 35 males and 27 females. A total of 33 patients underwent TKA and 29 underwent THA. On average, patients had a mean interval to TJA of 3.7 years from their treated native joint septic arthritis (range, 3 months–56 years). In all, 60 patients underwent surgical treatment in the form of open or arthroscopic débridement to treat the native joint septic arthritis while two patients were treated by intravenous antibiotics only. All patients received a minimum of 4 weeks of intravenous antibiotics. A one-stage procedure was performed in eight THAs and 21 TKAs, while two-stage procedures were performed in 15 THAs and 12 TKAs (Table 1). Two-stage revisions were performed in patients who required multiple débridement procedures to control their native joint septic arthritis. Standard technique was used, consisting of cement spacer implantation for 6 to 8 weeks followed by definitive prosthetic implantation. For all patients, we collected patient characteristics, time interval from treatment of septic arthritis to TJA, American Society of Anesthesiologists Physical Status scores (ASA), and Charlson comorbidity index adjusted for age.
Table 1.
Surgery-related variables for all patients
Variables, Outcome Measures, Data Sources, and Bias
We recorded the occurrence of postoperative PJI diagnosis after TJA using the MSIS criteria [13] in each patient. At the final followup, we compared the following factors among patients who developed a PJI versus those who did not: age, sex, body mass index (BMI), Charlson comorbidity index adjusted for age, ASA scores, smoking status, current history of diabetes mellitus, time interval between native joint septic arthritis treatment and TJA, the use of antibiotic-loaded cement during TJA, and status of the organism that caused the native joint septic arthritis (Staphylococcus aureus versus non-S. aureus). Positive smoking status was recorded for any patient with current tobacco use regardless of the amount. Diabetes status was also recorded as a dichotomous variable (having diabetes versus not having diabetes) if the patient was receiving any medical management for previously diagnosed diabetes (insulin or non-insulin dependent) during the study period. The use of antibiotic loaded-cement was indicated at our institutions for high-risk patients, including those with history of native joint septic arthritis. However, this was not consistently applied to all patients as some patients demonstrated higher clinical risk than the others, such as those with a shorter interval from native joint septic arthritis, equivocal ESR and CRP, organism virulence, etc.
Statistical Analysis, Study Size
For PJI risk, we used descriptive statistics to report on the PJI rate among all patients. In addition, we used the Kaplan-Meier analysis to estimate the overall survivorship among all TJAs as well as those who underwent THA versus TKA; we performed a statistical comparison using the Mantel-Cox log-rank test. We then performed a Cox regression hazard ratio (HR) survival analysis to identify independent risk factors for PJI among those who encountered the infection versus those who did not. PJI odds ratios (OR) for patients who underwent TJA within 2 years of septic arthritis were calculated as an additional temporal analysis. All statistical analyses were performed using SPSS version 24 (IBM Corp, Armonk, New York, USA). We determined that a cut off point of ≤ 0.05 was the point of statistical significance for the analysis.
Results
In patients with a history of treated same-joint native septic arthritis, the proportion of PJI was five of 62 patients (8%). The Kaplan-Meier analysis demonstrated an overall survivorship free from PJI of 92% at 14.5 ± 1.14 years (95% confidence interval [CI] = 12.3–16.8 years). All PJIs occurred only in patients who underwent TKA, which when analyzed separately yielded survivorship of 85% at 10.5 ± 0.9 years (95% CI = 8.7–12.3 years) (Fig. 1) versus 100% in patients who underwent THA (p = 0.068) (Fig. 2). Mean time to PJI occurrence was 10 months (range, 2–20 months). In all patients who developed PJI, routine intraoperative cultures were negative. Additionally, in three patients the same organism from native joint septic arthritis was cultured in PJI (S. aureus) while different organisms were isolated in the other two patients (Viridans streptococci to Candida albicans and pneumococci to S.aureus). In the three patients who had S. aureus infection, methicillin-sensitive S.aureus (MSSA) was isolated. This is in addition to another patient who developed PJI due to MSSA but initially had pneumococci cultured during his native joint septic arthritis (Table 2).
Fig. 1.
The Kaplan-Meier survivorship curve for entire cohort is shown here.
Fig. 2.
This Kaplan-Meier survivorship curve shows the comparison of total knee versus hip arthroplasty patients.
Table 2.
Summary of characteristics of patients who developed periprosthetic joint infection
After controlling for relevant confounding variables, such as age, sex, affected joint, comorbidities, we found smoking (HR, 8.06; 95% confidence interval (CI), 1.33–48.67; p = 0.023) to be associated with an increased risk for PJI development (Table 3). There were no associations between PJI risk and age (HR per year increase, 1.003; 95% CI, 0.939–1.072; p = 0.925), sex (female HR compared with male, 0.015; 95% CI, 0.000–20.76; p = 0.255), or BMI (HR, 0.99; 95% CI, 0.85–1.15; p = 0.927), joint affected (knees HR compared with hips, 46.86; 95% CI= 0.28–79662.16; p = 0.311). Patients with diabetes did not show higher risk (HR= 6.02; 95% CI = 0.995–36.517; p = 0.051). Patients who had TJA within 2 years from their treated septic arthritis compared with patients with more than 2 years interval showed OR of 3.022 (95% CI = 0.156–58.4347). However, this was not found to be statistically significant (p = 0.464).
Table 3.
Cox regression hazard ratio survival analysis for risk factors of periprosthetic joint infection in patients who had prior same-joint septic arthritis at mean followup of 4.4 years
Discussion
After primary TJA, PJI occurs in 1% to 2% of patients [8]. Currently, there is a lack of consensus as to whether native joint septic arthritis carries an increased risk of developing PJI, even when the TJA procedure is performed years after the primary infection has been eradicated. Therefore, in this study we aimed to quantify the risk in these patients and characterize the associated risk factors. We demonstrated that the risk of PJI was much higher in patients with history of native septic arthritis than the current reported rates after primary THA or TKA. This was demonstrated by the higher PJI rate (8%) and the relatively low survivorship at mean followup. In addition, current smoking status seemed to carry an additional independent risk in these patients.
There are several limitations to this study. Although our cohort had a mean followup of 4.4 years, our minimum followup was very short (3 months) and several patients did not reach the 2-year followup point. Patients who are lost to followup may have had complications that were not captured in our analysis, including PJI. Therefore, our risk estimation of PJI in this specific patient population may not be accurate, and our findings are more likely to underestimate the risk of PJI in these patients. Another important shortcoming is that our findings may exhibit some of the signs of sparse-data bias [9] due to the inherently underpowered study and small number of recorded events of interest. This has reflected on some of the recorded variables in our analysis demonstrating very wide confidence intervals, even for those that reached the point of statistical significance (CI for smoking HR ranges from just over 1 to nearly 50). A CI of such wide range suggests that even a few patients with different outcomes, such as a few more smokers who later developed PJI, could have substantially changed our findings. The same can be true for other variables included in the analysis. Another limitation is that some of the studied risk factors have been only analyzed as dichotomous variables. For example, we recorded having diabetes as yes or no status regardless of the diabetes subtype and control. Hence, significant details pertaining to the effect size of these variables were not considered. Additional limitations related to the retrospective nature of the study include a lack of analysis of other important variables, including rigorous preoperative confirmation of infection resolution (using aspiration, MRI, MRSA screening, and nuclear scans) and correlation of risk with intraoperative culture results. Also, in several patients, native joint septic arthritis was treated using both open and arthroscopic approaches, which limited our ability to study the relationship between each of the approaches and PJI risk after TJA. For TJA, our sample was heterogeneous and both one- and two-stage primary procedures were performed. Antibiotic-loaded cement was also used inconsistently in our patients. However, the main aim of the study was to investigate the risk of infection versus evaluating all the details of TJA failure in those patients.
We found the survivorship free from PJI after treated native joint septic arthritis in the replaced joint to be 92% at 14.5 ± 1.14 years; given that 21 patients were lost to followup before 2 years, this estimate is almost certainly conservative, and the real risk likely is worse, perhaps considerably so. We are aware of only three studies that have explored this topic. One small study [12] compared reinfection and occurrence of PJI in 18 patients who had hip native septic arthritis and were treated with THA using single- (n = 11 patients) and two-stage approaches (n = seven patients) compared with a control cohort of 18 patients who underwent routine primary THA with mean followup of 70 months. They reported no infections in either cohort. However, their conclusions remain highly limited by the inclusion of patients who underwent two-stage procedures, the smaller sample size, and the retrospective analysis. In addition, in their analysis, they did not aim to directly investigate the history of a treated same-joint native septic arthritis as a risk factor for PJI development after TJA and mainly aimed to assess staging THA as a treatment option for native joint septic arthritis, which may imply additional level of selection bias. Nevertheless, they found no PJIs among patients who underwent THA, a similar finding to our present study, which may suggest that the PJI risk is lower in this patient population compared with TKA. A similar study that reviewed 14 patients with either native joint septic arthritis or osteomyelitis about the knee reported no PJI in the cohort after two-stage TKA at mean followup of 4.5 years [11]. In another retrospective case series study [14], the authors reported a similarly high PJI risk, with 9.7% at mean followup of 6.1 years, which is similar to the present study. The results from Seo et al. [14] are limited by the fact that their sample lacked adequate homogeneity because the authors also included patients who were treated for osteomyelitis, which may be a confounding variable. In their analysis, the number of previous surgeries performed to treat the native joint septic arthritis was the only independent risk factor associated with developing PJI.
Patients who were smokers at the time of TJA had higher risk for PJI. It is well-established in the current evidence that smoking is associated with worse postoperative outcomes in general, including postoperative infections [15]. Previous studies suggested that nicotine-induced local tissue vasoconstriction leads to tissue hypoxia and poor healing that predisposes to infection [6, 7]. Azodi et al. [1] also identified smoking as one of the independent risk factors for PJI after primary THA in their retrospective analysis of 3309 patients. Our study has shown that smoking has an additional implicated risk in patients with history of native joint septic arthritis who undergo TJA, which is a novel finding to the present evidence.
Other factors did not statistically correlate with an increased risk of PJI but may point to a trend toward increasing risk. This included increasing Charlson comorbidity index adjusted for age scores, diabetes, and worse ASA status. Several studies have indicated that inadequate glycemic control, multiple comorbidities and poor nutritional status are associated with increased PJI risk [3, 5, 10]. Additionally, patients who had an S. aureus infection causing their native joint septic arthritis and those who had infection of their native knees, showed a trend towards greater PJI risk. On the other hand, the interval time from native joint septic arthritis treatment to TJA inversely correlated with PJI risk. In our experience, all patients who developed PJI contracted the infection less than 2 years from their TJA. In addition, only TKA patients developed PJI compared with those who had THA, which may point to a different risk profile that requires further analysis.
In conclusion, the present study demonstrated that patients who had a history of treated same-joint native septic arthritis had a high PJI risk after TJA. Despite our aforementioned limitations of the present study, including in particular sparse-data bias, we believe that we reached several meaningful conclusions that deserve consideration in clinical practice. In patients with history of a treated native joint septic arthritis, surgeons should be cognizant about several factors before offering TJA, particularly in smokers. Ideally, a minimum interval of 2 years should be allowed from the time of native joint septic arthritis resolution to the time of TJA. Patients who are undergoing TKA seem to be more prone to the PJI risk and may benefit from more aggressive planning. Lastly, medical optimization of comorbidities that may confer additional risk, such as diabetes, becomes exceptionally important in these patients Meanwhile, future research should include more comprehensive and detailed analyses of implicated demographic risk factors, the potential for a negative linear relation between interval from native joint septic arthritis to TJA and PJI risk, and compare the findings among hips versus knees in a robust patient sample.
Footnotes
One of the authors certifies that he (CAH) or a member of his immediate family has received or may receive payments or benefits, all outside the submitted work, in an amount of less than USD 10,000 from KCI (San Antonio, TX, USA); an amount of less than USD 10,000 from Zimmer Biomet (Warsaw, IN, USA); an amount of less than USD 10,000 from Covance (Princeton, NJ, USA); an amount of less than USD 10,000 from Pfizer (New York, NY, USA); an amount of less than USD 10,000 from TenNor Therapeutics Limited (Suzhou Industrial Park, China); and has received research support funding from Stryker (Kalamazoo, MI, USA), Myoscience (Fresno, CA, USA), CD Diagnostics (Claymont, DE, USA), KCI (San Antonio, TX, USA), OREF (Rosemont, IL, USA), Pacira (Parsippany, NJ, USA), 3M (Maplewood, MN, USA), Cymedica (Scottsdale, AZ, USA), Ferring Pharmaceuticals (Parsippany, NJ, USA), and Orthofix, Inc. (Lewisville, TX, USA).
All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.
Clinical Orthopaedics and Related Research® neither advocates nor endorses the use of any treatment, drug, or device. Readers are encouraged to always seek additional information, including FDA approval status, of any drug or device before clinical use.
Each author certifies that his or her institution approved or waived approval for the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.
This work was performed at the Cleveland Clinic, Cleveland, OH, USA.
References
- 1.Azodi OS, Bellocco R, Eriksson K, Adami J. The impact of tobacco use and body mass index on the length of stay in hospital and the risk of post-operative complications among patients undergoing total hip replacement. J Bone Joint Surg Br . 2006;88–B:1316–1320. [DOI] [PubMed] [Google Scholar]
- 2.Bauer T, Lacoste S, Lhotellier L, Mamoudy P, Lortat-Jacob A, Hardy P. Arthroplasty following a septic arthritis history: A 53 cases series. Orthop Traumatol Surg Res . 2010;96:840–843. [DOI] [PubMed] [Google Scholar]
- 3.Berbari EF, Osmon DR, Lahr B, Eckel-Passow JE, Tsaras G, Hanssen AD, Mabry T, Steckelberg J, Thompson R. The Mayo prosthetic joint infection risk score: implication for surgical site infection reporting and risk stratification. Infect Control Hosp Epidemiol . 2012;33:774–81. [DOI] [PubMed] [Google Scholar]
- 4.Farrell MJ, Bryan RS. Total knee arthroplasty after septic arthritis. Orthop Clin North Am . 1975;6:1057–62. [PubMed] [Google Scholar]
- 5.Kapadia BH, Berg RA, Daley JA, Fritz J, Bhave A, Mont MA. Periprosthetic joint infection. Lancet. 2016;387:386–394. [DOI] [PubMed] [Google Scholar]
- 6.Kapadia BH, Issa K, Pivec R, Bonutti PM, Mont MA. Tobacco use may be associated with increased revision and complication rates following total hip arthroplasty. J Arthroplasty. 2014;29:777–80. [DOI] [PubMed] [Google Scholar]
- 7.Kapadia BH, Johnson AJ, Naziri Q, Mont MA, Delanois RE, Bonutti PM. Increased revision rates after total knee arthroplasty in patients who smoke. J Arthroplasty. 2012;27:1690–1695.e1. [DOI] [PubMed] [Google Scholar]
- 8.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;23:984–991. [DOI] [PubMed] [Google Scholar]
- 9.Leopold SS, Porcher R. Editorial: sparse-data bias-what the savvy reader needs to know. Clin Orthop Relat Res . 2018;476:657–659. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Marchant MH, Viens NA, Cook C, Vail TP, Bolognesi MP. The impact of glycemic control and diabetes mellitus on perioperative outcomes after total joint arthroplasty. J Bone Joint Surg. Am. 2009;91:1621–9. [DOI] [PubMed] [Google Scholar]
- 11.Nazarian DG, de Jesus D, McGuigan F, Booth RE. A two-stage approach to primary knee arthroplasty in the infected arthritic knee. J Arthroplasty. 2003;18:16–21. [DOI] [PubMed] [Google Scholar]
- 12.Papanna MC, Chebbout R, Buckley S, Stockley I, Hamer A. Infection and failure rates following total hip arthroplasty for septic arthritis: a case-controlled study. Hip Int . 2018;28:63–67. [DOI] [PubMed] [Google Scholar]
- 13.Parvizi J, Zmistowski B, Berbari EF, Bauer TW, Springer BD, Della Valle CJ, Garvin KL, Mont MA, Wongworawat MD, Zalavras CG. New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society. Clin Orthop Relat Res . 2011;469:2992–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Seo J-G, Moon Y-W, Park S-H, Han K-Y, Kim S-M. Primary total knee arthroplasty in infection sequelae about the native knee. J Arthroplasty. 2014;29:2271–2275. [DOI] [PubMed] [Google Scholar]
- 15.Sørensen LT. Wound healing and infection in surgery. The clinical impact of smoking and smoking cessation: a systematic review and meta-analysis. Arch Surg . 2012;147:373–83. [DOI] [PubMed] [Google Scholar]