Prosthetic joint infection in a Tunisian tertiary care teaching hospital

Background: Prosthetic joint infections are a serious problem both in low and high income countries. Their prevention measures should be targeted by knowing the risk factors. Our study aimed to determine the incidence and current risk factors of prosthetic joint infections in our organization. Methods: We reviewed prospectively collected data from our database on 340 patients undergoing primary and revision hip or knee and hip arthroplasty between January 2010 and December 2014. Results: Twelve patients with infections were identified (3.5%). Infection rates were 6.5% and 1.5% after hip and knee arthroplasty respectively. The most common organism identified was Staphylococcus aureus (6 cases). Logistic regression analysis identified three variables independently associated with prosthetic joint infection: rheumatoid arthritis (Relative Risk [RR] = 44.90; 95% confidence interval (95% CI): [5.54–333.0]; p <0.001) remote infections (RR = 33.17; 95% CI: [2.57–528.7]; P = 0.007) and revision (RR = 11.57; 95% CI: [1.84–71.6]; p = 0.009). Conclusion: Our rate seems to be higher than those reported in the literature. To reduce this rate, we must take actions targeting identified risk factors. Correspondence to: Mohamed Ben Rejeb, Centre Hospitalo-Universitaire Sahloul Route ceinture Sahloul 4054 Sousse, Tunisie, Tel: +21654098709, Fax: +21673367451, E-mail: mohamed_benrejeb@yahoo.fr ; mohamed.benrejeb@rns.tn


Introduction
Total knee (TKA) and hip arthroplasties (THA) are the most common types of prosthetic surgery. Prosthetic joints improve the quality of life, but they may fail, necessitating revision or resection arthroplasty [1]. Infection is the most serious complication occurring in 0.8 to 1.9% of knee arthroplasties [2][3][4] and 0.3 to 1.7% of hip arthroplasties [4][5][6]. Prevent this infection remains the most formidable weapon and depends on a better understanding of risk factors.
Since of decades, our country is experiencing a demographic and epidemiological transition. These transitions were characterized by aging population, sedentary, smoking and lifestyle changes. Thus, degenerative diseases increased and joint arthroplasties became increasingly popular activities. To our knowledge, few Tunisian studies have estimated the incidence of prosthetic joint infection (PJI). Furthermore, we do not have figures reflecting the seriousness of the problem in our hospital. An evaluation of arthroplasty activities was required. Therefore, our study aimed to determine the incidence and current risk factors of PJI associated with TKA and THA.

Setting
Sahloul hospital is a tertiary care academic medical center in the upper Mideast of Tunisia; with 630 beds. More than 220.000 patients per year are cared for at this center. Department of Orthopedic surgery capacity was 59 beds with 5116 admissions in 2010.

Study design
It was a retrospective study. All primary and revision TKA and THA that were performed in department of orthopedic surgery of Sahloul Hospital from January 2010 to December 2014 were included and reviewed in the study. A 12-month follow-up was conducted in all cases.

Definition of prosthetic joint infection
All cases of potential infection were identified from positive cultures in the microbiology laboratory and were cross-referenced to patients who had undergone a hip or knee replacement at the hospital. Deep infection was defined as infection within the prosthetic cavity, whereas superficial infection involved only the wound without extension deep to the fascia [7]. Infection was diagnosed per the following criteria: two or more positive culture taken from inside a prosthetic hip or knee joint, radiological or hematological evidence of infection and clinical features consistent with infection (e.g. pain. fever. restricted movement of the joint. sinus. discharge).

Potential risk factors
The medical records of patients were abstracted for data on potential risk factors: age, gender, hypertension, diabetes mellitus, underlying native joint disease (degenerative diseases, rheumatoid rheumatoid arthritis, infection…), remote infection (Any other infection identified in preoperative period (e.g. gynecological. skin. urinary…)), immunosuppression (It is neutropenia with a rate of circulating polymorph nuclear <500 cells / mm 3 or another type of immunosuppression by treatments (chemotherapy. radiotherapy. immunosuppressant, corticosteroids long-term or high-dose recently) or by disease (leukemia. lymphoma. AIDS)). American Society of anesthesiologists (ASA) score, National Nosocomial Infections surveillance (NNIS) index, preoperative length of stay, antibioprophylaxis, emergency surgery and type of arthroplasty. The presence of any other risk factor was documented by the clinicians caring for the patient at the time of prosthesis implantation.

Data collection and statistical analysis
The medical records of all patients were abstracted by using a standardized data collection tool for demographic information and data related to host and index arthroplasty risk factors of PJI. The differences and similarities between categorical groups were compared using the chi-squared test (or Fisher exact tests where appropriate). Continuous variables were compared using the student t-test. Univariate and then multivariate regression analysis were used to examine the associations of various risk factors and PJI between patients with and without PJI. Multiple logistic regression utilizing a stepwise method of model was performed including only variables with p ≤ 0.20 in univariate regression. All analyses were conducted using SPSS version 19.0 (IBM Corporation. Armonk. NY) with significance set at the 5% level.

Discussion
Over the past three decades, much effort has been expended on reducing the incidence of PJI. However, recent studies reported high rates and still emphasize the acuteness and significance of this health problem [8,9]. Thus, several studies were carried out to examine the incidence of infections after THA and TKA and determine their risk factors. It is admitted that the rate of these infections depends on the activities of the center, monitoring mode, sample size and underlying diseases. Moreover, direct comparisons between different studies are not reliable and easy because of considerable difference in methodology, definitions of PJI and type of prosthesis used [10].
Our rates were higher than those reported by several studies. The lowest incidences have been reported in Scandinavian countries. Infection occur in 0.3 to 0.6% of patients with THA [10,11] and 0.52 to 1% of patients undergoing TKA [10,12]. Low rates were also found in United Kingdom [6], Germany [13], USA [4,8] The majority of these studies were based on epidemiological surveillance programs or registries already in place [2,6,[10][11][12]14]. Even such routine surveillance   [16]. Finally, the results of studies estimating its incidence have been conflicting [17]. Moreover, there are no standard criteria available to how infection is both defined and identified. Isolating a microorganism on tissue biopsies cultures was regarded as gold standard [18]; however, certain authors have instead used histological criteria [19] and it is well known that there are false negatives among the culture results.
In present study, 58% of isolated micro-organisms were Gram Positive Cocci. Of them, Staphylococcus aureus was the most common. It was resistant to methicillin (MRSA) in 33% of cases. Our results are consistent with those reported in the literature. However, our proportion of MRSA was significantly higher. It has been reported that many bacteria can cause PJIs, but Staphylococcus aureus, including MRSA, and coagulase-negative staphylococci (CNS) such as Staphylococcus epidermidis were the most common infecting organism [3,20,21] Rheumatoid arthritis, remote infection and revision were the independent risk factors of PJI in our study. Since the 80s, several studies incriminated inflammatory arthritis especially rheumatoid arthritis in the genesis of infection after THA or TKA [26][27][28][29]. Rheumatoid arthritis trend to increase from 2.5 to 4 times the risk of developing this infectious disease [30,31]. Positive association could be related to the defect in the immune system caused by this underlying disease and its treatment. Corticosteroids are known to increase the risk of such infection [32,33] but the effect of new anti-rheumatoid therapy not been clearly elucidated [34,35]. The second patient related factor in our study was active infections identified in preoperative period. Any remote infections especially those affecting the skin, genital, unary tract and oro-rhino-laryngeal sphere increase the risk of PJI through bacteremia [36]. Thus, remote infection sources must be detected and eradicated before surgery.
On the other hand, revision was the operation-related risk factor in present study. Hanssen et al. [37] concluded that the rates of PJI were higher after revision than after primary TKA and THA. While between 1.5% and 2.5% of patient undergoing primary joint replacement developed PJI, the infection rates were 3.2% and 5.6% after revision for THA and TKA respectively [37]. High risk of infection could be related to the quality of joint aspiration fluid procedure. Joint aspiration prior to revision arthroplasty is widely utilized. For knee patients the procedure is comparatively straightforward, whereas hip aspiration may impose a higher risk of iatrogenic infection [38]. Thus, it is often argued that invasive diagnostic samples from hip patients should be obtained only if a there is a high probability of infection [39,40].
Present study has a number of limitations. We did not include enough infected patients and our study was mono-centric. Thus, the results should not be generalized to other settings. The choice of 12 months' follow-up could underestimate the incidence rate of PJI in our study. In fact, late infections caused by slow multiplication microorganisms and occurred beyond this deadline may not be counted. Unfortunately, due to retrospective design of our study, it was not possible to assess the role of all risk factors and collect data was limited to those found in patients' records. Many potential confounding factors of comorbidities such as obesity, malignancy, malnutrition, smoking, alcohol, blood transfusion and operating room seems to be insensitive in detecting PJIs [15]. Thus, combining data from different sources and a manual review of all identified patient records probably leads to the most reliable recording of infectious complications and helps to classify them correctly [14]. Earlier studies that base their infection rates on routine recording of septic complications may therefore have underestimated the true infection burden [12,14]. Moreover, Well-executed studies examining factors associated with infection after joint replacement are difficult to conduct. Prospective studies with meticulous collection of data have problems in achieving sufficient numbers to be effectively statistically powerful because infection is a rare event [16]. Large registry-based studies often rely on readmission or re-operation as a surrogate measure of infection and hence miss numerous infections successfully treated conservatively  environment were not studied. These factors have been suggested to be associated with an increased risk of PJI.

Conclusion
In present study, infection rate seems to be higher than those reported by several studies. This confirms that PJI is a public health issue in developing countries such Tunisia and leads us to establish as soon as possible a strategy to promote joint replacement activities in our hospital. The utility of this study is to target our improvement actions. Those actions should be interested patients with Rheumatoid arthritis or remote infection and those undergoing a revision.