Occurrence, risk factors, and microbiology of surgical site infections after total knee arthroplasty: preliminary results of a retrospective study

To the Editor: Total knee arthroplasty (TKA) has became an established treatment for end-stage joint disease. In China, there was a 5.9-fold increase in the number of TKA cases between 2011 and 2019. All kinds of infections are catastrophic complications of TKA. Recent reports have highlighted the underestimated prevalence of surgical site infections (SSIs) and projected an increase in complex SSIs following hip and knee arthroplasties by 14% between 2020 and 2030.[1] However, there are few reports describing SSIs in Chinese population. This preliminary retrospective study aimed to gain some insights into the occurrence, risk factors, and microbiological patterns of SSIs in China. This study was approved by the Institutional Review Board at Peking Union Medical College Hospital under Protocol Number S-K1728. Written informed consent was obtained from all included patients. Patients underwent primary TKAs by the authors' surgical team at a tertiary medical center between January 2009 and September 2021 were included. All procedures were performed following conventional protocols. The medical records of the included patients were carefully reviewed according to the definitions of SSIs from National Healthcare Safety Network (NHSN). The longest SSI surveillance period for operations involving prostheses was 90 days; patients without 90-day follow-up records were excluded. Following factors were extracted from medical records: gender, age, preoperative comorbidities (autoimmune-related diseases, diabetes mellitus, and hemophilia), and surgical procedure. Surgical procedures were sorted into single (primary unilateral TKA) and complex TKA (bilateral TKA or unilateral TKA with other procedures, such as Achilles tendon lengthening, in the same anesthesia process). Patients with SSI collected microbiological tests data. Statistical analysis was performed using SPSS version 19.0 (IBM Corporation, Armonk, NY, USA). Numerical variables were expressed as median and interquartile range (IQR), whereas categorical variables were expressed as frequency and percentage. The incidence of SSI was defined as the percentage of patients who fulfilled the NHSN criteria for SSIs among all included patients. Extracted data were subjected to logistic regression analysis. First, univariate test was performed using the chi-squared test (Fisher's precision probability test when necessary) or Student's t-test (Mann–Whitney U test for non-normally distributed numerical variables) to screen the variables. Variables that achieved statistical significance in the univariate analysis—defined as P <0.05—were included in the binary logistic regression model. The number of SSI patients required for logistic regression model was based on the event per variable (EPV) method. For the Wald method based on maximum likelihood estimation, the EPV should not be <10 to preserve the stability of the regression. However, the EPV is only a "rule of thumb" and statisticians believe that an EPV <10 is acceptable. The minimum standard was EPV = 5. The distribution of microbes was assessed using descriptive analysis. A total of 1447 patients (2084 TKAs) were included in this study. All patients have follow-up records for 90 days and none were excluded. According to NHSN criteria, 14 patients (17 TKAs) were diagnosed with SSI, corresponding to an overall incidence rate of 0.97% (14/1447). The median duration of postoperative SSI time was 12 days (IQR: 1–36 days). According to definitions from NHSN, SSIs were classified into three types: superficial (involving skin and subcutaneous tissue), deep (involving deep soft tissue, such as fascia and muscle layer), and organ/space (involving any part of the anatomy that was opened or manipulated during surgery). Among 14 cases with SSI, seven were superficial, three were deep incisional, and four were organ/space. The NHSN defines periprosthetic joint infection (PJI) as a severe organ/space SSI meeting specific standard. All four organ/space SSI in the present study fulfilled the criteria for PJI and underwent surgical treatment. In the univariate analysis of extracted factors, statistical significance was found between SSI occurrence and age (P = 0.030), autoimmune-related disease(s) (P = 0.009), and hemophilia (P = 0.049). Forward stepwise logistic regression discarded age (P = 0.359) and preserved the other two factors. The results suggested that patients with autoimmune-related diseases had an approximately seven times higher risk for SSI (odds ratio [OR]: 7.190, 95% confidence interval [95% CI]: 2.166–23.865]; P = 0.001). Patients with hemophilia have a higher risk in SSI (OR: 8.106, 95% CI: 1.747–28.910; P = 0.006) [Table 1]. Table 1 - Risk factors for SSI and results of univariable and multivariable analysis. Factors Number of all patients Number of SSI patients SSI rate (%) Univariable analysis Multivariable analysis P value OR (95% CI) P value Gender 0.337 NA NA Male 338 5 1.48 Female 1109 9 0.81 Age (years) 66 (60–72)* 62 (50–69)* N/A 0.030 NA NA Autoimmune-related diseases 145 5 3.45 0.009 7.190 (2.166, 23.865) 0.001 Hemophilia 88 3 3.41 0.049 8.106 (1.747, 28.910) 0.006 Diabetes mellitus 166 2 1.20 0.670 NA NA Surgical procedure 0.435 NA NA Single TKA 776 6 0.77 Complex TKA 671 8 1.19 *The numerical variables were summarized as median and IQR. CI: Confidence interval; IQR: Interquartile range; OR: Odds ratio; SSI: Surgical site infection; TKA: Total knee arthroplasty. NA: Not applicable. Among 14 cases, only one lacked microbiological test records and eight of remaining patients exhibited positive results. Most pathogens were detected in incisional secretions or intra-articular fluid. The most common pathogen was Staphylococcus, which was confirmed in seven patients. The main targets of antibiotic resistance were β-lactams, macrolides, and cephalosporins. Multisite or multispecies infections are common among patients with SSI after TKA. The worst condition observed in the present study involved a 55-year-old woman, in whom with three types of pathogens in sputum sample and four types in the incisional secretion sample. Candida albicans and Pseudomonas aeruginosa were found both in sputum and incisional secretions, while methicillin-resistant Staphylococcus aureus was detected in the incisional secretions, drainage fluid, and venous blood. To control the infection, the patient received three courses of vacuum sealing drainage and implantation of an antibiotic cement spacer. Four years later, an above-knee amputation of the right side became the option of last-resort for uncontrolled mixed infection. In our study, the occurrence of SSI was 0.97%, which was lower than that of a previous report (2.03%) based on 986 cases and 1-year follow-up.[2] In 2010, NHSN reduced the surveillance period to 90 days. Studies using the "old" criteria always had a higher incidence. A review of 7737 TKAs conducted in Canada reported an SSI incidence rate of 1.38% within one year.[3] However, not all studies with shorter follow-up periods reported a lower occurrence. Słowik et al[4] reported an SSI incidence of 1.9% according to the new criteria. A cohort study from Singapore reported a 1.10% incidence of SSI within two weeks after TKA.[5] Under the same criteria, SSI occurrence also varies among different countries; therefore, geographical region or ethnicity may be factors influencing the occurrence of SSIs and more evidence from multicenter studies is required. Autoimmune-related diseases and hemophilia were screened using logistic regression as risk factors for SSI after TKA. Although glucocorticoids have immunosuppressive effects, they also usually used in patients with autoimmune diseases. Some studies highlighted the risk for infection with glucocorticoid exposure. In patients with hemophilia, the increased risk for SSI could be explained by coagulation deficiency. All three patients with SSI in the present study had hemophilia A, two with moderate coagulation deficiency (coagulation factor VIII activity was 1.5% and 1.6%), and one with severe deficiency (coagulation factor VIII activity was 0.5%). Although all patients underwent appropriate factor replacement under the guidance of a hematologist which could prevent hemorrhage, incision healing in hemophiliacs is prolonged compared to that of non-hemophiliacs. Our previous follow-up of patients with hemophilia who underwent TKA without infection demonstrated this tendency. However, the intrinsic relationship between hemophilia and infection remains unclear. Some cases of TKA infection had obvious clinical preferences, but negative microbiological test results. Culture-negative rate of PJI varies from 7% to 50%. Our study reported a negative rateio of 5/13 for patients who experienced SSI. The explanations lie in the application of antibiotics before sample collection, the weakness of the culture technique, and the characteristics of the pathogen. The most common pathogen among cases of SSI in the present study was Staphylococcus, which is also common in PJI. These two types of infections share the same antibiotic prophylaxis strategies. Owing to the popularity of antibiotics in infection prophylaxis and treatment, drug resistance among pathogens is growing rapidly, and the majority of detected pathogens in the present study were resistant to more than one antibiotic, including one gentamycin-resistant sample. Gentamycin is widely used in antibiotic-loaded bone cement and orthopedic surgery. Drug resistance will become a long-term challenge in infection control. As a preliminary study of SSI after TKA, some limitations should be addressed. First, the sample size was small and does not fully satisfy the EPV requirement; the results arise not sufficiently robust. The range between the upper and lower limits of the 95% CIs was too wide, thus affecting the effectiveness of the statistical model. However, SSI is not a common complication of TKA, and the incidence in our study was the same as previous investigations. The reliability of the statistical results needs to be corroborated by further studies with larger sample sizes and/or multiple centers. Gender, age and diabetes mellitus have been shown to be the risk factors for TKA, women and older adults appearing to experience more complications. Simultaneously, multisite surgeries prolonged the duration of the procedure and increased the possibility of exogenous infection. These factors were also identified in this study. Targeted preventive measures by surgeons could mitigate their effect on risk. However, there may be other confounding factors that were not different in this study. Besides, body weight, hemoglobin and albumin levels, liver or renal comorbidities, and preoperative medications are suspected risk factors for SSI after TKA. However, the statistical power of the logistic regression model was limited by the number of SSI cases which may have also affected reliability. There is no national joint arthroplasty registry in China, and such a single-center retrospective study could only provide a brief, preliminary investigation of SSI after TKA. With the rapid growth in the number of surgeries, the prevention of SSI has become an important issue for healthcare professionals. A detailed survey of the occurrence and risk factors for SSI after TKA will be the next hotspot to improve surgical outcomes. Multicenter prospective investigations are required to reveal present condition and benefit the prevention and treatment of SSI after TKA in China. Conflicts of interest None.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.