In this study, we summarized our experience with surgical site infections and perioperative antibiotic prophylaxis consisting of the first generation cephalosporin administered at a single dose 30 minutes prior to the initial skin incision. In the studied group of 254 deceased donor renal graft recipients, 10 (3.9%) developed surgical site infection. We calculated that reoperation adjusted for the presence of hematoma in the surgical site and BMI is a risk factor for this challenging complication (adjusted odds ratio: 6.963, 95% confidence interval: 1.523–31.842, P = .012).
Surgical site infections occur in 3–11% of patients after kidney transplantation (3–5,9,19). The prevalence of 3.9% in our transplant unit seems, therefore, to be relatively low. Even though the complication rarely leads to graft loss, it is proved to significantly increase morbidity and prolonged hospitalization time (20).
The identification of certain surgical site infection risk factors has been a significant problem for years as the conclusions vary in different publications. Most of them mention body mass index (BMI), immunosuppression regimen, older age, diabetes and accompanying heart diseases, cold ischemic time, duration of surgical procedure, and delayed graft function (20–23). We could verify those factors with a relatively high confidence, given the balanced study groups in terms of the baseline characteristics and satisfactory post-hoc power.
While evaluating the impact of BMI on surgical site infections, we expected that patients with a higher BMI would acquire more episodes of surgical site infections. Obese patients are proved to have an increased prevalence of postoperative wound colonization, as demonstrated by multiple studies (24–27). Some data on intraabdominal surgery show that especially the thicker subcutaneous fat tissue is related with higher wound morbidity (28). In our study, we observed that the median BMI was higher in patients with surgical site infections compared to those without infection but the difference was not statistically significant (28.0 vs 25.0, U = 842.0, P = .096). However, BMI was indeed confirmed to be an important confounding factor that was included in our final logistic regression model.
Literature clearly confirms that patients with more intense immunosuppression are expected to acquire more surgical site infections (25). In current study, each patient received the same, triple-drug immunosuppressive treatment, and almost 88% (224/254) of them did not receive antibody induction. As the protocol was universal, no differences could be found between the studied groups and we could not identify more intensive immunosuppression to be a predisposing risk factor for SSI. Similar conclusions can be drawn regarding patients’ age. However, reoperation was confirmed as a major risk factor for surgical site infection (25,29). It is most likely caused due to increased exposure of the surgical site on contamination (20). In such circumstances, the immunocompromising influence of anti-rejection drugs worsens the outcomes (29). We detected that the presence of a hematoma in the operated area also predisposes to the development of infection. This is not a frequent observation in the literature. In fact, the two variables (hematoma and reoperation) are correlated (not strongly, VIF = 1.075) as often they can cause one another. The multivariable logistic regression analysis confirmed that reoperation is an independent risk factor for surgical site infection. Therefore, we provided an odds ratio adjusted for the occurrence of hematoma in the surgical site and patient’s BMI. That leads us to the point, where in the future, this observation may be helpful in postoperative estimation of the risk of infection wound problems.
Evidence of the beneficial role of perioperative antibiotic prophylaxis in reducing postoperative surgical site infections in solid organ recipients is well-established (5,7,19). Only a few reports question its efficacy (9,30). However, specific antibiotic regimens and durations vary widely across transplant centers and procedures, and the quality of the evidence supporting specific practices is quite poor (5,31–33). Currently, there are no strict formal recommendations on perioperative antibiotic prophylaxis administration in kidney recipients. According to the IDSA/ASHP/SIS/SHEA guidelines, each transplant center should customize their own protocol that corresponds with their unique microbiological setting (patients’ and donor’s colonization statuses, local resistance patterns, and presence of nosocomial pathogens) (5). It should also be pointed out that such prophylaxis can lead to some problems including increased: financial costs, rates of Clostridioides difficile colonization and colitis, incidence of antimicrobial resistance, and other adverse events (34–37). Thus, targeted antibiotics should be used in each case based on individual risk factors and local patterns of resistance (38–40). Moreover, the spectrum of organisms involved in SSI in kidney recipients is more diverse than for the general population due to other important factors such as the underlying end-stage organ failure, immunosuppression, prolonged hospitalization, organ transportation/preservation, and previous exposures to antibiotics in donors and recipients that could predispose to infections with multidrug‐resistant organisms. Therefore, the use of antibiotic prophylaxis should be optimized as to be truly effective yet still minimal as to restrain the spread of multidrug resistant pathogens. As we modified our local protocol a few years earlier, we had a unique chance to compare our historical data with the current ones.
Before October 2015 our strategy was to administer 2.0 g of the third generation cephalosporin (ceftriaxone) in a single infusion 30 minutes prior to the first skin incision. We observed back then that surgical site infections occurred in 12 of 214 patients (5.6%) (7). The historic incidence is slightly higher, although we think that the difference is not clinically significant. This finding is very satisfying as we started using an antibiotic with a narrower spectrum and achieved similar results. We hope that such an approach can at least minimally help to decrease the risk of development of new multi drug resistant strains. Similar conclusions were demonstrated by other authors (4,21). Furthermore, other factors like: shortening surgical operative time, optimal sterile technique, proper perioperative management of patients’ comorbidities as well as good glucose and temperature regulation, are also assumed to be imperative to limit surgical site infections but those elements were outside the scope of our research (41–45).
The main limitations of our study are the limited number of cases and the retrospective, non-randomized design carried out at a single Central European institution. The incompleteness of medical records resulted in the presence of some missing data. Therefore, we cannot exclude bias due to missing data. Another bias that might be present in such case-control study is caused by confounding (46). We attempted to eliminate it by restriction (e.g., we excluded patients with Bricker ileal conduit due to significant deviation from a standard kidney transplant procedure that carries a higher risk for surgical site infection) and multivariable logistic regression (i.e., we calculated adjusted odds ratio).
Another limitation is related to the lack of definition of minimal clinically important difference (MCID) regarding the comparison of different perioperative antibiotic prophylaxis regimens. Such a concept would help transplant centers to confirm when a change in their perioperative antibiotic prophylaxis is reasonable. We were able to compare our current prophylaxis policy to the previous one concerning the prevalence of surgical site infection, although it was not the main scope of our study. We conclude that the decrease in prevalence of surgical site infection of 1.7% is not clinically relevant (47). The difference is also not statistically significant (data not shown) which increases our confidence in that matter. However, if MCIDs were defined by transplant societies regarding the outcomes of perioperative antibiotic prophylaxis, it could possibly improve the process of choosing the most optimal regimen in each transplant center (48).