In this study, cardiac surgery population was divided in patients at high risk of developing infections because of well-known risk factors and patients at low risk. As expected, patients at high risk (i.e., patients with severe comorbidities, immunosuppressive therapy, severe obesity or malnutrition) had a significant higher occurrence of SSIs as well as of BSIs, LRTIs and mortality compared with patients at low risk, independently of violation of the antimicrobial prophylaxis protocol. This finding reflects a population of patients who were more severely ill and therefore at higher risk for postoperative infectious complications.
The Society of Thoracic Surgeons Practice Guidelines on antibiotic prophylaxis in cardiac surgery recommended that in the setting of the institutional presence of a ‘high incidence’ of MRSA, it would be reasonable to combine a β-lactam (cefazolin) with a glycopeptide (vancomycin) for prophylaxis (Class IIB recommendation, Level of Evidence C) [16]. Optimal dosage regimens of vancomycin and protocol of administration still remain controversial [16, 22]. The Society of Thoracic Surgeons Guidelines mentioned that any of the following doses and durations may be used: 1000 mg, 1500 mg, or 15 mg/kg; and 24 h versus 48 h or 1 dose versus 2 doses [16, 22]. Specifically, guidelines for appropriate dosing of prophylactic antibiotics stated that ‘In patients for whom vancomycin is an appropriate prophylactic antibiotic for cardiac surgery, a dose of 1 to 1.5 g or a weight-adjusted dose of 15 mg/kg administered intravenously slowly over 1 h, with completion within 1 h of the skin incision, is recommended’ (Class I, Level of Evidence A) [16]. Similarly, the 2011 American College of Cardiology/American Heart Association guideline for CABG surgery recommended that ‘Antibiotic prophylaxis should be initiated 30 to 60 min before surgery, usually at the time of anaesthetic induction, except for vancomycin, which should be started 2 h before surgery and infused slowly’ [6]. Finally, the Scottish Intercollegiate Guidelines Network, updated April 2014, stated that ‘Vancomycin should be given by intravenous infusion starting 90 min prior to skin incision’ (Class B recommendation) [9].
Several studies investigated the association between measure(s) applied for reducing the rate of SSIs and their occurrence [23, 24]. Our study focused on the timing of antimicrobial prophylaxis; specifically, on the relationship between the first skin incision and the end of vancomycin infusion. We found that the initial surgical incision was performed before the vancomycin infusion had been completed in nearly 40% of patients. Generally, the reason for the violation of the antimicrobial prophylaxis protocol was due to our policy to start antimicrobial prophylaxis in the preoperative holding area under supervision of anaesthesiologists. This policy was adopted in our Cardiovascular Surgery Unit following the occurrence of some relevant adverse drug reactions due to vancomycin administration (i.e., mainly hypotension; occasionally, red man syndrome) [23, 24].
Differently, in the Garey’s study [25] cardiac surgery patients were assigned to five groups on the basis of the relation between the start time of the vancomycin infusion and the time of the initial surgical incision. In this study, antibiotic prophylaxis was started in the preoperative holding area only for the first surgical case of the day and in admission unit for all subsequent cases immediately prior to transferring the patient to the preoperative holding area. These Authors reported that of the 2048 patients in the study, 0.7% received vancomycin 0–15 min before incision, 8.6% 16–60 min before incision, 43.4% 61–120 min before incision, 34.2% 121–180 min before incision and 13.1% >180 min before incision.
The relationship between the timing of antimicrobial prophylaxis and the occurrence of SSIs has been studied with conflicting results. The Surgical Care Improvement Project measure assesses compliance for antimicrobial prophylaxis administration initiated within 60 min (or 120 min for vancomycin) prior to surgical incision [11]. The choice of the preincision 60-min window for antimicrobial prophylaxis was based on two types of evidence: pharmacokinetics of the antibiotics and one large cohort study analyzing the association between timing of antibiotic administration and SSIs in several types of surgical procedures [11].
However, following studies investigating this relationship did not clearly demonstrate the superiority of the 60-min window [17–20]; in particular, some studies showed lower risk of SSI with shorter times between antibiotic administration and skin incision. Garey et al. reported that SSI developed in 26.7% of cardiac surgery patients who received vancomycin 0–15 min before incision, 3.4% of patients between 16 and 60 min before incision, 7.7% of patients between 61 and 120 min before incision, 6.9% of patients between 121 and 180 min before incision and 7.8% of patients >180 min before incision [25]. Steinberg et al. in an observational study (43.6% were cardiac patients) found a trend toward lower risk of SSI occurring when antimicrobial prophylaxis with vancomycin or cephalosporins were given within 60 and 30 min prior to incision, respectively [18]. Hawn et al. in a retrospective study in noncardiac surgery patients found that the SSI risk was not significantly associated with prophylactic antibiotic timing [19].
SSIs are still among the most severe complications in cardiac surgery patients. The overall SSI rate observed in our study was 8.1%, which was similar or lower to that previously reported [1–3]. The main finding of our study was that violation of the timing of vancomycin prophylaxis protocol was a significant risk factor for development of SSI in patients undergoing cardiac surgery. Specifically, when the first surgical skin incision was performed before the end of the vancomycin infusion, we observed a 5-fold increased rate of SSIs both in low and high risk patients.
Nosocomial infections occur in 10 to 20% of cardiac surgery patients [6]; however, while SSIs incur significant morbidity and costs but rarely lead to death, conversely, postoperative LRTI, BSI and endocarditis are more frequently correlated with mortality [15]. In our study, we also found a consistent relationship between violation of vancomycin prophylaxis timing protocol and rates of postoperative infectious complications as well as mortality from infectious cause. Specifically, BSIs and mortality were increased 6-fold and more than 3-fold, respectively both in low and high risk patients. Moreover, LRTIs and UTIs were increased 2-fold and 7-fold, respectively in high risk patients. Also for mortality, violation of the timing of vancomycin prophylaxis protocol was a significant risk factor.
Actually, before starting this study we did not suspect that the timing of vancomycin prophylaxis administration was being violated at this rate as well as that this violation was associated with a significantly increased rate of SSIs, postoperative infectious complications and mortality from infectious cause. However, the overall SSI rate and mortality observed in our study were similar or lower to those previously reported in other studies in cardiac surgery patients [1–6].
Policies and practices aimed at reducing the risk of SSIs include performing surveillance for SSIs as well as measuring and providing feedback to healthcare providers on the rates of compliance with process measures, including antimicrobial prophylaxis [7, 12]. This study let us to discover that the violation of the protocol was due to the start of antimicrobial prophylaxis in the preoperative holding area. Indeed, the information obtained in this study was reported to our healthcare providers and has altered practice patterns for avoiding the persistent risk of violation of prophylactic vancomycin administration timing.
Strengths and limitations of the study
If compared with previous studies, our study has some relevant features: (1) it was a prospective study; (2) data were collected through a clinical study and not from a database or registry; (3) antibiotic timing data were collected in ‘real-time’ in the operating room and not from the patient chart; (4) only cardiac surgery patients were enrolled; (5) all patients received the same prophylactic administration of antibiotics; (6) the rate of BSIs, LRTIs, UTIs and mortality from infectious cause were also investigated; (7) it was concluded in only 12 months and (8) no patient was lost to follow-up. Moreover, to the best of our knowledge, this is the first study investigating the relationship between the rate of SSIs in cardiac surgery patients and the presence or absence of violation of the timing of antimicrobial prophylaxis administration, comparing patients at low and high risk of infections.
The study presented several limitations. First, it was a single-centre study. Second, neither a calculation was made on the number of subjects required nor an interim analysis was conducted since the study was designed by our statistician to be continuous over one year. Specifically, the duration of a year was necessary to enrol an adequate number of patients (i.e., 741) to obtain statistically significant differences among the groups. As a matter of fact, the earlier studies enrolled a number of patients ranging among 2048 and 4472 [18, 25].
Moreover, being an observational study, there was no randomization of patients to the two groups (with and without protocol violation), although the characteristics of patients in the two groups turned out not to be statistically different. Finally, in the study period patients were not screened for S. aureus colonization prior to surgery.