A national survey of antimicrobial prophylaxis in adult cardiac surgery across Canada

OBJECTIVE: To characterize national and regional patterns of antimicrobial prophylaxis in adult cardiac surgery across Canada. DESIGN: Retrospective, cross-sectional analysis. SETTING: Thirty-three adult cardiac surgical centres across Canada. INTERVENTIONS: A one-page questionnaire collecting information regarding institutional demographics and antimicrobial prophylaxis regimens for adult cardiac surgical procedures was mailed to all adult surgical centres across Canada. If a response was not received within one month, a second survey was mailed, followed by a telephone reminder within two weeks of the second mailing. MAIN RESULTS: The overall response rate was 100%. Prophylactic antimicrobials were used in all of the adult cardiac centres; single-agent prophylaxis was used in 97% (32 of 33) of centres. Single-dose antimicrobial prophylaxis was used in only 3% (one of 33) of centres. Preoperative and postoperative antimicrobial prophylaxis regimens varied both between provinces and within provinces across Canada. Cefazolin was the antimicrobial used in 88% (38 of 43) and 87% (33 of 38) of the reported preoperative and postoperative prophylaxis regimens, respectively. Antimicrobial prophylaxis was initiated in the operating room 72% (26 of 36) of the time, and intraoperative supplemental antimicrobial doses were administered for cardiac procedures that took longer than a median of 4 h (range 4 to 8 h). Overall, the median duration of antimicrobial prophylaxis treatment was 36 h (range 8 to 96 h). CONCLUSIONS: Despite the availability of various published guidelines, the present survey identified several areas for improvement with respect to antimicrobial prophylaxis in adult cardiac surgery across Canada.

The objective of the present nationwide survey was to characterize national and regional patterns of antimicrobial prophylaxis in adult cardiac surgery across Canada.

Adult cardiac surgical centres
Thirty-three adult cardiac surgical centres were identified from information provided by Sulzer Medica Canada Inc (Toronto, Ontario) and the Cardiac Care Network of Ontario (Toronto, Ontario).

Questionnaire
The survey consisted of a one-page questionnaire that was designed to collect the following information regarding institutional demographics and antimicrobial prophylaxis regimens for adult cardiac surgical procedures: size and type of hospital; number of surgeons and cardiac procedures per year; standard preoperative and postoperative antimicrobial prophylaxis regimens; alternate regimens for penicillinallergic patients and high risk patients; criteria for high risk; use of preprinted standard orders; initiation of antimicrobial prophylaxis; and the use of intraoperative antimicrobial doses.
In April 2000, a cover letter (addressed to the Head of Cardiovascular Surgery) and questionnaire were mailed to all 33 adult cardiac surgical centres across Canada; a French version was sent to the 11 centres in Quebec.If a response was not received within one month, a second survery was mailed in May 2000, followed by a telephone reminder within two weeks of the second mailing.This study was approved by the Ethics Research Board at Sunnybrook and Women's College Health Sciences Centre (Toronto, Ontario).

Statistical analysis
Descriptive analysis was completed using SPSS for Windows (Release 10.0.5, SPSS Inc, USA).

RESULTS
The overall survey response rate was 100%.Not all centres responded to every question.Institutional demographics are summarized in Table 1.Prophylactic antimicrobials were used by all of the adult cardiac surgical centres.Singleagent prophylaxis was used in 97% (32 of 33) of centres; combination prophylaxis (ampicillin and cloxacillin) was used in only one of the 33 centres.Single dose antimicrobial prophylaxis with cefazolin was used in only 3% (one of 33) of the centres.

Standard preoperative antimicrobial prophylaxis regimens
Cefazolin was the antimicrobial used in 88% (38 of 43) of the reported preoperative prophylactic regimens; vancomycin was used in only 7% (three of 43) of the regimens.The preoperative dosages of cefazolin varied widely both between provinces and within provinces across Canada (Table 2).

Initiation of antimicrobial prophylaxis
Antimicrobial prophylaxis was initiated in the operating room 72% (26 of 36) of the time.In four centres (12%), antimicrobial prophylaxis could be initiated in more than one location (Table 2).

Standard postoperative antimicrobial prophylaxis regimens
Cefazolin was used in 87% (33 of 38) of the various postoperative prophylaxis regimens.The postoperative dose regimens for cefazolin varied widely both between provinces and within provinces across Canada (Table 3).

Duration of antimicrobial prophylaxis
Overall, the median duration of antimicrobial prophylaxis was 36 h (range 8 to 96 h).Prophylactic antimicrobials, whether used as a single agent or in combination, were stopped within 24 and 48 h of surgery 43% and 82% of the time, respectively.Regional differences in the duration of antimicrobial prophylaxis are summarized in Figure 1.

Supplemental intraoperative antimicrobial doses
Overall, intraoperative supplemental antimicrobial doses were administered for cardiac surgical procedures taking longer than a median of 4 h (range 4 to 8 h).Intraoperative supplemental antimicrobial doses were administered 60% of the time for cardiac procedures lasting longer than 4 h.Regional differences are summarized in Figure 2.

Antimicrobial prophylaxis in high risk patients
The following criteria for high risk patients were provided

Canadian antimicrobial prophylaxis in cardiac surgery
Can J Infect Dis Vol 13 No 1 January/February 2002 23 *Some institutions reported more than one preoperative prophylaxis regimen; † Cefazolin 1 g for patients weighing less than 75 kg, cefazolin 2 g for patients weighing 75 kg or greater; ‡ Cefazolin 1 g for body mass index less than 30, cefazolin 2 g for body mass index 30 or greater; § High risk patients (not defined); ¶ Cefazolin 1 g for patients weighing less than 100 kg, cefazolin 2 g for patients weighing 100 kg or greater; **Patients undergoing valve replacements or receiving prosthetic implants, patients with endocarditis; † † Antimicrobial prophylaxis was initiated in more than one location in four centres; ‡ ‡ Administered with induction; § § Administered before skin incision.OR Operating room; NB New Brunswick; Nfld Newfoundland; Sask Saskatchewan by five centres (one in Nova Scotia, one in Quebec, two in Manitoba, one in Ontario): diabetes; older age (older than 75 years); endocarditis; sepsis; cardiogenic shock; steroid therapy; positive culture; and delayed sternal closure.Two centres (6%) modified their antimicrobial prophylaxis regimen for high risk patients.

Antimicrobial prophylaxis regimens in valve surgery
Two institutions (one in Vancouver, one in Quebec) used a different antimicrobial for prophylaxis in valve surgery (vancomycin) compared with coronary artery bypass surgery (cefazolin).The duration of antimicrobial prophylaxis was prolonged for valve replacement patients in two centres (one in Saskatchewan, one in Newfoundland).Patients undergoing valve replacement at these centres received five days of prophylaxis versus single-dose (Saskatchewan) or 24-h (Newfoundland) for coronary artery bypass surgery.

Antimicrobial prophylaxis regimens in penicillin-allergic patients
Vancomycin was used 96% (25 of 26) and 67% (16 of 24) of the time for antimicrobial prophylaxis in penicillin-allergic patients preoperatively and postoperatively, respectively (Table 4).

Paradiso-Hardy et al
Can J Infect Dis Vol 13 No 1 January/February 2002 24 *Some institutions reported more than one postoperative prophylaxis regimen; † Cefazolin 2 g administered once patient was off cardiopulmonary bypass; ‡ Cefazolin 1 g for body mass index less than 30, cefazolin 2 g for body mass index 30 or greater; § Cefazolin 1 g every 8 h for patients weighing less than 100 kg, cefazolin 2 g every 8 h for patients weighing 100 kg or greater; ¶ Patients with valve replacements

DISCUSSION
Antimicrobial prophylaxis is generally recommended in the following situations: surgical procedures with high infection rates; surgeries involving implantation of prosthetic material; and surgeries in which the consequences of infection are particularly serious (4)(5)(6)(7)(8).Although cardiac surgery is generally considered a 'clean' surgery with an intrinsically low infection rate (ie, less than 5%), postoperative infections complicating cardiac surgery are associated with potentially serious adverse outcomes (1,2).Before the use of antimicrobial prophylaxis, postoperative infection rates ranged from 0.8% to 25%, depending on the definition of wound infections and follow-up duration (14).A meta-analysis of 30 years of placebo controlled clinical trials demonstrated that antimicrobial prophylaxis for cardiac surgery (using penicillins or first-generation cephalosporins) reduces the risk of postoperative infections fivefold (odds ratio 4.96, 95% CI 2.06 to 9.72) (14).
It is recommended that the selection of prophylactic antimicrobials be directed at the most likely infecting agents, Gram-positive bacteria (ie, Staphylococcus aureus and Staphylococcus epidermidis) (15).Despite the large num-ber of prospective, randomized, controlled clinical studies using various antimicrobials alone or in combination (including penicillins; first-, second-and third-generation cephalosporins; aminoglycosides; and vancomycin), no single regimen has emerged as clearly superior in preventing postoperative infections (14).First-or second-generation cephalosporins (cefazolin, cefuroxime) are among the most effective antimicrobials in reducing postoperative infections in cardiac surgery and are the agents of choice according to the 1999 American College of Cardiology/American Heart Association guidelines (7).Data supporting the benefit of extending the antibacterial spectrum with combination antimicrobial prophylaxis are limited (14).In a recent randomized, double-blind, comparative study in 1641 adults, no significant differences in postoperative infection rates were observed among those taking cefazolin (firstgeneration cephalosporin) versus cefuroxime or cefamandole (second-generation cephalosporins), 8.4% versus 9.0% or 8.6%, respectively (16).
Studies have failed to demonstrate a clinically significant advantage of vancomycin over cephalosporins for routine antimicrobial prophylaxis (17)(18)(19).The emergence of

Canadian antimicrobial prophylaxis in cardiac surgery
Can J Infect Dis Vol 13 No 1 January/February 2002 25  vancomycin-resistant enterococci has restricted the use of vancomycin for antimicrobial prophylaxis to patients with a serious beta-lactam allergy or for hospitals in which methicillin-resistant S aureus and/or S epidermidis are a frequent cause of postoperative infections (20).
Our national survey demonstrated that, although the preoperative and postoperative antimicrobial prophylaxis dose regimens varied widely both between provinces and within provinces across Canada, the majority of centres used singleagent prophylaxis with the first-generation cephalosporin, cefazolin.The use of this agent may be attributed to its continued efficacy in preventing postoperative infections, favourable toxicity profile and reasonable cost.
Risk factors for postoperative infection include diabetes, obesity, smoking, prolonged preoperative hospital stay, chronic obstructive pulmonary disease and prolonged cardiopulmonary bypass (14,21).Prophylactic antimicrobial prophylaxis does not appear to decrease postoperative infections associated with these risk factors, whereas other techniques (such as aggressive perioperative glucose control) have been shown to have a beneficial effect (22,23).In our survey, 6% and 9% of centres modified their antimicrobial prophylaxis regimen for high risk patients and valve replacement patients, respectively.
It is established that for prophylaxis to be effective, the initial dose of the antimicrobial agent must be administered within 30 to 60 min of the time of incision to ensure a bactericidal concentration of the drug in serum and tissues (21,24).For most procedures, scheduling administration at the time of induction of anesthesia before skin incision ensures adequate concentrations during the period of potential contamination (5).In our survey, antimicrobial prophylaxis was initiated either on the ward, on-call to the operating room (OR) or in the OR holding room 26% of the time.Optimal initiation of antimicrobial prophylaxis in the OR occurred 74% of the time.
It is also established that when an antimicrobial with a relatively short half-life (eg, cefazolin) is administered preoperatively, an additional intraoperative dose may be necessary to maintain bactericidal tissue concentrations during lengthy procedures (ie, those lasting longer than 3 to 4 h) (4,25,26).In our survey, intraoperative supplemental antimicrobial doses were only administered 60% of the time for cardiac procedures lasting longer than 4 h.
The optimal duration of antimicrobial prophylaxis in cardiac surgery remains controversial.According to the most recent 1999 American College of Cardiology/ American Heart Association guidelines, a 24 h course of intravenous antimicrobial prophylaxis is as efficacious as prolonged regimens (48 h or longer), even in high risk patients (7).In our survey, antimicrobials were stopped within 24 h of surgery only 42% of the time.
Because prolonged antimicrobial prophylaxis use after cardiac surgery may result in increased drug toxicity, increased risk of acquired antibiotic resistance, alterations in a patient's gastrointestinal flora and increased cost, it should generally be avoided.Recently, an observational four-year cohort study demonstrated that prolonging antimicrobial prophylaxis beyond 48 h after cardiac surgery did not further decrease the risk of surgical site infections compared with shorter durations (shorter than 48 h) of prophylaxis (27).In our survey, prolonged antimicrobial prophylaxis (48 h or longer) occurred 19% of the time.There is no evidence to support continuing prophylaxis until removal of the chest tubes.
Table 5 provides a summary of recommended prophylactic regimens from various consensus references and highlights the diversity of opinion regarding postoperative

Recommended antibiotic regimen Alternative regimen* Duration of prophylaxis
The Medical Letter on Drugs Cefazolin or cefuroxime 1 to 2 g iv Vancomycin 1 g iv Single preoperative dose and and Therapeutics 1999 (4) one additional dose after CPB (optional) American Society of Cefazolin 1 g iv preoperatively and Vancomycin 1 g iv Up to 72 h Hospital Pharmacists (5) every 8 h or cefuroxime 1.5 g iv with or without gentamicin preoperatively and every 12 h 2 mg/kg iv Waddell et al ( 6) Cefazolin 1 g iv or cefuroxime 1.5 g iv No recommendation Optimal duration not clearly preoperatively defined -single dose may not be sufficient Eagle et al (7) Cefuroxime 1.5 g iv preoperatively, Vancomycin 1 g iv 24 to 48 h 1.5 g iv after CPB and 1.5 g iv preoperatively and 1 g every 12 h, or cefazolin 1 g iv iv every 12 h preoperatively, 1 g at sternotomy, 1 g after CPB and 1 g every 6 h Martin (8) Cefazolin 2 g iv preoperatively Vancomycin 15 mg/kg iv 24 to 48 h and 1 g every 8 h or cefuroxime 1.5 g preoperatively, 7.5 mg/kg iv preoperatively and 0.75 g after CPB and 10 mg/kg iv iv every 6 h every 8 h antimicrobial prophylaxis (4)(5)(6)(7)(8).According to these various guidelines, the preoperative antibiotic regimen was appropriate in 21% to 88% of institutions included in this survey, the postoperative antibiotic regimen was appropriate in 3% to 64% of institutions included in this survey, and the duration of antimicrobial prophylaxis was appropriate in 67% to 91% of institutions included in this survey.Antimicrobial prophylaxis practice patterns in cardiac surgery documented in the United Kingdom, the United States and Germany have demonstrated inconsistent and ineffective use of antimicrobial prophylaxis (9)(10)(11)(12)(13).Frequent errors included incorrect timing of initial dose, prolonged duration of prophylaxis, incorrect selection of antimicrobial and inappropriate dose regimens.To our knowledge, this is the first survey characterizing national and regional patterns of antimicrobial prophylaxis in adult cardiac surgery across Canada.
Randomized trials do not provide data on the effect of antimicrobial prophylaxis in actual clinical practice.A major limitation of our study was the inability to document patient outcomes to assess the effectiveness of the various prophylaxis regimens used in adult cardiac surgery across Canada.

CONCLUSIONS
Despite the availability of various published guidelines for antimicrobial prophylaxis in adult cardiac surgery, the present survey identified several areas for improvement in adult cardiac surgery across Canada.These include initiating antimicrobial prophylaxis in the OR, using intraoperative supplemental antimicrobial doses for cardiac procedures lasting longer than 4 h and shortening the duration of antimicrobial prophylaxis to 24 to 48 h.The results from this survey have been provided to all cardiac centres across Canada, and may be used to develop and/or revise evidence-based institutional guidelines.A follow-up survey is planned in one year to determine whether this information improved practice patterns.

Figure 1 )Figure 2 )
Figure 1) Duration of postoperative antimicrobial prophylaxis Figure 2) Duration of cardiac surgical procedure necessitating intraoperative antimicrobial supplement doses

*
Vancomycin 500 mg administered once patient off cardiopulmonary bypass; † Dose not available.BMI Body mass index; NA Not available

Paradiso-
Hardy et al Can J Infect Dis Vol 13 No 1 January/February 2002 26

TABLE 1 Demographics of Canadian cardiovascular surgical centres
*Mean ± SD; † Information not reported on two questionnaires.CABG Coronary artery bypass grafting; NA Not available; NB New Brunswick; Nfld Newfoundland; Sask Saskatchewan