D. H . W i t t m a n n , R. E. C o n d o n

Prophylaxis of Postoperative Infections* Summary: The antibiotic most appropriate for of anaerobes. Selection of resistant bacteria has not been prophylaxis of postoperative infections depends on the significant and is unlikely to become so with single-dose nature of the operation. In aseptic (clean) operations, prophylaxis. Occasionally, if there is a high probability gram-positive postoperative infections are the primary t h a t the operative field may be heavily contaminated by concern, and cefazolin is recommended because of its anaerobes, metronidazole should be added. Dosing excellent pharmacokinetics and good activity against should be sufficient to cover the operative period. Only a gram-positive pathogens, including staphylococci. In single prophylactic dose is necessary, given at the time of those operations where violation of the digestive tract induction of anesthesia. For particularly long operations, creates a contaminated field, a cefotaxime-generation a second dose of those antibiotics with half-lives shorter cephalosporin is the agent of choice because of the than 60 min is required two hours after the first. excellent safety profiles and the capability of agents of Single-injection prophylaxis is effective, inexpensive, has this class to kill essentially all pathogenic gram- no side effects and does not induce bacterial resistance. negative aerobes as well as a substantial portion Zusammenfassung: Prophylaxe postoperativer Infektio= Nebenwirkungen ~iul3erst selten sind. Die Selektion nen. Die Art der Operation bestimmt die Wahl prim~ir resistenter Bakterien durch die Einzeldosierung des besten Antibiotikums zur Prophylaxe post- ist nicht beobachtet worden und unwahrscheinlich. operativer Infektionen. Infektionen nach aseptischen Starke Kontamination des Operationsfeldes durch Operationen werden vorwiegend durch gram- obligate Anaerobier erfordert die gleichzeitige positive Bakterien verursacht, und wir empfehlen Verabreichung von Metronidazol. Die Dosierung soil Cefazolin z u r Prophylaxe, weil es gut grampositive ausreichend sein, um genfigend hohe Konzentrationen in Kokken, einschlie61ich der Staphylokokken, eliminiert der Operationswunde zu realisieren. Die bei und aul3erdem aus pharmakokinetischer Sicht Operationseinleitung verabreichte Einzeldosis hat die hervorragend zur Prophylaxe geeignet ist. Bei besten Resultate. Dauern die Operationen extrem lange, Operationen mit Kontamination der Operationswunde ist eine zweite Dosis zwei Stunden nach der ersten dutch Darmbakterien nach Magen-Darm-Er6ffnung, indiziert, falls das Antibiotikum eine kurze sind Cephalosporine der Cefotaxim-Generation Mittel Eliminationshalbwertzeit v o n weniger als 60 Minuten der Wahl, da sie alle wichtigen gramnegativen aeroben hat. Die Einzeldosisprophylaxe ist wirksam, billig, hat und wesentliche Teile der obligat anaeroben keine Nebenwirkungen und induziert keine bakertielle Bakterien erfassen und da mit der Einzeldosierung Resistenz. Introduction Since the discovery of sulfonamides by Paul Ehrlich and of penicillin by Alexander Fleming, antimicrobial agents have been used not only to treat but also to prevent infections associated with operative procedures. It was, however, only in 1959 and 1961 that the scientific basis for the use of prophylactic agents was laid [1, 2]. Burke [2] was able to show that infections could be prevented only when antimicrobial agents, independent of their action, were given prior to the infectious challenge. When given three hours following a challenge with infecting bacteria, antibiotics were ineffective in preventing infection. At about the same time, a multicenter study was conducted to evaluate the effect of ultraviolet irradiation on the hospital environment and the development of postoperative infection [3]. In this study the current practices of antibiotic prophylaxis were assessed. The results were astonishing. Patients receiving prophylactic antibiotics had an infection rate of 14.3%, while those who did not receive prophylaxis had infection rates of 4.4%. Even when adjusted for type of operation and operative

risk, a significantly higher infection rate was observed in patients receiving antimicrobial agents for prophylaxis. At that time, the use of prophylactic antibiotics differed markedly from accepted current practice: the antibiotic was first administered at the termination of the operation and was usually continued for as long as the patient was hospitalized. Such practices clearly have the potential for negatively affecting the incidence of postoperative wound infection and Offer no opportunity for any positive or desirable effect. We now know that if an antibiotic is injected intravenously just prior to the incision, it is effective in preventing wound infections. Timing is critical

(Figure 1).

D. 1t. Wittmann, M.D., Ph.D., R. E. Condon, M.D., Dept. of Surgery,

Section of Trauma and Emergency Surgery, Medical College of Wisconsin, 8700W. WisconsinAve., Milwaukee,WI 53226, USA. * See alsoAprahamian, C. and Wittmann, D. H.: Operative management of intraabdominal infection. Infection 19 (I991) 453-455.

Infection 19 (1991) Suppl. 6 © MMV Medizin Verlag GmbH Miinchen, Mfinchen 1991

S 337

D. It. Wittmann et al.: Prophylaxis of Postoperative Infections Table 1" Infection rates associated with operations in which antimicrobial prophylaxis was not employed (placebo groups).

Concentration ~SKIN INCISION t00

l i~: ![llil~il~i~,~ i l ii i~iiiiiii l iiii@::iii;iiiiFi@ ~]ilitlili(!~iiti!, ! @t~iii~Ji ,iiiiiii~ili!!4~i~!iiii@~i~!ii~ ~ii @i@,!i~:i:~iliNi:~ @ ]!i~!lN i@i !~jii i!ii~N!~il!i1! ~ @iiii li~ii~ii~ii;i~~ii! i@!iNi ,~ i I i/i~i~i;i!ilit?il:i!!

t~ \ I

\\

\

10

\. PERIOD i

i

TOO EARLY Serum

i

i

CORRECT TIMING' Tissue Fluid

il ~

I

TOO LATE

I

VULNERABLE PERIOD j |

Figure 1 : The relation between timing of antibiotic prophylaxis and the start of an operation. Administration of the antibiotic needs to be properly timed so that effective serum and tissue concentrations are present during the vulnerable period when t h e wound is open and the tissues are subject to contamination. For antibiotics with a serum half-life of 1-4 h, administration should occur as a bolus about 30 rnin before the skin incision, done most conveniently in conjunction with the induction of anesthesia. If given too early or too late, coverage during the vulnerable period will be inadequate. If the operation is longer than 4 h, so that the vulnerable period is extended, the antibiotic should be redosed, with the second dose given 2 h after the first dose.

Pathophysiology Following the surgical incision, normally sterile tissue is exposed to an unsterile environment and contamination occurs with any operation. T h e host defense mechanisms cannot react fast enough to effectively defend this sudden bacterial invasion. Bacteria, on the other hand, find excellent living conditions in tissues traumatized by the surgical intervention. Bacteria start multiplying before effective host defense is established. Host defenses recognize traumatized and dead tissue for elimination by the same humeral and cellular mechanisms that fight invading bacteria. Thus, surgical operations claim phagocytes twofold: through more traumatized tissue, and through high counts of contaminating bacteria. Eventually, when major operations are performed, fewer phagocytes and other defense mechanisms are available to eliminate bacteria, and pathogens may have multiplied to high numbers, resulting in a wound abscess. The amount of tissue damage/correlates negatively with the number of bacteria required to induce infection [4]. Following closure of the wound, its environment is sealed off due to intravascular coagulation and products of early inflammation. Any remaining blood within the wound cavity may harbor and nourish bacteria. Pathogens are protected from host defense and antimicrobial agents by this diffusion barrier, which has sealed the wound cavity and eventually will mature into an abscess membrane. This may explain why postoperative antibiotics are ineffective in preventing wound infection. Administered at S 338

Colorectal 1,449 Appendectomy Normal appendix 176 Inflamed appendix 379 Gangrenous appendix 149 Gastroduodenal Peptic ulcer, elective 165 Gastric outlet stenosis 50 Perforation 45 Gastric ulcer 66 Gastric tumor 61 Upper GI tract hemorrhage 41 Gastr0duodenal, not specified 252 Cholecystectomy Risk factors not specified 763 Risk factors present 97 No risk factors 81 Vascular 517

580

40.0

24 60 83

13.6 15.8 55.7

8 8 8 15 19 20 69

4.8 t6.0 17.8 22.7 31.1 48.8 27.4

115 20 10 66

15.1 30.0 12.5 12.8

GI = gastrointestinal. the right time, however, antibiotics diffuse into the peripheral compartment, in this case the wound fluid; the wound is saturated with antimicrobial agents at the time it is opened, and potentially invading bacteria will encounter bactericidal conditions long before host defenses are activated (Figure 1). This is the vulnerable period, defined in 1962 by Burke [2], during which host defense has not been established. To compensate for this deficit, antibiotics must be present in wounds early enough to interfere with bacterial growth, thus reducing their ability to cause postoperative infection. if, however, the antibiotic is given too early, it may diffuse into and then be eliminated out of the tissues before the incision and not be available in the wound during the vulnerable period, thus having no effect. For years the basic principles concerning the appropriate timing in administering antibiotics for prophylaxis were not accepted. It was not until nearly ten years after the description of the decisive or vulnerable period that a controlled clinical trial was conducted, demonstrating the importance of timing in a prospective, randomized study [5] that clearly showed that antibiotics were effective only when given immediately preoperatively. Antbiotics are ineffective for prophylaxis when given following termination of the operation [6]. Many studies have now been published, and today it is common practice to give prophylactic antibiotics during the hour prior to the incision and to maintain sufficient antimicrobial levels in the surgical wound during the vulnerable period by repeated dosing during long operations [%9].

Procedure-Specific Infection Rates It is difficult to assess the impact of prophylaxis on current infection rates associated with specific operations because

Infection 19 (1991) Suppl. 6 © MMV MedizinVerlag GmbH Mtinchen,Miinchen 1991

D. It. W i t t m a n n et al.: Prophylaxis o f Postoperative Infections

Table 2: Single-dose prophylaxis versus multiple-dose prophylaxis in colon operations (results of prospective randomized trials). Number of studies Single injection: Operations performed Wound infections seen Infection rate (%) Multiple injections: Operations performed Wound infections seen Infection rate (%)

35 mg/I

ANTIBIOTICTISSUE FLUID CONCENTRATION

3o =~

27 510 22 4.3

2 DOSES

493 34 6.9

VULNERABLE PERIODI VI

ONE SINGLEDOSE I

0

1

0~

prophylaxis is widely accepted and used. Therefore, studies involving control groups receiving no antibiotic prophylaxis can no longer be justified ethically. Data from old studies may be used to assess the situation, but it must be kept in mind that these studies are more than ten years old and that surgical techniques and other factors may have changed sufficiently to alter the infection rate without considering the use of antimicrobial agents for prophylaxis [7]. Table 1 summarizes site-specific infection rates published before the routine use of antimicrobial agents or taken from placebo groups of prospective controlled trials for various indications.

. . . . . . . . 2

3

I

4

N It ~ t

5

....... 6

j

~ ' ~

~

7

....... 8

.

9

10 HRS

Figure 2: Tissue fluid concentrations following two additive doses of an antibiotic with short serum half-life. The serum concentration is not shown. Note that the second dose is given 2 h following the. first dose to boost tissue concentrations. Maximal tissue saturation of an antibiotic with a half-life of about 1 h is achieved with most 13-1actam antibiotics in the second hour following administration of a bolus injection. If the second dose of an antibiotic is given after 6 or 8 h, most of the tissue compartment will be emptied and need to be refilled, and the high concentrations observed with a booster injection after 2 h will not be attained.

Single-Dose Prophylaxis The first prospective controlled trial that sought to find an answer to the question of the postoperative duration of antibiotic prophylaxis was published in 1977 [8]. A single preoperative dose of cefazolin versus a regimen of cefazolin given for a period of five postoperative days was tested. The infection rate with a single dose was 3% and with the multiple five-day dosing, 5%. Although no significant difference could be shown by statistical evaluation, the slight numerical difference in favor of the single dose seen in this study was found in many subsequent studies that tested the same hypothesis for various indications. The number of patients enrolled in each study was often too small to prove any difference between the single injection versus the muttidose prophylactic regimen statistically, but most studies showed a difference in favor of the single injection [7]. It is astonishing in light of this evidence (Table 2) that multiple-injection prophylaxis is still used in many institutions. The most prudent and conservative interpretation of the results of all of these studies is that at the very least, single-injection prophylaxis is as effective as multiple dosing and should be given preference because it is less likely to alter antibiotic resistance patterns of bacteria in an institution, which, in turn, influence the number of serious nosocomial infection problems. If, however, an antibiotic with a short half-life is used for prophylaxis and the operation time exceeds four hours, a second dose must be given two hours after the first dose to boost tissue concentrations (Figure 2) [9]. Infection 19 (1991) Suppl. 6

Results of Controlled Trials To give an overview of the efficacy of various antibiotic agents in specific clinical circumstances, we have summarized the results of published controlled studies that fulfilled most of the following criteria [7, 8, 10-41]: prospective study, definition of entrance criteria and drop-out conditions, consideration of risk factors, unbiased randomization, clear definition and double-blind Table 3: Prospective randomized parenteral antibiotic trials in

colorectal operations.

iiiiiiii i!ii!Ni?i:!ii?iii)!i!}iiiii! Sulfonamides Aminoglycosides Lincomycin Metronidazole Nitroimidazoles Tetracyclines Penicillins Cephalothin Cephaloridine Cefazolin Cefamandole Cefoxitin Moxalactam Cefotaxime/ceftizidime Cephalosporin + metronidazole

© MMV Medizin Verlag GmbH M~nchen, MLinchen 1991

ii i iiNi ii iil 4 5 4 18 21 11 7 5 4 4 7 13 8 7 10

170 177 105 744 941 693 338 190 225 135 429 650 340 287 333

iiI 106 72 18 126 139 64 26 42 28 16 53 71 27 16 45

62 41 17 17 15 9 8 22 12 12 12 11 8 6 14

S 339

D. H. Wittmann et al.: Prophylaxis of Postoperative Infections assessment of clinical and bacteriological outcome, and appropriate statistical analysis.

Colon Operations Placebo: The risk of wound infection following colorectal operations without prophylaxis can be assessed by examining placebo groups of early antibiotic prophylaxis studies. In 38 studies, a total of 1,449 patients received placebol Of these, 580 (40%) developed a wound infection (Table 1). Single-injection antibiotic prophylaxis or effective oral antibiotic bowel preparation reduced postoperative infections to rates of less than 10% 0"able 3).

Penicillins: Twenty-six of 338 patients developed a wound infection following prophylaxis with penicillins alone. In five studies, mezlocillin was used, resulting in an infection rate of 8.4% (21 of 251 cases).

Cephalosporins: Various first- and second-generation cephalosporins have been used. In 22 studies, 160 wound infections occurred following 1,079 operations, resulting in a rate of 14.8%. The best results were obtained with cefazolin. In four studies, 16 wound infections were seen in 135 patients, giving an 11.8% infection rate with cefazolin prophylaxis. Thirteen studies were conducted with cefoxitin, and 71 wound infections w e r e seen following 650 colorectal operations (10.9%). The use of third-generation cephalosporins seems to increase the efficacy. Eight studies have been performed with cefotaxime, ceftriaxone or moxalactam, and 27 infections have been seen following 340 colorectal operations (7.9%). Due to the long half-life of ceftriaxone, conditions of multi-dose prophylaxis are created with single-dose application and therefore this drug c a n n o t be considered for single-dose prophylaxis. In five studies, the infection rate in 210 patients who received cefotaxime o r ceftizoxime perioperatively was 6.2%.

Aminoglycosides: In five studies, aminoglycosides alone have been used. Seventy-two infections were observed following 177 colorectal operations, resulting in an infection rate of 41%, equal to that obtained with placebo.

Tetracyclines: The use of tetracycline seems to work extremely well. Most studies have been carried out in Scandanavian countries. Eleven studies fulfilled the criteria of properly conducted investigations. In these, 693 patients were enrolled and 64 infections seen (9.2%). Of these 11 studies, nine were performed using doxycycline as the prophylactic antibiotic. In 508 colorectai operations, 43 infections (8.5%) were observed. Tetracyclines are not considered to be the first choice for prophylaxis of colorectal operations, but the results of these studies cannot be denied. The bacteriostatic activity of tetracyclines against aerobic as well as anaerobic bacteria may explain these favorable results. While bacteria contaminate the wound, they usually will not be in a S 340

Table 4: Preparation for elective colorectal operation.

Pre-operative Day: 1. Oral administration of polyethylene glycol-electrolytesolution (Colyte, Golytely),41 over 2 h beginningat 10:00 a.m. 2. If chronic constipation or extensive diverticulosisexists, double the purgation volumeand start 2-3 h earlier. 3. If the patient cannot swallowthe required volumein the required time-frame, pass a nasogastric tube into duodenum and administerpurgationby lavage,removingthe tube whencomplete. 4. Clear liquid diet only on this day. 5. Neomycin, 1 g, and erythromycin1 g together orally at 1 p.m., 2 p.m. and 10 p.m. 6. Nothing per os after midnight Operation Day: 1. Evacuatethe rectum completelyinto a commodebefore transport to the operating theater. 2. Administeri.v. antibioticwith inductionof anesthesia. 3. Commenceoperation at 8:00 a.m. 4. If the operation is scheduledlater in the day, appropriatelyadjust the oral neomycin-erythromycindosing schedule so that the operation starts 19-20 h after the first dose of oral antibiotics. dividing phase of growth and thus bacteriostatic tetracyclines may be more effective than bactericidal drugs like the cell-wall-active [3-1actams, which require multiplying bacteria to exert their antibacterial activity.

Anti-anaerobic antibiotics: Many studies have been performed using nitroimidazoles alone as a monosubstance, but the results are not favorable. Metronidazole alone reduces infection rates to 16.9% (18 studies, 114 infections following 674 procedures). The same is true in studies where only lincomycins'or clindamycin have been used, with a similar infection rate of 17.1%. These antibiotics act mainly against anaerobic bacteria, which play an important role in the development of wound infection following colorectal surgery, but aerobic coverage must be provided too. Thus, combinations including metronidazole generally yield better results. Recommended antibiotic prophylaxis for an elective colon operation involves the following: 1. effective mechanical cleansing of the bowel; 2. oral administration of antibiotics; and 3. parenteral administration of single-injection intravenous antibiotic(s) Mechanical cleansing is an essential preparatory step; polyethylene glycol purgation is p r e f e r r e d . Oral administration of neomycin and erythromycin base was documented years ago to be effective [10] and remains our preference O-able 4). In seven studies, 1,010 patients underwent colorectal operations using this combination for prophylaxis and only 52 (5.1%) developed an infection postoperatively [7]. We do not recommend the sole use of parenteral antibiotics under elective circumstances, but this is the only prophylactic regimen available for emergency operations.

Infection 19 (1991) Suppl. 6 © MMV MedizinVerlag GmbH M~nchen, Mtinchen 1991

D. H. Wittmann et al.: Prophylaxisof Postoperative Infections Table 5: Antibiotics in appendicitis, all disease stages.

Placebo Povidone-iodine Cephalosporins Metronidazole Extended-spectrum penicillins Anti-aerobe/anaerobe combinations

Table 6: Antibiotic prophylaxis in gastroduodenal operations.

17 1 5 15

856 16 53 143

3,951 60 336 1,325

21.7 26.7 15.8 10.8

2

10

114

8.8

4

7

261

2.7

Appendectomies Of 553 patients who underwent appendectomy without antibiotic prophylaxis, 14% developed an infection when the appendix was normal, 12% developed an infection when there was only lymph node hyperplasia, and 16% developed an infection following removal of a chronically inflamed appendix. In another study, 40 of 288 patients developed a wound infection following acute appendicitis (14%), and 41 out of 59 (70%) patients with perforated appendicitis developed a wound infection (Table 1). The presence of abscess increased the wound infection rate to 94% [7]. Antibiotic prophylaxis in acute appendicitis is partly treatment and partly prophylaxis. The antibiotic may be given for the treatment of the intraabdominal infection and for prophylaxis of infection of the abdominal wall wound. Antibiotics are of questionable therapeutic value in cases of appendectomy to treat simple acute appendicitis but are certainly therapeutic in the presence of perforated appendicitis with peritonitis. For all other circumstances, the term prophylaxis is appropriate and single-injection prophylaxis is justified. The best results have been obtained with antibiotics that are active against aerobic as well as anaerobic bacteria. In four studies where an aminoglycoside or a cephalosporin was used in combination with either clindamycin or metronidazole, only seven of 261 patients developed wound infections (2.7%). The best regimen seems to be metronidazole combined with a cephalosporin. Studies performed using local antibiotics for prophylaxis of wound infections, meaning antibiotic irrigation of the wound, included 390 patients. Rates of wound infection ranged from 0 to 21%. None of the studies were conducted in a proper scientific fashion [7]- In many studies the degree of inflammation of the appendix was graded into normal appendix, inflamed appendix, and purulent, gangrenous or perforated appendix. Twenty-four wound infections were seen in the placebo groups in studies utilizing this classification [7] following removal of 176 cases of normal appendix (13.6%). In 379 inflamed appendices, 60 wound infections were seen in the placebo group (15.8%). Eighty-three out of 149 patients with gangrenous or perforated appendix developed wound

Placebo Penicillins Cefuroxime Aerobe/anaerobe combinations Cephaloridine Cefazolinor cefamandole

11 3 4

67 11 5

250 72 102

26.8 15.3 4.9

4 2 3

16 2 3

378 54 135

4.2 3.7 2.2

infections (55.7%). With the use of prophylactic antibiotics, infections following appendectomy can be reduced with the normal appendix from 13.6 to 5.4% and with the inflamed appendix from 15.8 to 9.6%. When the appendix is gangrenous or perforated with or without peritonitis, infection rates are reduced from 55% to 40%. This supports the theory that in perforated appendicitis the wound would be better left open. Cumulative results published are shown in Table 5.

Gastroduodenal Operations The incidence of postoperative wound infections following stomach operations correlates with the number of bacteria within the stomach, which in turn correlates with the acidity of the stomach and underlying pathology. The "acid barrier", which functions effectively to kill swallowed bacteria in normal individuals, may be altered by H2 receptor antagonists when given during the 12 preoperative hours. All groups of antibiotics have been used for prophylaxis following gastroduodenal operations. Table 6 summarizes the results of the best trials. In four studies, local antibiotics were used, resulting in an infection rate of 10.8% [7]. Penicillins, first- and second-generation cephalosporins, aminoglycosides and a combination of antibiotics active against anaerobic and aerobic bacteria have been used. The use of penicillins resulted in 11 infections following 72 stomach operations (15%) in three studies. Four acceptable studies with first-generation cephalosporins have been published. Four infections occurred in 136 patients (2.9%), indicating that cefazolin may be regarded as the gold standard. In seven studies [7, 11-12] the value of second- and third-generation cephalosporins has been investigated. Eleven wound infections following 205 gastroduodenal operations were observed (4%). The use of aminoglycosides did not significantly reduce infection rates. The use of combinations of two antibiotics also did not significantly improve the results. Recommended prophylaxis for gastroduodenal operations includes stopping acid secretion by H 2 blockers or neutralizing agents for one or two days preoperatively,

Infection 19 (1991) Suppl. 6 © MMV MedizinVerlag GmbHMfinchen,Mtinchen1991

S 341

D. H. W i t t m a n n et al.: Prophylaxis of Postoperative Infections

Table 8: Antibiotic prophylaxis in biliary operations.

Table 7: Bacteria isolated from the biliary tract.

Aerobes Escherichia coli Klebsiella spp. Enterobacter spp. Proteus spp. Pseudomonas spp. Other gram-negative pathogens Streptococci Enterococci Staphylococci Obligate anaerobes Bacteroides fragilis Peptostreptococci Clostridia Other anaerobes All bacteria

1,183 434 154 45 99 19 36 162 159 75 65 9 20 29 1 1,248

95 35 12 4 8 2 3 13 13 6 5 1 2 2 100

even when such agents are ordered as part of the anesthetic premedication, and administration of 2 g of cefazolin on induction of anesthesia.

Operations of the Biliary Tract Although the normal biliary tree rarely harbors any bacteria, in the presence of disease the biliary tract should be viewed as bacterially contaminated. The number of bacteria is influenced by a variety of factors; those which are associated with a risk of heavy bacterial contamination of bile are listed in Table 7. The infection risk of biliary operations without antibiotic coverage is about 15.8% in 14 studies comparing antibiotic prophylaxis to a placebo group (Table 1). Numerous clinical trials have employed a variety of antibiotic regimens in biliary operations [12-31] (Table 8). Cefazolin as a single prophylactic agent was used in six studies, and infections were seen in 26 of a total of 494 patients enrolled, resulting in an infection rate of 5.2% (range 0-10.4%). Cefamandole was used in five studies. Thirteen patients developed infections following a total of 364 bile tract operations (3.7%). In two well controlled studies, cefoxitin was used and three infections were seen in 92 patients (3.6%). In one study the infection rate was 5%, in the other 0%. The use of cefotetan resulted in an infection rate of 13 out of 90 (14.4%) in all types of biliary pathology. Cefuroxime was used in five studies, and ten infections were seen in 353 patients, resulting in a rate of 2.8%. Gentamicin was the prophylactic agent in three studies, and infections were seen in 11 out of 202 patients (5.5%). Mezlocillin was given peri-operatively in three studies, and infections were seen in 24 out of 359 patients (6.8%). One well controlled study tested the efficacy of cotrimoxazole, and two out of 48 patients developed an infection following biliary tract operations. In some studies, no infections were seen at all following prophylaxis. Third-generation cephalosporins were S 342

!!!iiiiiiiiiiiiiiiiiiiiiii!iiiiiigiigii!iiiiiiiiiiiiNNgiiii1 iilNiiiiN i iii! Cefotetan Mezlocillin Cefamandole Gentamicin Cefazolin Cefonicid Cefoxitin Cefuroxime Piperacillin Cefotaxime

1 2 5 2 7 2 2 5 1 2

90 359 431 202 494 99 92 353 50 87

13 24 24 11 26 4 3 10 1 0

14 7 6 5 5 4 3 3 2 0

utilized in four studies (cefotaxime in two, ceftriaxone in one and moxalactam in one); three infections were seen in 178 patients. All three patients were in the moxalactam group. The use of cefonicid was associated with infection in four out of 94 (4%) patients. In summary, single-injection antibiotic prophylaxis in bile tract surgery is justified. The agent of choice remains 2 g cefazolin given about 30-45 minutes prior to the operation. A second choice may be cefuroxime or cefamandole. Third-generation cephalosporins may be used in situations of local resistance. Ceftriaxone is not recommended because of its long half-life. One dose of ceftriaxone is pharmacokinetically equivalent to up to six doses of cefamandole or cefoxitin and represents multiple dose application and therefore is not justified (Table 8). Although enterococci are frequently recovered from bile, their presence can be ignored in choosing an antibiotic for prophylaxis. Cholangitis (fever, chills, jaundice) involves a more feculent spectrum of bacteria, including anaerobes. In this circumstance, ampicillin-sulbactam is preferred [34].

Vascular Operations The risk of infection is increased by the clinical settings in which most vascular reconstructions are done: electively for arterial insufficiency, implying some degree of tissue ischemia, or as an emergency for trauma. The frequent need to insert a prosthesis for bypass or further repair disables host defense mechanisms and increases the risk of infection. Overall, the risk of infection in vascular surgery is approximately 4%. The bacteria of concern are Staphylococcus aureus, Escherichia coli and other enteric gram-negative rods that cause early postoperative infections, and organisms of the Staphylococcus epidermidis group that cause indolent late graft infections. A few controlled clinical trials have been reported (Table 9), but recent increases in resistance among staphylococci, especially S. epidermidis, mandate care in the choice of antibiotic for prophylaxis. Cefazolin, administered at induction of anesthesia, is recommended as prophylaxis as long as two-thirds or more of S. epidermidis group isolates on the ward or in the institution are sensitive to this antibiotic. If S. epidermidis

Infection 19 (1991) Suppl. 6

© MMV Medizin Verlag GmbH Miinchen, Mranchen 1991

D. H. Wittmann et al.: Prophylaxis of Postoperative Infections

Table 9: Antibiotic prophylaxis in vascular operations.

[Ii Cephaloridine Cephradine Cefazolin Cefotaxime Oxacillin Cephalothin

iiii?i;i!ii!i i 1 2 1 1 1 1

27 75 225 134 168 232

6 0 2 4 4 2

22 0 1 3 2 1

resistance is a problem, ampicillin-sulbactam and trimethoprim-sulfamethoxazole are currently effective alternatives. If methicillin-resistant staphylococci are a problem, vancomycin (for the staphylococci) plus cefotaxime or aztreonam (for the coliforms) should be effective.

Aseptic (Clean) Operations Antibiotic prophylaxis is not recommended in clean

operations unless a prosthesis is to be inserted during the operation or the patient is immunosuppressed or has impaired host defense mechanisms from another cause. One of the major reasons for not recommending antibiotic prophylaxis in clean operations has been the inability to demonstrate a benefit of prophylaxis due to the low rates of infection (< 3%) which are thought to characterize clean operations [42]. The inability to demonstrate benefit, coupled with the risks of administering antibiotics, which are low but not zero for each patient and are Clearly negative for the environment, led to a judgement that the potential benefits did not outweigh the potential risks. Recent reports [43-45], however, document that for selected clean operations, the infection risk may be higher than has been assumed and that antibiotic prophylaxis may be of benefit. Much more data is needed to settle this issue. If antibiotic prophylaxis is found to be warranted in some clean operations, it will only confirm the present practice of many surgeons.

References 13. 1. Miles, A. A., Miles, E. M., Burke, J.: The value and duration of defense reactions of the skin to primary lodgment of bacteria. Br. J. Exp. Pathol. 38 (1957) 79-96. 2. Burke, J. F.: The effective period of preventive antibiotic action in experimental incisions and dermal lesions. Surgery 50 (1961) 161-168. 3. National Research Council: Postoperative wound infections: the influence of ultraviolet irradiation of the operating room and of various other factors. Ann. Surg. 160 (Suppl. I) (1964) 192. 4. Elek, S. D.: Experimental staphylococcal infections in the skin of man. Ann. NY Acad. Sci. 65 (1956) 85-90. 5. Polk Jr., H. C., Lopez-Mayor, J. F.: Postoperative wound infection: a postoperative study of determinant factors and prevention. Surgery 66 (1969) 97-103. 6. Stone, H. H., Hooper, A., Kolb, L. D., Geheber, C. E., Dawkins, E. J.: Antibiotic prophylaxis in gastric, biliary and colonic surgery. Ann. Surg. 184 (1976) 443-452. 7. Diihgen, R.: Prophylaxis of infections with antibiotics in colorectal and gastroduodenal operations and appendectomies. A validation of clinical trials (German). Dissertation, Hamburg University Medical School. Hamburg, Germany, 1988. 8. Straehan, C. J. L, Black, J., Powis, S. J., Waterworth,T. A., Wise, 1L, Wilkinson, A. R., Burdon, D. W., Severn, M., Mitra, B., Norcott, H.: Prophylactic use of cefazolin against sepsis after cholecystectomy. Br. Med. J. 1 (1977) 1254-1256. 9. Wittmann, D. H.: Antibiotic concentration in tissue fluid during the vulnerable period as rational basis for prophylaxis of post operative infections: focus on infections after operations of the colon, biliary tree and bone. In: Ishigami, J. (ed.): Recent advances in chemotherapy. University of Tokyo Press, Tokyo 1985, pp. 189-192. 10. Nichols, R. L: Bowel preparation. In: Wilmore, D. W., Brennan, M. F., Harken, A. H., Holcroft, J. W., Meakins, J. L (eds,): Care of the surgical patient. Volume 2. Scientific American, New York 1988. 11. Cainzos, M., Potel, J., Puente, J. L : Prospective randomized controlled study of prophylaxis with cefamandole in high risk patients undergoing operations upon the biliary tract. Surg. Gynecol. Obstet. 160 (1985) 27-32. 12. Bolufer, J. M., Armananzas, E., Puehades, F., Cervero, P., Trullenque, R., Cano, J.: Profilaxis antibiotica en cirugia biliar

14.

15.

16.

17.

18.

19.

20.

21. 22. 23.

24. 25.

26.

electiva? Indicacion sistematica o electiva? Rev. Quir. Esp. 14 (1987) 323-32Z Halsaii, A. K., Welsh, C. L., Craven, J. L., Hopton, D. W., Peel, R. N.: Prophylactic use of metronidazole in preventing wound sepsis after elective cholecystectomy. Br. J. Surg. 67 (1980) 55t-552. Karran, S. J., Allen, S., Lewington, V.: Cefuroxime prophylaxis in biliary surgery: In: International Congress and Symposium Series: Royal Society of Medicine. Academic Press, London, 20 (1981) 27-34. Kanfman, Z., Dinbar, A.: Single dose prophylaxis in elective cholecystectomy. Am. J. Surg. 152 (1986) 513-516. Keighley, M. R. B., Baddeley, R. M,, Burdon, D. W., Edwards, J. A. C., Quoraishi, A. H., Oates, G. D., Watts, G. T., Alexander-Williams, J.: A controlled trial of parenteral prophylactic gentamicin therapy in biliary surgery. Br. J. Surg. 63 (1975) 275-279. Kune, G. A., Hunt, R. F., Jed, A.: Wound iiffection in elective biliary surgery. Controlled trial using one dose of cephamandole. Aust. NZ J. Surg. 55 (1985) 19-22. Lewis, R. T., Goodaii, R. G., Marien, B.: Biliary bacteria, antibiotic use and wound infection in surgery of the gall bladder and common bite duct. Arch. Surg. 122 (1987) 44--47. Morran, C., NcNaught, W., McArdle, C. S.: Prophylactic cotrimoxazole in bitiary surgery. Br. Med. J. 2 (1978) 462-464. Morran, C. G., Thomson, G., White, A., NcNaught, W., Smith, D. C., McArdle, C. S.: Wound sepsis after low risk elective cholecystectomy: the effect of cefuroxime. Br. J. Surg. 71 (1984) 540-542. McArdle, C. S., Moran, C. G., Thomson, G., White, A., Piekard, R., NcNaught, W., Smith, D. C.: Prophylactic use of cefuroxime in elective biliary surgery. Res. Clin. Forums 5 (1983) 65-68. Sykes, D., Basu, P. K.: Prophylactic use of cefotaxime in elective biliary surgery. J. Antimicrob. Chemother. 14 (1984) 237-239. Elke, R., Widmer, M., Gerber, H., Trippei, M., Gruber, U. F.: Single-dose of mezlocillin for antibiotic prophylaxis in biliary surgery. Eur. Surg. Res. 15 (1983) 101-106. Lewis, R. T.: Wound infection after gastroduodenal operations: a 10-year review. Can. J. Surg. 20 (1977) 435-440. Drumm, J., Donovan, I. A., Wise, R.: A comparison of cefotetan and cefazolin for prophylaxis against wound infection after elective cholecystectomy. J. Hosp. Infect. 6 (1985) 277-280. Kellum, J. M., Duma, 1L J., Gorbaeh, S. L., Sugarman, H. J,, Haynes, B. W., Gervin, A., Newsome, H. H.: Single-dose antibiotic prophylaxis

Infection 19 (1991) Suppl. 6 © MMV Medizin Verlag GmbH Miinchen, Miinchen 1991

S 343

D. H. W i t t m a n n et al.: Prophylaxis of Postoperative Infections

27.

28.

29.

30.

31.

32. 33.

34.

35.

36.

for biliary surgery: cefazolin vs. moxalactam. Arch. Surg. 122 (1987) 918-922. Hurlow, R. A., Streaehan, C. J. L., Wise, R.: A comparative study of the efficacy of cefuroxime for preventing wound sepsis after cholecystectomy. In: International Congress and Symposiums: Royal Society of Medicine. Academic Press, London 20 (1981)1-8. Kauffman, Z., Engelberg, M., Eliashav, A., Reiss, R.: Systemic prophylactic antibiotics in elective biliary surgery. Arch. Surg. 119 (1984) 1002-1004. Maid, D. G., Lammers, J. L , Aughey, D. R.: Comparative studies of multiple dose cefoxitin vs. single dose cefonicid for surgical prophylaxis in patients undergoing biliary tract operations and hysterectomy. Rev. Infect. Dis. 6 (1984) S 887-S 895. Prada, D., Garcia, S., Ortega, J. M., Jimenez, A., Garcia, S.: Profilaxis antibiotica en cirugia biliar. Indicaciones y resultados. Rev. Esp. Enferm. Apar. Dig. 68 (1985) 111-116. Roufail, W. M.: Comparison of cefonicid and cefoxitin for prophylaxis in biliary tract surgery. Adv. Ther. 2 (1985) 225-232. Morris, D. L., Mojadeddi, Z. J., Burdon, D. W., Keighley, M. R. B.: Clinical and microbiological evaluation of piperacillin in elective biliary surgery. J. Hosp. Infect. 4 (1983) 159-164. Wilson, S. E., Hopkins, J. A., Williams, R. A.: A comparison of cefotaxime versus cefamandole in prophylaxis for surgical treatment of the biliary tract. Surg. Gynecol. Obstet. 164 (1987) 207-212. Wittmann, D. H., Koltowski, P., Oleszkiewicz, J., Walker, A. P.: Infectious complications after 809 biliary tract operations and results of a randomized single blind study comparing cefoxitin versus ampicillin plus an inhibitor of beta-lactamases. Infection 18 (1990) 41-47. Chester, J. F., Fergusson, C. M., Chant, A. D.: The effect of cephradine prophylaxis on wound infection after arterial surgery through a groin incision. Ann. R. Coll. Surg. Engl. 65 (1983) 389-390. Evans, C,, Pollock, A. V.: The reduction of surgical wound infections by prophylactic parenteral cephaloridine. A controlled clinical trial. Br. J. Surg. 60 (1973) 434-437.

S 344

37. Geroulanos, S., Oxelbark, S., Donfried, B.: Antimicrobial prophylaxis in cardiovascular surgery. Thorac. Cardiovasc. Surg. 35 (1987) 199-205. 38. Kaiser, A. B., Clayson, K. R., Mulherin, J. L, Roach, A. C., Allen, T. R., Edwards, W. H., Dale, W. A.: Antibiotic prophylaxis in vascular surgery. Ann. Surg. (1978) 283-289. 39. May, A. 1L L., Darling, C., Brewster, D. C.: A comparison of the use of cephalothin and oxacillin in vascular surgery. Arch. Surg. 115 (1980) 56--59. 40. Pitt, H. A., Postier, R. G., MacGowan, W. A. L.: Prophylactic antibiotics in vascular surgery. Topical, systemic, or both? Ann. Surg. 192 (1980) 356-364. 41. Salzmann, G.: Perioperative infection prophylaxis in vascular surgery - a randomized, prospective study. Thorac. Cardiovasc. Surg. 31 (1983) 239-242. 42. Wittmann, D. H., Garwig, J., Fliedner, E., Toenies, H.: Is antibiotic prophylaxis justified in clean operations? Results of a randomized double blind study in 374 neurosurgical operations. In: Abstracts of the 14th International Congress of Chemotherapy, Kyoto, Japan. International Society of Chemotherapy 1985, Abstract no. P-13-71, p. 313. 43. Platt, R., Zaleznik, D. F., Hopkins, C. C., Dellinger, E. P., Karehmer, A. W., Bryan, C. S, Burke, J. F., Wikier, M, A., Marino, S. K., Holbrook, K. F., Tosteson, T. D., Segal, M. R.: Perioperative antibiotic prophylaxis for herniorrhaphy and breast surgery. N. Engl. J. Med. 322 (1990) 153-160. 44. Garcia-Rodriguez, J. A., Puig-La Calle, J., Arnau, C., Porta, M., Vallv6, C.: Antibiotic prophylaxis with cefotaxime in gastroduodenal and biliary surgery. Am. J. Surg. 158 (1989) 428-434. 45. Wittmann, D. H., Bergstein, J. M., Aprahamian, C., Kiichlin, J.: Broad Spectrum coverage versus antistaphylococcal prophylaxis alone reduces bone infections after open fractures. Society for Academic Emergency Medicine, Annual Meeting 1990, Abstract no.

Infection 19 (1991) Suppl. 6 © MMV Medizin Verlag GmbH Miinchen, M~inchen 1991