Prophylactic Antibiotics in Vascular Surgery

Prophylactic Antibiotics in Vascular Surgery Topical, Systemic, or Both? HENRY A. PITT, M.D., RUSSELL G. POSTIER, M.D., WILLIAM A. L. MAcGOWAN, F.R...
Author: Dorthy Lynch
6 downloads 0 Views 1MB Size
Prophylactic Antibiotics in Vascular Surgery Topical, Systemic,

or

Both?

HENRY A. PITT, M.D., RUSSELL G. POSTIER, M.D., WILLIAM A. L. MAcGOWAN, F.R.C.S., F.R.C.S.I., LINDA W. FRANK, M.D., ANDREW J. SURMAK, M.D., JAMES V. SITZMAN, M.D., DAVID BOUCHIER-HAYES, M.CH., F.R.C.S., F.R.C.S.I.

From the Departments of Surgery, The Johns Hopkins Medical Institutions, Baltimore, Maryland and St. Laurence's Hospital, Dublin, Ireland

A prospective, randomized, blinded study was performed to determine whether prophylactic antibiotics would reduce the incidence of infection in peripheral vascular surgery and whether the route of antibiotic administration was important. Patients undergoing a vascular procedure with a groin incision were allocated to one of four groups with respect to prophylactic antibiotics. Group I received no antibiotic. Group II had topical cephradine instilled in their incisions prior to closure. Group III received a 24-hour perioperative course of intravenous cephradine, and Group IV received both topical and intravenous cephradine. Groin and abdominal incisional infections were significantly reduced (p < 0.01) among patients who received prophylactic antibiotics by either the topical, systemic, or combined routes of administration. No significant differences were noted among the three antibiotic groups. Profundoplasty, femoral embolectomy, and femoral aneurysm repair were each associated with an increased incidence of infection (p < 0.01). Other risk factors were only important in patients not receiving antibiotics. Either intraoperative topical antibiotics or perioperative systemic antibiotics prevent infection in peripheral vascular surgery, but antibiotic administration by both routes is unnecessary.

INFECTION OF A SURGICAL wound is always a costly

and distressing problem, and in vascular surgery the consequences may be particularly devastating with potential loss of limb or life. Once a synthetic arterial graft becomes infected, management may be extremely difficult, time consuming, and costly. 1-4 For these reasons, every attempt should be made to prevent infection in patients having vascular surgery. Experimental studies have demonstrated that the incidence of infection following vascular procedures may be reduced by systemic antibiotics,5 by topical antibiotics,6 or by soaking the graft in antibiotic solution.7 Although the value of systemic or topical antibiotics in reducing wound infections has been well documented in patients having various orthopedic, gynecological, and abdominal operations,8'9 only one prospective, randomized, Presented at the Annual Meeting of the American Surgical Association, Atlanta, Georgia, April 23-25, 1980. Reprint requests: Henry A. Pitt, M.D., Department of Surgery, University of California at Los Angeles, Center for the Health Sciences, Los Angeles, California 90024.

blinded study documenting the efficacy of prophylactic systemic antibiotics in patients having vascular reconstructive surgery has been published.10 Properly designed clinical studies to access whether topical antibiotics or presoaking a graft with antibiotics will reduce infection have not been reported. The importance of the groin incision as the source of the majority of synthetic graft infections has been emphasized by Szilagyi and colleagues.4 Although the wound infection rate for "clean" elective surgery should be less than 2%,11 the incidence of wound infection in the groin incision following vascular surgery has been reported to be as high as 17-26%.12-14 Various explanations have been given for this high incidence of infection including the use of prosthetic materials, prolonged operative time, the presence of ischemic tissues, and the transection of infected groin lymphatics. Since the groin incision is the source of the majority of synthetic graft infections and since this incision is associated with a relatively high incidence of infection, it would seem logical that a study designed to evaluate prophylactic antibiotics in vascular surgery should focus on procedures involving this incision. Therefore, in an effort to determine: 1) whether the incidence of infection in the groin incision could be reduced by prophylactic antibiotics, and 2) whether the route of antibiotic administration was important, a prospective, randomized, blinded study was undertaken to evaluate both systemic and topical antibiotics in patients having vascular procedures with a groin incision. Methods Patient Selection and Randomization This study was conducted from April 1977 through September 1979 at the St. Laurence's Hospital, Dublin,

0003-4932/80/0900/0356 $00.95 © J. B. Lippincott Company

356

Vol. 192 * No. 3

357

ANTIBIOTICS IN VASCULAR SURGERY

Ireland. Only patients scheduled for vascular surgery in whom a groin incision was anticipated were considered eligible for the study. Patients having abdominal aortic surgery in whom an aortoiliac bypass was considered most probable from preoperative evaluation were not included in order to minimize the number of patients having prosthetic grafts without prophylactic antibiotic coverage. Patients were also excluded from participation in the study if: 1) antibiotics had been administered within five days prior to surgery, 2) there was a significant history of penicillin or cephalosporin allergy, or 3) cellulitis or wet gangrene was present in the lower extremity. Patients with ischemic ulcers without clinical evidence of cellulitis and those with dry gangrene were included. Patients having urgent femoral embolectomy and patients with expanding femoral aneurysms without a break in the skin were also included. Patients considered eligible for inclusion were randomly allocated to receive either topical saline, topical cephradine, systemic cephradine, or both topical and systemic cephradine. During the 30-month study period, 217 patients were allocated to one of the four treatment. groups. Five patients required additional antimicrobial therapy during the first postoperative week for pulmonary or urinary tract infections, and these patients were included in the final analysis. This study was performed in accordance with the standards of the Committee on Human Experimentation of the St. Laurence's Hospital.

Antibiotic Administration Cephradine was chosen as the study drug because of its low systemic toxicity and adequate tissue distribution. In addition, a prestudy analysis revealed that cephradine provided a broader antibacterial spectrum than the other cephalosporins tested against organisms cultured from wound infections at the St. Laurence's Hospital. Patients having a vascular procedure with a groin incision were randomly assigned to one of four antibiotic groups. Group I (control) received no topical or systemic antibiotics. Group II (topical) had a solution of cephradine (1.0 g) dissolved in 25 ml of saline instilled in their incisions prior to closure but did not receive systemic antibiotics. Group III (intravenous) received cephradine (1.0 g) intravenously one hour preoperatively and every six hours thereafter for a total of four doses. Group IV (topical plus intravenous) had their incisions instilled with cephradine (1.0 g) and also received four intravenous doses of cephradine (1.0 g) starting one hour preoperatively. All incisions were irrigated with saline prior to closure, and wound closure was standardized with 3-0 Dexon being used for the deep and subcutaneous layers and 4-0 nylon for skin closure. Preoperative skin preparation was also

standardized with the operative area being scrubbed the day prior to and immediately before surgery with a providone-iodine solution. Hair clipping was performed immediately preoperatively. on

Patient Evaluation

Each patient was evaluated daily for fever, signs of allergic reaction, and wound infection. All wounds were observed for a minimum of four weeks postoperatively. Wound infections were classified in accordance with the definitions of Szilagyi, et al.4 with Class I representing superficial cellulitis or stitch abscess, Class II indicating infection of the subcutaneous tissue, and Class III denoting involvement of the graft. Wounds were considered to be infected when purulent material drained from the incision without evidence of prior ischemia of the skin edges. Evaluation of possible wound infection was made by physicians who were blinded with respect to the patient's study group. Incisions which drained lymph or sanguineous fluid which was sterile and resolved within three days were not considered to be infected. All purulent and nonpurulent drainage was cultured using standard bacteriologic methods. Patients who died within four weeks of surgery were excluded from the final analysis.

Study Monitoring A physician not directly involved with patient care or data collection served as the study monitor. Data was reviewed at intervals, and this physician participated in the decision to terminate the study in September 1979. Statistical Analysis The chi Square test was the statistical analysis employed in evaluating risk factors and the incidence of infection. Results

Study Population Of the 217 patients randomized and participating in the study protocol 12 were excluded because they died within four weeks of surgery (Table 1). Three of these patients had been randomized to Group I and received no antibiotic while the other nine patients had received cephradine. Ten of the 12 deaths occurred within the first postoperative week, and six of these patients had undergone femoral embolectomy and died as a result of their underlying disease process. The remaining six deaths were due to perioperative myocardial infarction (3 patients), ischemic bowel (1 patient), aspiration pneumonia (1 patient), and combined respiratory and

358

PITT AND OTHERS

Ann. Surg. * September 1980

TABLE 1. Characteristics of Study Population and Distribution of Risk Factors Antibiotic Prophylaxis I No Antibiotic Study population (number of patients) randomized excluded analysed Age and sex mean age (years) age > 70 (years) male (%) Risk factors (number of patients) obesity synthetic graft operation > 4 hr leg ulcer dry gangrene

II Topical Cephradine

III Intravenous Cephradine

IV Topical and IV Cephradine

All Patients

56 3 53

51 5 46

57 2 55

53 2 51

58.3 12 63

57.2 8 80

60.9 12 73

62.5 10 78

59.8 42 73

18 13 4 10 8

14 10 17 10 3

17 14 9 14 12

15 15 13 8 5

64 52 43 42 28

renal failure (1 patient). None of the 12 patients excluded from the final analysis had developed a wound infection prior to death. The mean age of the 205 patients included in the study was 59.8 years, and 42 of the study patients were 70 years of age or older (Table 1). Seventy-three per cent of the patients studied were men, and there were no significant differences among the four groups with respect to mean age, number of patients over age 70, or sex. In addition to advanced age, a number of other risk factors that might have affected the incidence of infection were evaluated. In general, these patients were not immunosuppressed, were not receiving steroids or chemotherapeutic agents, were not malnourished, and had no underlying malignancy. Despite the fact that the study population consisted mainly of patients with vascular disease of the lower extremity, the number of patients with insulin-dependent diabetes mellitus was quite low and insufficient for evaluation as a separate risk factor. The parameters that were considered as potential risk factors for infection were obesity, the presence of a synthetic graft within the groin incision, an operation lasting more than four hours, an ischemic ulcer of the lower extremity, or dry gangrene (Table 1). Patients were considered to be obese if their weight was greater than 110 per cent of ideal body weight. Patients with infected ischemic ulcers with associated cellulitis and those with wet gangrene had been excluded from randomization. However, 70 of the 205 patients included in the study had either an ischemic ulcer without cellulitis or a toe with dry gangrene. Although there was some variance in the distribution of patients having a prolonged operation or dry gangrene, the relative frequency of obesity, synthetic grafts, and leg ulcers was evenly distributed among the four study

217 12 205

groups (Table 1). In addition to these factors, the relative incidence of the various operative procedures was also evenly distributed among the study groups. Groin Wound Infection

The incidence of groin wound infection among the four study groups is presented in Table 2. These data are presented with respect to both the number of patients infected and the number of incisions infected. Only Grade II and Grade III infections are included in this analysis since these infections have the potential to or already involve a synthetic graft and, therefore, are the infections of greatest clinical significance. Sixteen of 205 patients (7.8%) developed an infection in the groin incision. Of the 16 patients who developed a groin infection 13 were in Group I (no antibiotic), and the remaining three were in Group IV (topical plus intravenous cephradine). No infections occurred in Group II (topical cephradine) or in Group III (intravenous cephradine). In 26 patients two groin incisions were performed, and in one patient in Group I (no antibiotic) both incisions became infected. Of the 231 incisions at risk there were 17 infections (7.4%). Thus, the groin wound infection rate was significantly reduced when a cephalosporin was applied topically (p < 0.01), given intravenously (p < 0.01) or administered by both routes (p < 0.01). However, no significant difference in the groin wound infection rate was noted among the topical, systemic, or combined routes of administration (Group II vs. III, II vs. IV, III vs. IV, or II + IV vs. III + IV). These conclusions were the same whether infections were considered on the basis of patients or incisions infected and also hold true if Grade I infections had been included in this analysis.

Vol. 192

*

359

ANTIBIOTICS IN VASCULAR SURGERY

No. 3

TABLE 2. Incidence of Groin Wound Infection by Route of Prophylactic Antibiotic Administration

Antibiotic Prophylaxis

Patient infections infections*/patients patients infected (%) Incisional infections infections*/incisions incisions infected (%) *

t

I No Antibiotic

II Topical Cephradine

III Intravenous Cephradine

Topical and IV Cephradine

All Groups

13/53 24.5%

0/46 O%t

0/55 O%t

3/51

5.9ott§

16/205 7.8%

14/62 22.6%

0/54 0%t

0/56 0%t

3/59 5. 1%*tt§

17/231 7.4%

Grade II or III groin wound infection. p < 0.01, no antibiotic vs topical (I vs

II

or

I

vs

II

and IV).

Other Wound Infections

The incidence of abdominal and leg as well as groin wound infection is presented in Table 3. In addition to 17 groin infections, two abdominal incisions and three leg incisions became infected. One of the two patients with an abdominal infection also had a groin wound infection. The three leg infections each involved only one of several leg incisions with no involvement of the groin incision. Both abdominal and groin infection rates among those patients receiving no antibiotic (Group I) were significantly higher (p < 0.01) than among patients receiving cephradine (Groups II, III, and IV). The infection rate for all patients was greatest in groin incisions (7.4%), less in abdominal incisions (5.0%), and least in leg incisions (0.8%). Synthetic Grafts In 52 patients synthetic graft material was present within the groin incision (Table 4). Twenty-four of these patients had two groin incisions, placing 76 incisions at risk for infection. Four of these 52 patients (7.7%) developed an infection in the groin incision, and this infection rate was almost identical to the infection rate (7.8%) noted among the 153 patients studied in whom no synthetic material was present in the groin incision. Three of the four infections developing in patients with a synthetic graft occurred in Group I (no antibiotic)

IV

t

p < 0.01, no antibiotic vs intravenous (I vs III or I vs III and IV).

§

p