Prophylactic antibiotics and wound infections following laminectomy for lumber disc herniation NORMAN H. HORWITZ, M.D., AND JAMES A. CURTIN, M.D.

Transcription

1 Prophylactic antibiotics and wound infections following laminectomy for lumber disc herniation A retrospective study NORMAN H. HORWITZ, M.D., AND JAMES A. CURTIN, M.D. Departments of Neurosurgery and Medicine of the Washington Hospital Center and the George Washington University School of Medicine, Washington, D.C. v" The authors review 531 consecutive operations for lumbar disc herniation performed on 496 patients by one neurosurgeon to determine the effect of prophylactic antibiotics upon postoperative wound infections. In this retrospective analysis 16 instances of sepsis were found, 11 considered to be major and five minor. In the 128 cases in which no antibacterial agents were given, 11 major and 1 minor infection occurred. Four minor infections developed in the 402 occasions when antibiotics were given in the perioperative period. Men had a significantly greater risk of developing infection than women. These data suggest that pre- and postoperative antibiotic therapy directed at a narrow spectrum of microorganisms reduced the incidence of significant wound infections in patients undergoing laminectomy for lumbar disc herniation. KEVWORDS 9 lumbarlaminectomy 9 infection 9 prophylaxis 9 disc 9 lincomycin T HIS paper reviews the occurrence of postoperative wound infections in individuals operated on by one of us (N.H.H.) for lumbar disc herniation, and the role of prophylactic antibiotics administered. Clinical Material In the 151/2-year period ending in 1973, 531 consecutive operations for lumbar disc herniation were performed on 496 patients by one neurosurgeon at the Washington Hospital Center. In 81 instances, spine fusion was performed by various orthopedists following the disc exploration. The surgery was performed under general anesthesia, with the patient in the prone position. No drains were employed in the laminectomy wounds, but on occasion subcutaneous drains were used in separate donor sites by the orthopedists. The laminectomy wounds were closed with nonabsorbable suture material (steel wire, silk, or synthetics). In addition to being fully responsible for the aftercare of those having disc removal alone, the neurosurgeon was closely involved in the postoperative supervision of the patients subjected to fusion. Incisions were examined at frequent intervals and cultures taken of any exudate, however slight. The period of office follow-up ranged from a minimum of 3 months to many years in some J. Neurosurg. / Volume 43 / December,

2 TABLE 1 Clinical data relating to 16 patients with postoperative infection N. H. Horwitz and J. A. Curtin Infection Prophylactic Case No. Age, Sex Fusion Month/Yr Ctdtured Pathogen Major Minor Antibiotics 1 47 M 0 11,/60 Staph. aureus M 0 2/62 Staph. aureus + -k M 0 9/62 Staph. aureus M Jr- 10/62 Staph. aureus M + 9/63 Beta-hemolytic strep. -Jr M 0 12/63 Staph. epidermidis + 0 7* 60 M 0 1/64 Staph. aureus M q-- 1/64 Staph. aureus q F 0 9/64 Staph. aureus q- 0 10" 33 M 0 10/64 Staph. aureus Jr M 0 6/65 Staph. aureus -Jr- q M 0 7/65 Beta-hemolytic strep, q M 0 4/67 E. coil -Jr M q-- 5/68 Staph. aureus, Beta- q- 0 hemolytic strep., Klebsiella M 0 7/68 Staph. aureus M q- 10/70 Klebsiella- "Jr- q- enterobacter * Denotes patient with diabetes. instances. No cases manifesting infection after hospitalization are known to have been treated elsewhere. The follow-up period terminated on October 1, There was one postsurgical death in the entire series; this occurred 20 days after operation, from a pulmonary embolus. The patient was on the urology service, having undergone cystoscopy and urethral dilatation 9 days before death. Nevertheless, it was considered a neurosurgical mortality. This case is not included in the statistical analysis that follows. The antibiotics used varied in the early years of this investigation, and included chloramphenicol, erythromycin, and penicillin. The antibiotic of choice in the years 1965 to 1967 was oxacillin, but since 1968 we have used exclusively lincomycin. In the period before lincomycin became available, medication was begun 24 to 36 hours before surgery and was resumed within 3 hours of the completion of the procedure, intramuscularly or orally, every 6 hours for 5 days. As will be detailed later, some patients received only postoperative drugs. With lincomycin, a consistent pattern of administration was eventually established, namely, 600 mg intramuscularly 8 hours and 2 hours before the operation. After surgery, 600 mg was given intramuscularly 6 and 12 hours respectively following the last preoperative dose, which meant that the first injection after surgery usually was received within 2 hours of the completion of the operation. An oral continuation of 500 mg every 6 hours was pursued thereafter for 3 to 4 days. No intraoperative medication was given. Results Sixteen instances of sepsis, confirmed by culture, were identified, an incidence of 3%. These were divided into two clinical categories, major and minor. The former group comprised superficial or deep wound infections which required prolonged aftercare and in some instances incision and drainage with subsequent secondary wound closure. The latter group was made up of individuals whose wounds healed per primam, but whose incisions manifested slight amounts of drainage from stitch openings which cultured pathogens. Induration and low-grade fever sometimes were present, but the hospital stays were not prolonged beyond a few days and the infection did not substantially alter the postsurgical course or ultimate outcome. Eleven cases (2%) were considered to be major infections and five (1%) designated minor (Table 1). Fifteen of these patients 728 J. Neurosurg. / Volume 43 / December, 1975

3 Antibiotic prophylaxis after lumber disc surgery were males, including all those with major infections, despite the fact that 163 (33%) of the total patient population were females. Five of the 16 patients were subjected to fusion procedures, representing an infection rate of 6.2% as against 2.4% in instances when no fusion followed. Staphylococcus aureus, coagulase positive, was present in 11 of the cultures, beta-hemolytic streptococcus in three, and klebsiella in two. Neither the patient's age nor the mode of skin preparation used in the operating room seemed important variables. Two of the infected individuals had mild diabetes mellitus. The length of time required to perform the surgery and its relation if any to subsequent sepsis is unclear in this review. While it is true that those with fusions whose wounds became infected underwent surgery for periods in excess of 3 hours, five others who had disc removals alone required less than 2 hours for the completion of the procedure. Sepsis was most frequently encountered during the months of September and October. Table 2 is a summary of the number of operations performed during each of the years under consideration, the percentage of patients receiving antibiotics, and the resulting infection rate. During the study period, 128 patients received no prophylactic antibacterial therapy. Of the 402 who received such therapy, 320 received both preand postoperative drugs, and 75 were given medications only in the postoperative period. Three patients received preoperative antibiotics, but through oversight, none after surgery. The records of four individuals are incomplete as to dosage schedule and agent employed. The infection rate for the 402 patients receiving antibiotics was 1%. Of the 323 patients to whom medication was administered both pre- and postsurgery, including the three receiving preoperative dosage only, the infection rate was 0.6%, two minor infections. One of these occurred in one of the three patients to whom postoperative medications were not given. For the 75 patients receiving only postsurgical treatment, the rate of infection was 2.7%. In the 128 instances in which no effective antibacterial program was instituted, the rate was 9.3% for wound infection. All 11 in the major category and one in the minor group occurred in these untreated patients. A Year TABLE 2 Anmtal incidence of infection durit;g the study period Cases % Receiving Infections Prophylactic Operated Antibiotics No. Rate (~) I t l ! I scrutiny of the 398 complete protocols among the 402 receiving antibiotics reveals a mention of loose stools or frank diarrhea in 15.1%; in the 265 patients receiving lincomycin, the incidence of bowel disturbance rose to 20.4%. This complication was not usually a serious event and subsided promptly after cessation of the medication. However, a few severe reactions necessitated the assistance of gastrointestinal consultants and the hospital stay was prolonged. One patient required readmission because of gastroenteritis attributed to her antibiotics. She and others responded to an appropriate remedial regimen, and none suffered any permanent disability from this side-effect. No other known drug-induced complication occurred. Discussion According to the chi-square method of analysis, the foregoing data clearly indicate a significant statistical difference between the treated and untreated patients. This applies specifically to the heightened rate of all infections in those not receiving drug protection as well as to the incidence of major infections only in the untreated group as compared to the category of antibiotic recipients. Likewise, the male sex dominance among infection victims is sustained statistically. There was no effort on the part of the surgeon to J. Neurosurg. / Volume 43 / December,

4 select patients for therapy based on physical condition. In the years through 1968, the decision as to therapy was related to the current beliefs about the value of a program of antibacterial prophylaxis in clean surgical procedures. In the last 5 years an effort has been made to treat every patient with pre- and postoperative lincomycin. During this time we treated 240 cases; 238 of these received the drug as outlined above; one received it only postoperatively through oversight, and one individual received antibiotics but the complete record could not be retrieved. One minor infection occurred in this 5-year period, an incidence of 0.4%. Table 2 shows that in the years 1958, 1959, 1961, and 1966, 100% of patients operated on received antibiotics without the appearance of a single infection. In the first three of these years, the antibiotics were given only postoperatively. Comparison of our postoperative infection rate with those of others shows some similarity. Wright '7 studied the figures at the Massachusetts General Hospital from 1952 to He found a 3.1% infection rate for disc surgery alone and 8.2% for those undergoing fusion as well. During part of this time, postoperative penicillin and streptomycin were tried. Donor site infections were not included as they were in our series. He also summarized the available literature up until that time. The highest recorded infection rate for disc surgery alone was 8%, ~ and for disc with fusion, 8.3%. TM The lowest rate was 0.6% for disc removal alone, 5,7,'3 and 0.9% for disc removal combined with fusion. ~8 Of particular interest is that deaths from sepsis following disc surgery do occur. Waris 15 recorded five deaths in 374 operations for lumbar disc, a mortality figure of 1.3%. These patients' wounds were drained and the author felt that the practice contributed to the outcome. The concept in the use of antibiotics to prevent infection is based upon their ability to develop adequate levels in the target tissues of an antimicrobial agent to combat a specific sensitive organism. Attempts to avert infection by the use of broad-spectrum antibiotics for an extended period of time are almost invariably unsuccessful. The prototype of effective prophylaxis is the use of penicillin or other antimicrobial substances to prevent streptococcal sore throat and hence reduce the incidence of primary or recurrent N. H. Horwitz and J. A. Curtin rheumatic fever. 2,' This principle can be applied to surgical procedures in which it is recognized that certain specific organisms are involved in perioperative contamination. However, earlier reports of the lack of effectiveness of postoperative antibiotics 9'1~ led many to condemn any regimen in clean surgical cases that provided for antibacterial adjuncts. Nevertheless, the experimental work of Burke 1 suggested that pre- and intraoperative antibiotics would be effective since significant blood levels were necessary within 3 hours of wound contamination in order to prevent sepsis. More recent studies in the clinical field embodying this principle support his thesis: Fogelberg, et al., 3 noted a significant reduction in infections in clean orthopedic cases (arthroplasties and spine fusions) as compared to controls, by giving 600,000 units of procaine penicillin intramuscularly at 8 PM and 6 AM prior to surgery and 5,000,000 units intravenously during surgery and 600,000 units intramuscularly every 6 hours for 8 days afterward. In a double-blind study of 150 clean orthopedic cases, Pavel, et al, 8 report a significant reduction of infection in those given 1 gm cephaloridine intramuscularly 1 hour before surgery and 1 gm intravenously during the procedure. Savitz, et al., ~2 compared the year , when clean neurosurgical cases were given only ampicillin postoperatively for 10 days, with a similar number of cases in when lincomycin was given preoperatively and intraoperatively. The respective infection rates were 5.1% and 2.3%. The authors concluded that in clean cranial and spinal neurosurgical cases, one or two doses of a specific antistaphylococcal agent given immediately preand intraoperatively is the equivalent of 40 doses of a broad-spectrum antibiotic administered over the first 10 days postoperatively. It is our view that the role of prophylactic antibiotics in clean laminectomy procedures for disc removal should be reassessed in the light of the very recent contributions of the foregoing authors and the retrospective analysis of our own case material. We recognize a contemporary skepticism concerning the value of retrospective studies. That views on this matter, too, may change is suggested by a current defense of these studies by Sartwell. ~ 730 J. Neurosurg. / Volume 43 / December, 1975

5 Antibiotic prophylaxis after lumber disc surgery References I. Burke JF: Pre-operative antibiotics. Surg Clin N Am 43: , Denny FW, Wannamaker LW, Brink WR, et al: Prevention of rheumatic fever. Treatment of the preceding streptococcic infection. JAMA 143: , Fogelberg EV, Zitzmann EK, Stinchfield FE: Prophylactic penicillin in orthopedic surgery. J Bone Joint Surg (Am) 52:95-98, Frank PF, Stollerman GH, Miller LF: Protection of military population from rheumatic fever. Routine administration of benzathine penicillin G to healthy individuals. JAMA 193: , Gurdjian ES, Ostrowski AZ, Hardy WG, et al: Results of operative treatment of protruded and ruptured lumbar discs. Based on 1176 operative cases with 82 percent follow-up of 3 to 13 years. J Neurosurg 18: , Malmros R: Den lumbale discusprolaps og ligamentaeve rodkompression. Diss KCebenhavn, E Munksgaard, Odom GL, Hart D, Johnson P, et al: Neurosurgical operation infections. A seventeen year survey of the use of ultraviolet radiation. Presented at the 23rd meeting of the American Academy of Neurological Surgeons, New Orleans, La., November, Pavel A, Smith RL, Ballard A, et al: Prophylactic antibiotics in clean orthopedic surgery. J Bone Joint Surg (Am) 56: , Prothero SR, Parkes JC, Stinchfield FE: Complications after low-back fusion in 1000 patients. A comparison of two series one decade apart. J Bone Joint Surg (Am) 48:57-65, Sanchez-Ubeda R, Fernand E, Rousselot LM: Complication rate in general surgical cases. The value of penicillin and streptomycin as postoperative prophylaxis. A study of 511 cases. N Engl J Med 259: , Sartwell PE: Retrospective studies. A review for the clinician. Ann Int Med 81: , Savitz MH, Malis LI, Meyers BR: Prophylactic antibiotics in neurosurgery. Surg Neurol 2:95-100, Shinners BM, Hamby WB: The results of surgical removal of protruded lumbar intervertebral discs. J Neurosurg 1: , Stevens DB: Postoperative orthopedic infections. A study of etiological mechanisms. J Bone Joint Surg (Am) 46:96-102, Waris W: Lumbar disc herniation. Clinical studies and late results of 374 cases of sciatica operated on the diagnosis or suspicion of lumbar disc herniation. Acta Chir Scand [Suppl] 140:1-134, White JC: Results in surgical treatment of herniated lumbar intervertebral discs: investigation of the late results in subjects with and without supplementary spinal fusion -- a preliminary report. Clin Neurosurg 13:42-51, Wright RL: Septic Complications of Neurosurgical Spinal Procedures, Springfield, Ill, Charles C Thomas, 1970 Address reprint requests to." Norman H. Horwitz, M.D., Room 2A-60, Washington Hospital Center, 110 Irving Street, N.W., Washington, D.C J. Neurosurg. / Volume 43 / December,