J Neurosurg 73:383-386, 1990
Antibiotic prophylaxis during prolonged clean neurosurgery Results of a randomized double-blind study using oxacillin MICHEL DJINDJIAN, M.D., ELISABETH LEPRESLE, M.D., AND JEAN-BERNARD HOMS, M.D. Neurosurgery Service and Department of Anesthesia, Hbpital Henri Mondor, Crbteil, France
u- The efficacy of oxacillin as a prophylaxis for infection was analyzed in a 27-month randomized doubleblind study of 400 patients who had undergone clean neurosurgical interventions lasting longer than 2 hours. Four neurosurgeons took part in the study and 356 patients were eligible for final analysis. Among the 171 patients treated with oxacillin, there was one case of infection (0.6%), compared to nine (4.9%) of the 185 patients given a placebo. The difference between the two groups was statistically significant (p = 0.0398). This study, together with others (randomized or not), clearly demonstrates the efficacy of antibiotic prophylaxis in prolonged clean neurosurge~'. KEY WORDS
"
antibiotic prophylaxis
HE majority of relevant publications indicate that antibiotic prophylaxis is effective in clean neurosurgery. However, although a recent review by Haines s suggests that a "clear and definitive response was given," we believe that the latest publications leave certain points open to discussion. In 1986, we reported on the efficacy of antibiotic prophylaxis using oxacillin in shunt procedures 6 and suggested the possible extension of prophylaxis with oxacillin for use in clean neurosurgical procedures, excluding operations for shunt placement and patients at low risk of infection (short surgical interventions).
T
Clinical Material and Methods
Four senior surgeons (A, B, C, and D) participated in the study which was carried out from October, 1986, to January, 1989, with a minimum observation period of 3 months. The results were collated in April, 1989. Only patients scheduled for surgery lasting 2 hours or longer were included. All minor neurosurgical procecures, for example, those involving sciatica and peripheral nerves, were excluded while stereotactic operations and surgery for cerebral tumors and aneurysms were considered. Interventions via the rhinoseptal route were omitted, but those patients in whom the frontal or mastoid sinus was involuntarily opened were included J. Neurosurg. / Volume 73 /September, 1990
9 wound infection
9 oxacillin
.
prophylaxis
in a "clean contaminated" group and were analyzed separately. The surgeons insured that preparation for surgery was identical in every patient. This included shaving the patient on the evening before surgery, disinfecting with povidone-iodine (Betadine), and use of plastic adhesive at the operation site (surgical barrier*). Redon suction drainage, used where necessary, was removed after 24 hours and systematically cultured. The data recorded included: the identify of the surgeon; preoperative prescription of corticosteroids; duration of the procedure; use, if any, of a microscope or Cavitron ultrasonic surgical aspirator (CUSA); the pathology involved; and any opening of the sinuses. Infection was defined by the criteria proposed by Malis ~~ and was confirmed jointly by the surgeon directing the study and one of the two intensive-care workers. The usual criteria for exclusion from analysis were applied. We omitted from the data allergic patients, those who had been prescribed antibiotics in the 7 days preceding surgery, those with planned opening of the frontal or mastoid sinus, children weighing less than 3 kg, patients subjected to preoperative ventricular * Surgical barrier manufactured by Johnson and Johnson, New Brunswick, New Jersey. 3813
M. Djindjian, E. Lepresle, and J. B. Homs TABLE 1
TABLE 2
Reasons for exclusion of patients from final analysis"
Clinical summary of 356 cases in this study
Reason for Exclusion total cases lost to follow-upreview died < 90 days postop (no sepsis) protocol deviations total cases excluded
Oxacillin Group 198 4 7
Placebo Group 202 2 5
16 27
10 17
drainage or in whom the drain had been removed less than 48 hours before surgery, patients undergoing surgery in the casualty operating theater, and those treated by stereotactic surgery under a scanner. Oxacillin or a placebo identical in appearance was prescribed according to a previously generated randomized list and administered in 1-gm injectable doses. Administration began upon induction of anesthesia, at a dose of 200 mg/kg/24 hrs by direct intravenous injection every 4 hours; the total treatment lasted 24 hours.
Results Initially, 400 patients were included in this study, but 44 patients were omitted from evaluation for reasons summarized in Table 1. Table 2 presents a summary of the 356 patients for whom results could be analyzed. No side effects such as seizures were observed during the study. In the group treated with oxacillin, infection was noted in only one case (0.6%), compared with nine cases (4.9%) in the control group (Table 3). The difference was statistically significant (chi-squared test with Yates' correction: R = 4.22, p = 0.0398). All infections were deep except for one case of superficial infection following stereotactic surgery. Ten cases of meningitis without signs of infection were observed in each group and were systematically treated by a short course of antibiotics. In accordance with the criteria of Malisfl ~ these patients were not scored as infected (aseptic meningitis).
Discussion There have been m a n y publications on antibiotic prophylaxis in clean neurosurgery since 1979, when Malis ~~ reported the absence of infection in 1732 patients given a combination of intravenous vancomycin, intramuscular tobramycin, and topical streptomycin. A recent article by Haines 8 reviewed the subject in detail. While it is not necessary to repeat the same detailed discussion here, certain important points should be stressed: 1. The current goal of antibiotic prophylaxis in clean neurosurgery is to keep the infection level at less than the 1% achieved by Cushing 4 at the beginning of the century using nothing but soap and water. 2. The percentage infection in the absence of antibi384
Feature sex (M:F) duration of surgery 2-4 hrs 4-6 hrs > 6 hrs open sinuses type of surgery cerebral tumor (meningioma vascular posterior fossa spinal stereotactic other totals * No incidence of sepsis.
Oxacillin Group 90:81
Placebo Group 101:84
132 27 12 10*
148 25 12 6*
52 17 28 29 20 39 3 171
44 16) 22 41 25 46 7 185
otic prophylaxis as reported in the literature varies from 1% to 7% among series, but also, significantly, from year to year within any one series? 3 3. The risk of infection in clean neurosurgery varies according to the pathology involved.ll'~5 It also depends on a number of factors associated with infection such as shunt procedures, craniotomy as compared to spinal surgery, long operations, the necessity for drainage, 2~ the use of a microscope, CUSA, or other such device, certain pathological entities (glioma), and t u m o r recurrences.18 For this study, operations of predicted length of 2 hours or more were selected, thereby excluding low-risk interventions such as sciatica or peripheral nerve neurosurgery. 4. All studies have found the same distribution of infectious agents, with about 85 % Gram-positive bacteria of which 80% are Staphylococci. All the antibiotics tested therefore are anti-staphylococcal and have subsequently shown themselves to be 85% effective in single-drug or multidrug therapy. 12.2 5. The majority of nonrandomized studies demonstrate the efficacy of antibiotic prophylaxis. The following figures for the incidence of infection following surgery with antibiotic prophylaxis have been reported by authors using either Malis' antibiotic treatment or another single-drug therapy: Malis, ~~ 0%; Haines and Goodman, 9 0.9%; Savitz and Malis, ~60%; Quartey and Polyzoidis, ~3 0.8%; Urciuoli, et aL, ~9 0%; and Malis as cited by Dempsey, et al.,S 0.18 %. However, three studies indicate the ineffectiveness of antibiotic prophylaxis: Quadery, et al., 12 who used penicillin/sulfadiazine; Wright, 21 who used penicillin/streptomycin; and Raggueneau, et al., ~4 who used cephalothin (except for operations lasting more than 5 hours). Nonetheless, only the third study described true antibiotic prophylaxis. 6. Three randomized single-blind studies led to the conclusion that antibiotic prophylaxis is effective: GerJ. Neurosurg. / Volume 73/September, 1990
O x a c i l l i n p r o p h y l a x i s in p r o l o n g e d c l e a n n e u r o s u r g e r y TABLE 3 Details of sepsis in 10 patients with postoperative infection*
Surgeon (A, B, C, or D) A B B A B, C D C
Etiology
Age Durationof (yrs) & Intervention Sex
cerebral menin54, M gioma glioblastoma(stereo) 65, F glioblastoma (stereo) 68, F
4 hrs
cerebral meningioma cervicalneurinoma medulloblastoma cerebral AVM
74, F 40, F 37, M 15, F
Type of Infection
Isolated MicroOrganisms
Treatment Group
3 hrs
woundinfection, osteomyelitis of bone flap woundinfection woundinfection, empyema (died) meningitis
< 1 wk
S. epidermidis, MS
placebo
5 hrs 4 hrs 6 hrs
extraduralempyema meningitis ventriculitis
< 1 wk < 1 wk < 1 wk
S. aureus, MS Enterobacter cloacae Acinetobacter& S. aureus, MS Streptococcus S. epidermidis, MS
placebo placebo placebo placebo placebo
?
oxacillin
2 hrs 2 hrs
A A
spinal ependymoma 54, M 8 hrs meningitis(died) cerebral menin63, M 6 hrs woundinfection, osteomyelitis gioma of bone flap A-D cerebralAVM 54, F 8 hrs empyema * AVM = arteriovenous malformation.S. = Staphylococcus; MS = methicillin susceptible.
aghty and Feely, 7 who used Malis' treatment (0.5% vs. 3.5% in the control group); Young and Lawner, 22 who used cefazolin/gentamicin (0.73% vs. 4.09%); and Blomstedt and Kytt~i,2 who used vancomycin (2% vs. 8%). 7. Four randomized double-blind studies comparing antibiotics against placebo have been reported. Two of the four studies were stopped before completion due to the elevated number of infections in the control group (Savitz and Malis, ~6clindamycin vs. placebo, 1.2% and 10.9% infections, respectively; ~6 and Shapiro, et al., ~7 who used Malis' antibiotic prophylaxis treatment with 2.8% infection compared to 11% infection with placebo). In the two other completed studies (Bullock, et al., 3 who used pipericillin, and van Ek, et al., 2~ who used cloxacillin), antibiotic prophylaxis was shown to be effective: 2.1% infection vs. 5.9% and 3.3% vs. 10.3%, respectively. 8. All of the trials, in particular the four randomized double-blind studies, included operations for shunt placement. In our view, this affected the results. The placement of a shunt is known to involve a risk of infection higher than for neurosurgery in general. The use of antibiotic prophylaxis has proved effective in neurosurgery involving only shunt procedures. J,6 If the data from those subjects who had shunt procedures are excluded from the two completed randomized doubleblind studies, 3'2~the difference between the percent of infections in the treated group and the placebo group is not statistically significant. For this reason, patients undergoing shunt procedures have been excluded from the present study. 9. Our results, derived from patients undergoing long surgical interventions, show a drop in the level of infection from 4.9% in the placebo group to 0.6% in the group with antibiotic prophylaxis and give inconJ. Neurosurg. / Volume 7 3 / S e p t e m b e r , 1990
Time Between Surgery & Sepsis
3 wks S. epidermidis, MS
placebo
4 wks ? 4 wks S. epidermidis, MS
placebo placebo
& Streptococcus
< 1 wk < 1 wk 10 wks
testable support for antibiotic prophylaxis. The level of infection in our placebo group (nine cases, 4.9%) was consistent with that reported in the literature in the absence of antibiotic prophylaxis. Severe risk factors for infection were present among these nine infected patients: a very long operation (> 4 hours) in six cases; craniotomies in seven cases; high-risk pathology (meningiomas) in three patients; age over 50 years in six patients; and the use of a microscope in five and CUSA in six procedures. Six of the nine infections in the placebo group could be considered severe. There were two deaths due to infection and three repeat operations, including two flap ablations and one serious sequela (Table 3). These few cases show that the risk from infection is serious, both in terms of malfunction and threat to life, even for a neurosurgeon who maintains a low level of infection (< 2%, for example). Antibiotic prophylaxis is therefore justified. Despite the efficacy of antibiotic prophylaxis, several points are worth further consideration: 1) With few exceptions, infections were persistent in all reported studies. The same infectious agents are found, predominately S t a p h y l o c o c c i (76%), mainly S. a u r e u s , which appears not to be resistant to antibiotics. 3'7'9'22It should be noted, therefore, that rigorous aseptic practices are extremely important to combat S t a p h y l o c o c c i . 2) Other factors should be further considered for their role in risk of infection particularly the issues of shaving, the use of adhesive barriers at operation sites and drains, and the choice of antiseptics. 3) The occurrence of methicillin-resistant S t a p h y l o c o c c i has been largely overcome in recent years and does not appear to be a problem in clean neurosurgery, although it may become one. 4) The choice of antibiotic agent and the selection of a single-drug or multidrug therapy protocol must be 385
M. Djindjian, E. Lepresle, and J. B. Horns made on a patient-by-patient basis. The issue of antibiotic prophylaxis is inextricably linked to that of aseptic practices, which must be considered first. Our preferred choice of antibiotic treatment for interventions of more than 2 hours' duration is simple and inexpensive with the broad-spectrum antibiotic oxacillin, which we have used successfully during shunt procedures. 6 References 1. Blomstedt GC: Results of trimethoprim-sulfamethoxazole prophylaxis in ventriculostomy and shunting procedures. A double-blind randomized study. J Neurosurg 62:694-697, 1985 2. Blomstedt GC, Kytt/i J: Results of a randomized trial of vancomycin prophylaxis in craniotomy. J Neurosurg 69: 216-220, 1988 3. Bullock R, van Dellen JR, Ketelbey W, et al: A doubleblind placebo-controlled trial of perioperative prophylactic antibiotics for elective neurosurgery. 3 Neurosurg 69: 687-691, 1988 4. Cushing H: Concerning the results of operations for brain tumor. 3AMA 64:189-195, 1915 5. Dempsey R, Rapp RP, Young B, et al: Prophylactic parenteral antibiotics in clean neurosurgical procedures: a review. J Neurosurg 69:52-57, 1988 6. Djindjian M, Fevrier MJ, Otterbein G, et al: Oxacitlin prophylaxis in cerebrospinal fluid shunt procedures: results of a randomized open study in 60 hydrocephalic patients. Surg Neurol 25:178-180, 1986 7. Geraghty J, Feely M: Antibiotic prophylaxis in neurosurgery. A randomized controlled trial. J Neurosurg 60: 724-726, 1984 8. Haines SJ: Efficacy of antibiotic prophylaxis in clean neurosurgical operations. Neurosurgery 24:401-405, 1989 9. Haines SJ, Goodman ML: Antibiotic prophylaxis of postoperative neurosurgical wound infection. J Neurosurg 56: 103-105, 1982 10. Malis LI: Prevention of neurosurgical infection by extraoperative antibiotics. Neurosurgery 5:339-343, 1979 11. Mollman HD, Haines SJ: Risk factors for postoperative neurosurgical wound infection. A case-control study. J
386
Neurosurg 64:902-907, 1986 12. Quadery LA, Medlery AV, Miles J: Factors affecting the incidence of wound infection in neurosurgery. Acta Neurochir 39:133-141, 1973 13. Quartey GRC, Polyzoidis K: Intraoperative antibiotic prophylaxis in neurosurgery: a clinical study. Neurosurgery 8:669-67t, 1981 14. Raggueneau JL, Cophignon J, Kind A, et al: Analyse des suites infectieuses de 1000 interventions neurochirurgicales. Incidence de l'antibiothrrapie prophylactique. Nenrochirurgie 29:229-233, 1983 15. Savitz MH, Katz SS: Prevention of primary wound infection in neurosurgical patients: a 10-year study. Neurosurgery 18:685-688, 1986 16. Savitz MH, Malis LI: Prophylactic clindamycin for neurosurgical patients. NY State J Med 76:64-67, 1976 17, Shapiro M, Wald U, Simchen E, et al: Randomized clinical trial of intra-operative antimicrobial prophylaxis of infection after neurosurgical procedures. J Hosp Infect 8:283-295, 1986 18. Tenney JH, Vlahov D, Salcman M, et al: Wide variation in risk of wound infection following clean neurosurgery. Implications for perioperative antibiotic prophylaxis, J Neurosurg 62:243-247, 1985 19. Urciuoli R, Lo Russo G, Zeme S, et al: Traitement antibiotique prophylactique des complications infectieuses post-oprratoires en neurochirnrgie. Agressologie 24:237-238, 1983 20. van Ek B, Dijkmans B, van Dulken H, et al: Antibiotic prophylaxis in craniotomy. A prospective double-blind controlled study. Scand J Infect Dis 20:630-639, 1988 21. Wright RL: A survey of possible etiologic agents in postoperative craniotomy infections. J Neurosurg 25: 125-132, 1966 22. Young RF, Lawner PM: Perioperative antibiotic prophylaxis for prevention of postoperative neurosurgical infections. A randomized clinical trial. J Neurosurg 66: 701-705, 1987 Manuscript received October 12, 1989. Accepted in final form February 6, 1990. Address reprint requests to: Michel Djindjian, M.D., Service de Neurochirurgie, HSpital Henri Mondor, 94000 Crrteil, France.
.I. Neurosurg. / Volume 73/September, 1990
Antibiotic prophylaxis during prolonged clean neurosurgery Results of a randomized double-blind study using oxacillin MICHEL DJINDJIAN, M.D., ELISABETH LEPRESLE, M.D., AND JEAN-BERNARD HOMS, M.D. Neurosurgery Service and Department of Anesthesia, Hbpital Henri Mondor, Crbteil, France
u- The efficacy of oxacillin as a prophylaxis for infection was analyzed in a 27-month randomized doubleblind study of 400 patients who had undergone clean neurosurgical interventions lasting longer than 2 hours. Four neurosurgeons took part in the study and 356 patients were eligible for final analysis. Among the 171 patients treated with oxacillin, there was one case of infection (0.6%), compared to nine (4.9%) of the 185 patients given a placebo. The difference between the two groups was statistically significant (p = 0.0398). This study, together with others (randomized or not), clearly demonstrates the efficacy of antibiotic prophylaxis in prolonged clean neurosurge~'. KEY WORDS
"
antibiotic prophylaxis
HE majority of relevant publications indicate that antibiotic prophylaxis is effective in clean neurosurgery. However, although a recent review by Haines s suggests that a "clear and definitive response was given," we believe that the latest publications leave certain points open to discussion. In 1986, we reported on the efficacy of antibiotic prophylaxis using oxacillin in shunt procedures 6 and suggested the possible extension of prophylaxis with oxacillin for use in clean neurosurgical procedures, excluding operations for shunt placement and patients at low risk of infection (short surgical interventions).
T
Clinical Material and Methods
Four senior surgeons (A, B, C, and D) participated in the study which was carried out from October, 1986, to January, 1989, with a minimum observation period of 3 months. The results were collated in April, 1989. Only patients scheduled for surgery lasting 2 hours or longer were included. All minor neurosurgical procecures, for example, those involving sciatica and peripheral nerves, were excluded while stereotactic operations and surgery for cerebral tumors and aneurysms were considered. Interventions via the rhinoseptal route were omitted, but those patients in whom the frontal or mastoid sinus was involuntarily opened were included J. Neurosurg. / Volume 73 /September, 1990
9 wound infection
9 oxacillin
.
prophylaxis
in a "clean contaminated" group and were analyzed separately. The surgeons insured that preparation for surgery was identical in every patient. This included shaving the patient on the evening before surgery, disinfecting with povidone-iodine (Betadine), and use of plastic adhesive at the operation site (surgical barrier*). Redon suction drainage, used where necessary, was removed after 24 hours and systematically cultured. The data recorded included: the identify of the surgeon; preoperative prescription of corticosteroids; duration of the procedure; use, if any, of a microscope or Cavitron ultrasonic surgical aspirator (CUSA); the pathology involved; and any opening of the sinuses. Infection was defined by the criteria proposed by Malis ~~ and was confirmed jointly by the surgeon directing the study and one of the two intensive-care workers. The usual criteria for exclusion from analysis were applied. We omitted from the data allergic patients, those who had been prescribed antibiotics in the 7 days preceding surgery, those with planned opening of the frontal or mastoid sinus, children weighing less than 3 kg, patients subjected to preoperative ventricular * Surgical barrier manufactured by Johnson and Johnson, New Brunswick, New Jersey. 3813
M. Djindjian, E. Lepresle, and J. B. Homs TABLE 1
TABLE 2
Reasons for exclusion of patients from final analysis"
Clinical summary of 356 cases in this study
Reason for Exclusion total cases lost to follow-upreview died < 90 days postop (no sepsis) protocol deviations total cases excluded
Oxacillin Group 198 4 7
Placebo Group 202 2 5
16 27
10 17
drainage or in whom the drain had been removed less than 48 hours before surgery, patients undergoing surgery in the casualty operating theater, and those treated by stereotactic surgery under a scanner. Oxacillin or a placebo identical in appearance was prescribed according to a previously generated randomized list and administered in 1-gm injectable doses. Administration began upon induction of anesthesia, at a dose of 200 mg/kg/24 hrs by direct intravenous injection every 4 hours; the total treatment lasted 24 hours.
Results Initially, 400 patients were included in this study, but 44 patients were omitted from evaluation for reasons summarized in Table 1. Table 2 presents a summary of the 356 patients for whom results could be analyzed. No side effects such as seizures were observed during the study. In the group treated with oxacillin, infection was noted in only one case (0.6%), compared with nine cases (4.9%) in the control group (Table 3). The difference was statistically significant (chi-squared test with Yates' correction: R = 4.22, p = 0.0398). All infections were deep except for one case of superficial infection following stereotactic surgery. Ten cases of meningitis without signs of infection were observed in each group and were systematically treated by a short course of antibiotics. In accordance with the criteria of Malisfl ~ these patients were not scored as infected (aseptic meningitis).
Discussion There have been m a n y publications on antibiotic prophylaxis in clean neurosurgery since 1979, when Malis ~~ reported the absence of infection in 1732 patients given a combination of intravenous vancomycin, intramuscular tobramycin, and topical streptomycin. A recent article by Haines 8 reviewed the subject in detail. While it is not necessary to repeat the same detailed discussion here, certain important points should be stressed: 1. The current goal of antibiotic prophylaxis in clean neurosurgery is to keep the infection level at less than the 1% achieved by Cushing 4 at the beginning of the century using nothing but soap and water. 2. The percentage infection in the absence of antibi384
Feature sex (M:F) duration of surgery 2-4 hrs 4-6 hrs > 6 hrs open sinuses type of surgery cerebral tumor (meningioma vascular posterior fossa spinal stereotactic other totals * No incidence of sepsis.
Oxacillin Group 90:81
Placebo Group 101:84
132 27 12 10*
148 25 12 6*
52 17 28 29 20 39 3 171
44 16) 22 41 25 46 7 185
otic prophylaxis as reported in the literature varies from 1% to 7% among series, but also, significantly, from year to year within any one series? 3 3. The risk of infection in clean neurosurgery varies according to the pathology involved.ll'~5 It also depends on a number of factors associated with infection such as shunt procedures, craniotomy as compared to spinal surgery, long operations, the necessity for drainage, 2~ the use of a microscope, CUSA, or other such device, certain pathological entities (glioma), and t u m o r recurrences.18 For this study, operations of predicted length of 2 hours or more were selected, thereby excluding low-risk interventions such as sciatica or peripheral nerve neurosurgery. 4. All studies have found the same distribution of infectious agents, with about 85 % Gram-positive bacteria of which 80% are Staphylococci. All the antibiotics tested therefore are anti-staphylococcal and have subsequently shown themselves to be 85% effective in single-drug or multidrug therapy. 12.2 5. The majority of nonrandomized studies demonstrate the efficacy of antibiotic prophylaxis. The following figures for the incidence of infection following surgery with antibiotic prophylaxis have been reported by authors using either Malis' antibiotic treatment or another single-drug therapy: Malis, ~~ 0%; Haines and Goodman, 9 0.9%; Savitz and Malis, ~60%; Quartey and Polyzoidis, ~3 0.8%; Urciuoli, et aL, ~9 0%; and Malis as cited by Dempsey, et al.,S 0.18 %. However, three studies indicate the ineffectiveness of antibiotic prophylaxis: Quadery, et al., 12 who used penicillin/sulfadiazine; Wright, 21 who used penicillin/streptomycin; and Raggueneau, et al., ~4 who used cephalothin (except for operations lasting more than 5 hours). Nonetheless, only the third study described true antibiotic prophylaxis. 6. Three randomized single-blind studies led to the conclusion that antibiotic prophylaxis is effective: GerJ. Neurosurg. / Volume 73/September, 1990
O x a c i l l i n p r o p h y l a x i s in p r o l o n g e d c l e a n n e u r o s u r g e r y TABLE 3 Details of sepsis in 10 patients with postoperative infection*
Surgeon (A, B, C, or D) A B B A B, C D C
Etiology
Age Durationof (yrs) & Intervention Sex
cerebral menin54, M gioma glioblastoma(stereo) 65, F glioblastoma (stereo) 68, F
4 hrs
cerebral meningioma cervicalneurinoma medulloblastoma cerebral AVM
74, F 40, F 37, M 15, F
Type of Infection
Isolated MicroOrganisms
Treatment Group
3 hrs
woundinfection, osteomyelitis of bone flap woundinfection woundinfection, empyema (died) meningitis
< 1 wk
S. epidermidis, MS
placebo
5 hrs 4 hrs 6 hrs
extraduralempyema meningitis ventriculitis
< 1 wk < 1 wk < 1 wk
S. aureus, MS Enterobacter cloacae Acinetobacter& S. aureus, MS Streptococcus S. epidermidis, MS
placebo placebo placebo placebo placebo
?
oxacillin
2 hrs 2 hrs
A A
spinal ependymoma 54, M 8 hrs meningitis(died) cerebral menin63, M 6 hrs woundinfection, osteomyelitis gioma of bone flap A-D cerebralAVM 54, F 8 hrs empyema * AVM = arteriovenous malformation.S. = Staphylococcus; MS = methicillin susceptible.
aghty and Feely, 7 who used Malis' treatment (0.5% vs. 3.5% in the control group); Young and Lawner, 22 who used cefazolin/gentamicin (0.73% vs. 4.09%); and Blomstedt and Kytt~i,2 who used vancomycin (2% vs. 8%). 7. Four randomized double-blind studies comparing antibiotics against placebo have been reported. Two of the four studies were stopped before completion due to the elevated number of infections in the control group (Savitz and Malis, ~6clindamycin vs. placebo, 1.2% and 10.9% infections, respectively; ~6 and Shapiro, et al., ~7 who used Malis' antibiotic prophylaxis treatment with 2.8% infection compared to 11% infection with placebo). In the two other completed studies (Bullock, et al., 3 who used pipericillin, and van Ek, et al., 2~ who used cloxacillin), antibiotic prophylaxis was shown to be effective: 2.1% infection vs. 5.9% and 3.3% vs. 10.3%, respectively. 8. All of the trials, in particular the four randomized double-blind studies, included operations for shunt placement. In our view, this affected the results. The placement of a shunt is known to involve a risk of infection higher than for neurosurgery in general. The use of antibiotic prophylaxis has proved effective in neurosurgery involving only shunt procedures. J,6 If the data from those subjects who had shunt procedures are excluded from the two completed randomized doubleblind studies, 3'2~the difference between the percent of infections in the treated group and the placebo group is not statistically significant. For this reason, patients undergoing shunt procedures have been excluded from the present study. 9. Our results, derived from patients undergoing long surgical interventions, show a drop in the level of infection from 4.9% in the placebo group to 0.6% in the group with antibiotic prophylaxis and give inconJ. Neurosurg. / Volume 7 3 / S e p t e m b e r , 1990
Time Between Surgery & Sepsis
3 wks S. epidermidis, MS
placebo
4 wks ? 4 wks S. epidermidis, MS
placebo placebo
& Streptococcus
< 1 wk < 1 wk 10 wks
testable support for antibiotic prophylaxis. The level of infection in our placebo group (nine cases, 4.9%) was consistent with that reported in the literature in the absence of antibiotic prophylaxis. Severe risk factors for infection were present among these nine infected patients: a very long operation (> 4 hours) in six cases; craniotomies in seven cases; high-risk pathology (meningiomas) in three patients; age over 50 years in six patients; and the use of a microscope in five and CUSA in six procedures. Six of the nine infections in the placebo group could be considered severe. There were two deaths due to infection and three repeat operations, including two flap ablations and one serious sequela (Table 3). These few cases show that the risk from infection is serious, both in terms of malfunction and threat to life, even for a neurosurgeon who maintains a low level of infection (< 2%, for example). Antibiotic prophylaxis is therefore justified. Despite the efficacy of antibiotic prophylaxis, several points are worth further consideration: 1) With few exceptions, infections were persistent in all reported studies. The same infectious agents are found, predominately S t a p h y l o c o c c i (76%), mainly S. a u r e u s , which appears not to be resistant to antibiotics. 3'7'9'22It should be noted, therefore, that rigorous aseptic practices are extremely important to combat S t a p h y l o c o c c i . 2) Other factors should be further considered for their role in risk of infection particularly the issues of shaving, the use of adhesive barriers at operation sites and drains, and the choice of antiseptics. 3) The occurrence of methicillin-resistant S t a p h y l o c o c c i has been largely overcome in recent years and does not appear to be a problem in clean neurosurgery, although it may become one. 4) The choice of antibiotic agent and the selection of a single-drug or multidrug therapy protocol must be 385
M. Djindjian, E. Lepresle, and J. B. Horns made on a patient-by-patient basis. The issue of antibiotic prophylaxis is inextricably linked to that of aseptic practices, which must be considered first. Our preferred choice of antibiotic treatment for interventions of more than 2 hours' duration is simple and inexpensive with the broad-spectrum antibiotic oxacillin, which we have used successfully during shunt procedures. 6 References 1. Blomstedt GC: Results of trimethoprim-sulfamethoxazole prophylaxis in ventriculostomy and shunting procedures. A double-blind randomized study. J Neurosurg 62:694-697, 1985 2. Blomstedt GC, Kytt/i J: Results of a randomized trial of vancomycin prophylaxis in craniotomy. J Neurosurg 69: 216-220, 1988 3. Bullock R, van Dellen JR, Ketelbey W, et al: A doubleblind placebo-controlled trial of perioperative prophylactic antibiotics for elective neurosurgery. 3 Neurosurg 69: 687-691, 1988 4. Cushing H: Concerning the results of operations for brain tumor. 3AMA 64:189-195, 1915 5. Dempsey R, Rapp RP, Young B, et al: Prophylactic parenteral antibiotics in clean neurosurgical procedures: a review. J Neurosurg 69:52-57, 1988 6. Djindjian M, Fevrier MJ, Otterbein G, et al: Oxacitlin prophylaxis in cerebrospinal fluid shunt procedures: results of a randomized open study in 60 hydrocephalic patients. Surg Neurol 25:178-180, 1986 7. Geraghty J, Feely M: Antibiotic prophylaxis in neurosurgery. A randomized controlled trial. J Neurosurg 60: 724-726, 1984 8. Haines SJ: Efficacy of antibiotic prophylaxis in clean neurosurgical operations. Neurosurgery 24:401-405, 1989 9. Haines SJ, Goodman ML: Antibiotic prophylaxis of postoperative neurosurgical wound infection. J Neurosurg 56: 103-105, 1982 10. Malis LI: Prevention of neurosurgical infection by extraoperative antibiotics. Neurosurgery 5:339-343, 1979 11. Mollman HD, Haines SJ: Risk factors for postoperative neurosurgical wound infection. A case-control study. J
386
Neurosurg 64:902-907, 1986 12. Quadery LA, Medlery AV, Miles J: Factors affecting the incidence of wound infection in neurosurgery. Acta Neurochir 39:133-141, 1973 13. Quartey GRC, Polyzoidis K: Intraoperative antibiotic prophylaxis in neurosurgery: a clinical study. Neurosurgery 8:669-67t, 1981 14. Raggueneau JL, Cophignon J, Kind A, et al: Analyse des suites infectieuses de 1000 interventions neurochirurgicales. Incidence de l'antibiothrrapie prophylactique. Nenrochirurgie 29:229-233, 1983 15. Savitz MH, Katz SS: Prevention of primary wound infection in neurosurgical patients: a 10-year study. Neurosurgery 18:685-688, 1986 16. Savitz MH, Malis LI: Prophylactic clindamycin for neurosurgical patients. NY State J Med 76:64-67, 1976 17, Shapiro M, Wald U, Simchen E, et al: Randomized clinical trial of intra-operative antimicrobial prophylaxis of infection after neurosurgical procedures. J Hosp Infect 8:283-295, 1986 18. Tenney JH, Vlahov D, Salcman M, et al: Wide variation in risk of wound infection following clean neurosurgery. Implications for perioperative antibiotic prophylaxis, J Neurosurg 62:243-247, 1985 19. Urciuoli R, Lo Russo G, Zeme S, et al: Traitement antibiotique prophylactique des complications infectieuses post-oprratoires en neurochirnrgie. Agressologie 24:237-238, 1983 20. van Ek B, Dijkmans B, van Dulken H, et al: Antibiotic prophylaxis in craniotomy. A prospective double-blind controlled study. Scand J Infect Dis 20:630-639, 1988 21. Wright RL: A survey of possible etiologic agents in postoperative craniotomy infections. J Neurosurg 25: 125-132, 1966 22. Young RF, Lawner PM: Perioperative antibiotic prophylaxis for prevention of postoperative neurosurgical infections. A randomized clinical trial. J Neurosurg 66: 701-705, 1987 Manuscript received October 12, 1989. Accepted in final form February 6, 1990. Address reprint requests to: Michel Djindjian, M.D., Service de Neurochirurgie, HSpital Henri Mondor, 94000 Crrteil, France.
.I. Neurosurg. / Volume 73/September, 1990
