Child’s Brain /; 243-250 (1975)
An Analysis of External Ventricular Drainage as a Treatment for Infected Shunts K oreaki M ori an d A n th o n y
J. R a im o n d i
Division of Neurological Surgery, Northwestern University and the Children’s Memorial Hospital, Chicago
Key Words. External ventricular drainage * Hydrocephalus • Infected shunts * Intracranial pressure
Infection is the major complication of shunts for primary and second ary hydrocephalus [4], as well as the most common cause for shunt fail ure. Control of intracranial pressure in hydrocephalic children becomes a serious problem when a shunt gets infected. Two commonly accepted means of treatment are: (1) removal of the infected shunt and immediate reinsertion of a new shunt, combined with antibiotic therapy [3], and (2) removal of the infected shunt and initiation of antibiotic therapy, with reinsertion of a new shunt after clearing of the infection. Each of these two approaches has serious limitations, especially when ventriculitis is present. In an effort to eradicate the infection and control intracranial pres sure, we instituted external ventricular drainage (EVD), with appropriate antibiotic treatment, as our treatment of choice. The present study analyz es EVD as a treatment for infected shunts.
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Abstract. An analysis of EVD as a treatment for infected shunts is reported. The study consists of 6 patients with ventriculo-jugular shunts and 17 patients with vcntriculo-peritoneal shunts. When infection of the shunting system is suspected on a clinical basis, the shunt should be removed immediately. EVD is adequate to con trol intraventricular pressure. After the infection clears, reimplantation of shunt should be considered.
244
M ori/R aimondi
Fig. 1. External ventricular drainage. 1 = Ventricular catheter; 2 = right angle connector; 3 = drill hole; 4 = skin incision; 5 = stab incision; 6 = Raimondi spring tubing; 7 = 18-gauge medicut; 8 = three-way stopcock; 9 = double male connec tor; 10 = Holter tubing; 11 = Holter valve; 12 = IV extension tubing; 13 = bur ette; 14 = manometer.
Materials and Methods
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
At the Children’s Memorial Hospital in Chicago, during the period January 1969 to November 1972, we implanted 1,153 shunts on 623 children with primary and secondary hydrocephalus. Prophylactic antibiotics were not routinely given. Specific antibiotics, based upon culture-sensitivity studies, were given intravenously and, in the presence of persistent ventriculitis, intraventricular instillation of appro priate antibiotics was performed in children with infected shunts. 128 of 1,153 shunts, or 11%> got infected. 77 of 623 patients or 12%> suffered infection of the shunting system. There were 128 infections in 77 children. Of 77 patients with an infected shunt system, 45 were complicated by ventriculitis. 26 of these 45 patients with ventriculitis were treated by EVD. The EVD system is as shown in figure 1. The patients were put in isolation. The intraventricular pressure was maintained at a set range and the EVD was replaced hourly, milliliter for milliliter, with saline solution. Daily cultures and sensitivities, cell counts and chemistries were performed. Serum electrolytes and protein were checked daily and replaced as needed. All procedures were carried out under strict aseptic technique. In analyzing EVD in these 26 children as a treatment for shunt infection with ventriculitis, we sought the answers to eight questions: (1) What was the primary mode of the initial cerebrospinal fluid deviation? (2) How long after the last shunt did the infection become obvious? (3) What were the identifiable signs and symp-
Analysis of External Ventricular Drainage
245
Table /. How long after the last shunt did the infection become obvious? Patients < 1 week 1-4 weeks 1-5 months 5-12 months > 1 year Unknown
10 21 10 2 1 3
Average 8 weeks.
Table //. What are the identifiable signs and symptoms of an infected shunt? Wound infection Fever Lethargy Vomiting Irritability Anorexia Swelling along shunt Abdominal pain Tense fontanelle Diarrhea
18 17 9 8 7 6 5 4 3 3
Exposure of tubing Seizure Meningismus Headache Increased sun set sign Limitation of upward gaze Decreased activity Insomnia Apneic spell
2 1 1 1 I 1 1 1 1
toms of the infected shunt? (4) What was the primary bacterial organism? (5) What antibiotics were used? (6) How long was EVD maintained? (7) What type of shunt was inserted after EVD? (8) What was the mortality and morbidity of EVD?
17 of 26 of the primary cerebrospinal fluid deviations were ventriculo peritoneal (V-P) shunts and 6 were ventriculo-jugular (V-J) shunts. Two children had ventriculo-gallbladder (V-G) shunts. In one child, we do not know what the primary shunt was. More than half of the shunts got infected within 1 month after the last shunt with the maximum incidence occurring in the first week. The rest of
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Observations
246
Mori/R aimondi
Table III. What were the primary baeterial organisms? Staphylococcus aureus Staphylococcus epidermidis Escherichia coli Klebsiella
15 14 6 3
Enterococci Pneumococci No growth Unknown
2 1 6 2
Table IV. What antibiotic treatment was used? Parenteral and oral 27 18 17 9 9 jj
Cloxacillin Tetracyclin Erythromycin Polymyxin B
8 7 7 6 4
Kanamycin Gentamycin Bacitracin Methicillin Polymyxin B
2 1 1 1
Intraventricular 15 5 5 2 1
the shunts were delayed infections but only three were known to occur af ter 5 months (table I). 20 of 26 patients had ventriculitis at the time of shunt removal. Wound infection, swelling along the shunt tract, irritability and ano rexia were frequently seen (table II). In this report, a definite diagnosis of ventriculitis was made only by cultures of ventricular fluid. Interestingly enough, ventricular CSF glucose was within the range of normal values in many cases. The primary organisms were mostly Staphylococcus aureus and Staphylococcus epidermidis (table III). Antibiotics were used systemically and intraventricularly in accordance with sensitivities as determined from the cultures (table IV). The EVD was maintained until the infections cleared. The average du ration was about 3 weeks, ranging from 5 days to 3 months (table V). If EVD is left in for more than 2 weeks, we revised it electively. After three consecutive 72-hour negative cultures of CSF, the antibiotics were stopped for 24 h. If the culture still remained negative during that time, the EVD was internalized.
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Ampicillin Kanamycin Methicillin Penicillin Cephalothin Gentamycin Oxacillin Cephaloridine Cephalexin Chloramphenicol Carbenicillin
Analysis of External Ventricular Drainage
247
Table V. How long was external ventricular drainage maintained? Patients < 1 week 1-4 weeks 4-8 weeks > 8 weeks
9 24 6 5
Average 3 weeks.
Table VI. What type of shunt was inserted after external ventricular drainage?
32 4 3 2 4
Ventriculoperitoneal Ventriculo-atrial Ventriculo-gallbladdcr Posterior fossa-peritoneal No shunt (no longer necessary — self cured hydrocephalus?)
Table VII. What was the morbidity and mortality of external ventricular drainage? Patients 4 1 1 1 1 1 1
After removal of EVD, the shunt was usually internalized into the per itoneal cavity, using the Raimondi spring catheter. However, if the patient was no longer able to absorb fluid from the peritoneum, either a vascular or gallbladder shunt was inserted (table VI). Interestingly enough, in four patients a shunt was no longer necessary. We have noted this before and after ventriculitis. After internalization of the EVD, the patients were put on prophylactic antibiotics for varying periods of time.
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Convulsions Severe dehydration and electrolyte imbalance Hypoproteinemia Subdural hematoma Cardiorespiratory arrest - expired Status epilcpticus and apneic spells - expired Pyocephalus - expired
248
M ori/R aimondi
Four children expired during the treatment (table VII). None suffered increased intracranial pressure. Mostly, the morbidity consisted of such problems as low serum protein, electrolyte imbalance, and superinfection. These were readily controlled.
Infection is the most serious problem of any type of shunt for hydroce phalus. Our results agree with the reported average incidence of shunt infection: about 12% [5J. Though infection within 1 week is readily un derstandable, one has no explanation for these infections becoming clini cally obvious after the 5th postoperative week. Since bacteria do get into the venous system, one may understand delayed infection of V-J shunts. The infections in our patients ranged from wound abscess to meningitis and ventriculitis. The primary organisms were S. epidermidis (albus) and S. aureus. The vast majority of infections reported in the literature have been due to these two organsims. Staphylococci are known to be skin org anisms. On the scalp dtese two organisms compose more than 90% of the normal resident skin flora [2]. The skin of the operative site is a likely lo cus for entry of organisms into the wound or shunt tubing. The role of 5. aureus as a cause of surgical wound infections has been well studied [1]. However, the neurosurgical shunt differs from most other surgical proce dures because a foreign body is implanted under the skin. In this proce dure, S. epidermidis, an organism usually considered nonpathogenic, is in deed responsible for slightly less than one third of all our primary infection. A prospective study of the bacteriology of V-P shunting procedures showed that skin contamination was related to the duration of the proce dure and the scalp was found to be contaminated more often than the ab dominal skin [1], S. epidermidis was the most commonly isolated poten tial pathogen recovered from the skin during surgery. During the period of this retrospective study, prophylactic antibiotics were given to randomly selected patients. The surgical team was not aware of which patients were to receive prophylaxis. The antibiotic select ed was cephalothin 80 mg/kg/day in divided doses every 6 h, i.v., postoperatively. No definitive conclusions concerning the efficacy of the anti biotic prophylaxis were observed. In our series, during the treatment by EVD, all procedures were per formed under strict aseptic technique. Despite these precautions, in some cases, daily cultures revealed secondary invaders. These secondary invad
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Discussion
249
ers, mostly gram-negative rods such as Klebsiella, Mima, Proteus, Pseu domonas, etc. and fungi, indicated superinfection or hospital-acquired transfer infection. The EVD is not a completely closed system. Therefore, the longer the EVD is maintained, the greater the likelihood of contami nation and subsequent superinfection. At the present time there is no satisfactory way to deal with infected shunts. Most neurosurgeons acknowledge the necessity of removing the infected shunt system regardless of the type of shunt. Of our 77 patients with infected shunts, 45 (60%) were complicated by ventriculitis. 26 of these 45 patients were treated by EVD. When the drainage is regulated by a one way valve, we cannot assume that the possibility of retrograde flow (bacterial migration) is either nonexistent or remonte. The number of our secondary infections attests to this. This is contrary to the reports that ex ternalized valve drainage systems [6] diminish reflux infections. EVD, therefore, has many very real drawbacks, such as retrograde secondary infection, continuous loss of protein and electrolytes, the need to keep the children horizontal or to have highly qualified personnel with them, needs for extensive isolation facilities, the risk of contaminating personnel and other patients. 19 of 45 patients with infected shunts complicated by ven triculitis were treated by removal of the infected shunts and either imme diate reinsertion of a new shunt or repeat ventricular taps. A comparison of the results of treatment for shunt infections associat ed with ventriculitis in the two groups, the one treated by EVD and the other without EVD, was considered to be meaningful, especially from the prognostic point of view. Regarding recurrence of infections, the average number of episodes of infection was 1.9 in the EVD group and 1.4 in the non-EVD group. The slightly higher incidence of recurrence may reflect that the EVD group had clinically high-grade infections, or that they were more prone to secondary infections. Regarding morbidtiy and mortality, in the EVD group four patients suffered convulsions, severe dehydration and electrolyte imbalance, hypoproteinemia, and subdural hematoma, re spectively. Four of 26 patients (15%) expired (table VIT). In the nonEVD group, opisthotonos, septicemia and spastic paralysis developed in one patient. Four of 19 patients (21%) expired. The significant advan tages of EVD are that it allows for both a precise means of controlling in traventricular pressure and the removal of intraventricular pyrogenic exudate from the ventricular system. We found it to be of particular interest that four of our EVD patients with hydrocephalus no longer needed a shunt (table VI).
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Analysis of External Ventricular Drainage
250
M ori Raimondi
References
Koreaki M ori. Department of Neurosurgery Kyoto University Medical School. 54 Kawahara-cho, Shogoin, Sakyo-ku, 606 Kyoto (Japan)
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
1 Burker, J. F.: Identification of the sources of staphylococci contaminating the surgical wound during operation. Ann. Surg. 158: 898-904 (1963). 2 K ligman , A. M.: The bacteriology of normal skin; in M aibach, H oward and H ildick -Sm ith Skin bacteria and their role in infection, p. 26 (Cavin-BIakiston McGraw Hill. New York 1965). 3 P errin , J. C. and M c L aurin, R. L.: Infected ventriculo-atrial shunts. A method of treatment. J. Neurosurg. 27: 21-26 (1967). 4 R obinson , J. S. and R aimondi, A. J.: Complications of V-P shunting procedures for congenital and secondary hydrocephalus in childhood. Proc. Amer. Ass. Neurol. Surg., St. Louis 1974. 5 Scarff , J. E.: Treatment of hydrocephalus: an historical and critical view of methods and results. J. Neurol. Neurosurg. Psychiat. 26: 1-26 (1963). 6 W hite , R. J.; D arters, G. J.; Y ashon, D., and Albin , M. S.: Temporary control of cerebrospinal fluid volume and pressure by means of an externalized valvedrainage system. J. Neurosurg. 30: 264—269 (1969).
An Analysis of External Ventricular Drainage as a Treatment for Infected Shunts K oreaki M ori an d A n th o n y
J. R a im o n d i
Division of Neurological Surgery, Northwestern University and the Children’s Memorial Hospital, Chicago
Key Words. External ventricular drainage * Hydrocephalus • Infected shunts * Intracranial pressure
Infection is the major complication of shunts for primary and second ary hydrocephalus [4], as well as the most common cause for shunt fail ure. Control of intracranial pressure in hydrocephalic children becomes a serious problem when a shunt gets infected. Two commonly accepted means of treatment are: (1) removal of the infected shunt and immediate reinsertion of a new shunt, combined with antibiotic therapy [3], and (2) removal of the infected shunt and initiation of antibiotic therapy, with reinsertion of a new shunt after clearing of the infection. Each of these two approaches has serious limitations, especially when ventriculitis is present. In an effort to eradicate the infection and control intracranial pres sure, we instituted external ventricular drainage (EVD), with appropriate antibiotic treatment, as our treatment of choice. The present study analyz es EVD as a treatment for infected shunts.
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Abstract. An analysis of EVD as a treatment for infected shunts is reported. The study consists of 6 patients with ventriculo-jugular shunts and 17 patients with vcntriculo-peritoneal shunts. When infection of the shunting system is suspected on a clinical basis, the shunt should be removed immediately. EVD is adequate to con trol intraventricular pressure. After the infection clears, reimplantation of shunt should be considered.
244
M ori/R aimondi
Fig. 1. External ventricular drainage. 1 = Ventricular catheter; 2 = right angle connector; 3 = drill hole; 4 = skin incision; 5 = stab incision; 6 = Raimondi spring tubing; 7 = 18-gauge medicut; 8 = three-way stopcock; 9 = double male connec tor; 10 = Holter tubing; 11 = Holter valve; 12 = IV extension tubing; 13 = bur ette; 14 = manometer.
Materials and Methods
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
At the Children’s Memorial Hospital in Chicago, during the period January 1969 to November 1972, we implanted 1,153 shunts on 623 children with primary and secondary hydrocephalus. Prophylactic antibiotics were not routinely given. Specific antibiotics, based upon culture-sensitivity studies, were given intravenously and, in the presence of persistent ventriculitis, intraventricular instillation of appro priate antibiotics was performed in children with infected shunts. 128 of 1,153 shunts, or 11%> got infected. 77 of 623 patients or 12%> suffered infection of the shunting system. There were 128 infections in 77 children. Of 77 patients with an infected shunt system, 45 were complicated by ventriculitis. 26 of these 45 patients with ventriculitis were treated by EVD. The EVD system is as shown in figure 1. The patients were put in isolation. The intraventricular pressure was maintained at a set range and the EVD was replaced hourly, milliliter for milliliter, with saline solution. Daily cultures and sensitivities, cell counts and chemistries were performed. Serum electrolytes and protein were checked daily and replaced as needed. All procedures were carried out under strict aseptic technique. In analyzing EVD in these 26 children as a treatment for shunt infection with ventriculitis, we sought the answers to eight questions: (1) What was the primary mode of the initial cerebrospinal fluid deviation? (2) How long after the last shunt did the infection become obvious? (3) What were the identifiable signs and symp-
Analysis of External Ventricular Drainage
245
Table /. How long after the last shunt did the infection become obvious? Patients < 1 week 1-4 weeks 1-5 months 5-12 months > 1 year Unknown
10 21 10 2 1 3
Average 8 weeks.
Table //. What are the identifiable signs and symptoms of an infected shunt? Wound infection Fever Lethargy Vomiting Irritability Anorexia Swelling along shunt Abdominal pain Tense fontanelle Diarrhea
18 17 9 8 7 6 5 4 3 3
Exposure of tubing Seizure Meningismus Headache Increased sun set sign Limitation of upward gaze Decreased activity Insomnia Apneic spell
2 1 1 1 I 1 1 1 1
toms of the infected shunt? (4) What was the primary bacterial organism? (5) What antibiotics were used? (6) How long was EVD maintained? (7) What type of shunt was inserted after EVD? (8) What was the mortality and morbidity of EVD?
17 of 26 of the primary cerebrospinal fluid deviations were ventriculo peritoneal (V-P) shunts and 6 were ventriculo-jugular (V-J) shunts. Two children had ventriculo-gallbladder (V-G) shunts. In one child, we do not know what the primary shunt was. More than half of the shunts got infected within 1 month after the last shunt with the maximum incidence occurring in the first week. The rest of
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Observations
246
Mori/R aimondi
Table III. What were the primary baeterial organisms? Staphylococcus aureus Staphylococcus epidermidis Escherichia coli Klebsiella
15 14 6 3
Enterococci Pneumococci No growth Unknown
2 1 6 2
Table IV. What antibiotic treatment was used? Parenteral and oral 27 18 17 9 9 jj
Cloxacillin Tetracyclin Erythromycin Polymyxin B
8 7 7 6 4
Kanamycin Gentamycin Bacitracin Methicillin Polymyxin B
2 1 1 1
Intraventricular 15 5 5 2 1
the shunts were delayed infections but only three were known to occur af ter 5 months (table I). 20 of 26 patients had ventriculitis at the time of shunt removal. Wound infection, swelling along the shunt tract, irritability and ano rexia were frequently seen (table II). In this report, a definite diagnosis of ventriculitis was made only by cultures of ventricular fluid. Interestingly enough, ventricular CSF glucose was within the range of normal values in many cases. The primary organisms were mostly Staphylococcus aureus and Staphylococcus epidermidis (table III). Antibiotics were used systemically and intraventricularly in accordance with sensitivities as determined from the cultures (table IV). The EVD was maintained until the infections cleared. The average du ration was about 3 weeks, ranging from 5 days to 3 months (table V). If EVD is left in for more than 2 weeks, we revised it electively. After three consecutive 72-hour negative cultures of CSF, the antibiotics were stopped for 24 h. If the culture still remained negative during that time, the EVD was internalized.
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Ampicillin Kanamycin Methicillin Penicillin Cephalothin Gentamycin Oxacillin Cephaloridine Cephalexin Chloramphenicol Carbenicillin
Analysis of External Ventricular Drainage
247
Table V. How long was external ventricular drainage maintained? Patients < 1 week 1-4 weeks 4-8 weeks > 8 weeks
9 24 6 5
Average 3 weeks.
Table VI. What type of shunt was inserted after external ventricular drainage?
32 4 3 2 4
Ventriculoperitoneal Ventriculo-atrial Ventriculo-gallbladdcr Posterior fossa-peritoneal No shunt (no longer necessary — self cured hydrocephalus?)
Table VII. What was the morbidity and mortality of external ventricular drainage? Patients 4 1 1 1 1 1 1
After removal of EVD, the shunt was usually internalized into the per itoneal cavity, using the Raimondi spring catheter. However, if the patient was no longer able to absorb fluid from the peritoneum, either a vascular or gallbladder shunt was inserted (table VI). Interestingly enough, in four patients a shunt was no longer necessary. We have noted this before and after ventriculitis. After internalization of the EVD, the patients were put on prophylactic antibiotics for varying periods of time.
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Convulsions Severe dehydration and electrolyte imbalance Hypoproteinemia Subdural hematoma Cardiorespiratory arrest - expired Status epilcpticus and apneic spells - expired Pyocephalus - expired
248
M ori/R aimondi
Four children expired during the treatment (table VII). None suffered increased intracranial pressure. Mostly, the morbidity consisted of such problems as low serum protein, electrolyte imbalance, and superinfection. These were readily controlled.
Infection is the most serious problem of any type of shunt for hydroce phalus. Our results agree with the reported average incidence of shunt infection: about 12% [5J. Though infection within 1 week is readily un derstandable, one has no explanation for these infections becoming clini cally obvious after the 5th postoperative week. Since bacteria do get into the venous system, one may understand delayed infection of V-J shunts. The infections in our patients ranged from wound abscess to meningitis and ventriculitis. The primary organisms were S. epidermidis (albus) and S. aureus. The vast majority of infections reported in the literature have been due to these two organsims. Staphylococci are known to be skin org anisms. On the scalp dtese two organisms compose more than 90% of the normal resident skin flora [2]. The skin of the operative site is a likely lo cus for entry of organisms into the wound or shunt tubing. The role of 5. aureus as a cause of surgical wound infections has been well studied [1]. However, the neurosurgical shunt differs from most other surgical proce dures because a foreign body is implanted under the skin. In this proce dure, S. epidermidis, an organism usually considered nonpathogenic, is in deed responsible for slightly less than one third of all our primary infection. A prospective study of the bacteriology of V-P shunting procedures showed that skin contamination was related to the duration of the proce dure and the scalp was found to be contaminated more often than the ab dominal skin [1], S. epidermidis was the most commonly isolated poten tial pathogen recovered from the skin during surgery. During the period of this retrospective study, prophylactic antibiotics were given to randomly selected patients. The surgical team was not aware of which patients were to receive prophylaxis. The antibiotic select ed was cephalothin 80 mg/kg/day in divided doses every 6 h, i.v., postoperatively. No definitive conclusions concerning the efficacy of the anti biotic prophylaxis were observed. In our series, during the treatment by EVD, all procedures were per formed under strict aseptic technique. Despite these precautions, in some cases, daily cultures revealed secondary invaders. These secondary invad
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Discussion
249
ers, mostly gram-negative rods such as Klebsiella, Mima, Proteus, Pseu domonas, etc. and fungi, indicated superinfection or hospital-acquired transfer infection. The EVD is not a completely closed system. Therefore, the longer the EVD is maintained, the greater the likelihood of contami nation and subsequent superinfection. At the present time there is no satisfactory way to deal with infected shunts. Most neurosurgeons acknowledge the necessity of removing the infected shunt system regardless of the type of shunt. Of our 77 patients with infected shunts, 45 (60%) were complicated by ventriculitis. 26 of these 45 patients were treated by EVD. When the drainage is regulated by a one way valve, we cannot assume that the possibility of retrograde flow (bacterial migration) is either nonexistent or remonte. The number of our secondary infections attests to this. This is contrary to the reports that ex ternalized valve drainage systems [6] diminish reflux infections. EVD, therefore, has many very real drawbacks, such as retrograde secondary infection, continuous loss of protein and electrolytes, the need to keep the children horizontal or to have highly qualified personnel with them, needs for extensive isolation facilities, the risk of contaminating personnel and other patients. 19 of 45 patients with infected shunts complicated by ven triculitis were treated by removal of the infected shunts and either imme diate reinsertion of a new shunt or repeat ventricular taps. A comparison of the results of treatment for shunt infections associat ed with ventriculitis in the two groups, the one treated by EVD and the other without EVD, was considered to be meaningful, especially from the prognostic point of view. Regarding recurrence of infections, the average number of episodes of infection was 1.9 in the EVD group and 1.4 in the non-EVD group. The slightly higher incidence of recurrence may reflect that the EVD group had clinically high-grade infections, or that they were more prone to secondary infections. Regarding morbidtiy and mortality, in the EVD group four patients suffered convulsions, severe dehydration and electrolyte imbalance, hypoproteinemia, and subdural hematoma, re spectively. Four of 26 patients (15%) expired (table VIT). In the nonEVD group, opisthotonos, septicemia and spastic paralysis developed in one patient. Four of 19 patients (21%) expired. The significant advan tages of EVD are that it allows for both a precise means of controlling in traventricular pressure and the removal of intraventricular pyrogenic exudate from the ventricular system. We found it to be of particular interest that four of our EVD patients with hydrocephalus no longer needed a shunt (table VI).
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
Analysis of External Ventricular Drainage
250
M ori Raimondi
References
Koreaki M ori. Department of Neurosurgery Kyoto University Medical School. 54 Kawahara-cho, Shogoin, Sakyo-ku, 606 Kyoto (Japan)
Downloaded by: Washington University 128.252.67.66 - 5/14/2018 3:33:10 AM
1 Burker, J. F.: Identification of the sources of staphylococci contaminating the surgical wound during operation. Ann. Surg. 158: 898-904 (1963). 2 K ligman , A. M.: The bacteriology of normal skin; in M aibach, H oward and H ildick -Sm ith Skin bacteria and their role in infection, p. 26 (Cavin-BIakiston McGraw Hill. New York 1965). 3 P errin , J. C. and M c L aurin, R. L.: Infected ventriculo-atrial shunts. A method of treatment. J. Neurosurg. 27: 21-26 (1967). 4 R obinson , J. S. and R aimondi, A. J.: Complications of V-P shunting procedures for congenital and secondary hydrocephalus in childhood. Proc. Amer. Ass. Neurol. Surg., St. Louis 1974. 5 Scarff , J. E.: Treatment of hydrocephalus: an historical and critical view of methods and results. J. Neurol. Neurosurg. Psychiat. 26: 1-26 (1963). 6 W hite , R. J.; D arters, G. J.; Y ashon, D., and Albin , M. S.: Temporary control of cerebrospinal fluid volume and pressure by means of an externalized valvedrainage system. J. Neurosurg. 30: 264—269 (1969).
