Follow-up analysis of ventriculoperitoneal and ventriculoatrial shunts for hydrocephalus RONALD J. IGNELZI, M.D., AND WOLFF M. KIRSCH, M.D. Department of Neurosurgery, University of Colorado Medical Center, Denver, Colorado
~" The authors report an analysis of 300 patients in whom primary treatment for hydrocephalus was either a ventriculoperitoneal (VP) or ventriculoatrial (VA) shunt. Although differences in primary and total revision rates between placements were not statistically significant, the VA shunt failures carried a higher morbidity than the VP placements. KEy WOADS 9 hydrocephalus ventriculoatrial shunt
9 ventriculoperitoneal shunt
9
REATMENT of hydrocephalus remains tion over 15 years in an attempt to limit or a major neurosurgical enigma. The standardize the variables. A total of 664 majority of current shunting procedures operative procedures were required to control involve either a ventriculoperitoneal (VP) or a the hydrocephalus in these patients. Table 1 ventriculoatrial (VA) placement. This report categorizes the patients by hydrocephalus analyzes the experience at the University of etiology. One-third of the patients had conColorado Medical Center over the last 15 genital hydrocephalus on the basis of aqueyears with regard to the effectiveness and ductal stenosis with or without myelomencomplications of these two placements. Our ingocele; the others had hydrocephalus analysis indicates that the peritoneal associated with subarachnoid hemorrhage, placements have less serious associated mor- trauma, porencephaly, meningitis, and the sobidity than the atrial placements. called syndrome of "normal-pressure hydrocephalus." Virtually all patients seen were treated with either VA or VP shunts (Table Materials and Methods 2); the placements were related to prevailing Inherent in all reviews of the treatment of practices rather than to patient preselection. hydrocephalus is a multitude of variables that The age of the patient when first receiving the make comparison difficult, such as the shunt ranged from 1 day to 90 years. At this etiology of the hydrocephalus, the type of time the current status is known in 82% of the valve system used, different surgeons and patients, 71% of whom have been followed for techniques, and methods of reporting follow- longer than 1 year. The initial shunt valve up. We conducted our own review of 300 systems were either the Holter, Pudenz, or cases of hydrocephalus treated at one institu- Denver design.
T
J. Neurosurg. / Volume 42 / June, 1975
679
R. J. Ignelzi and W. M. Kirsch TABLE 1
Etiology of hydrocephalus in 300 patients with shunts
Etiology obstructive hydrocephalus: aqueductal stenosis aqueductal stenosis and myelomeningocele tumor communicating hydrocephalus: subarachnoid hemorrhage, trauma porencephaly, normal-pressure hydrocephalus, meningitis total
No. of Cases
~o
54 43
18.0 14.3
55
18.4
148
49.3
300
100.0
TABLE 2 Type of initial shunt placements
Type of Shunt
No. Placed
VP VA other* total
114 177 9 300
7o 38 59 3 100
* Includes ventriculo-pleural,spinal-peritoneal, subdural peritoneal, subdural atrial, and ventriculocisternostomy.
Results
A comparison between thc VP and VA shunt revisions was made to ascertain the advantages and complications of each. It was found that 58% of the VP shunts and 48% of the VA shunts required at least one revision. This is not statistically different (p > .05). Table 3 summarizes the reasons for primary revisions of the VP and VA shunts regardless of the valve system used. Most revisions were distal, and related to the hazards of a foreign body in either the vascular system or peritoneal cavity that led to clotting in the shunt tubing or caused vascular occlusion of a major venous channel with atrial shunts and adhesions forming around the distal shunt tubing in the peritoneal placements. The VP distal revisions were often very minor and at times corrected under local anesthesia while the VA distal revisions usually required 680
general anesthesia and a search for a new portal because the vascular channel being used was thrombosed. Attention was then turned to an analysis of all the nonelective revisions to see if any consistent trends could be found. In the long run, the VP shunts required many more total revisions (58%) than did VA shunts (48%). However, the difference is not significant (p > .10). We next made an analysis of the time interval between the initial placement and first nonelective revision. Using the mean for comparison, we found that the interval between initial placement and primary revision was 4.8 months for the VP placement and 14.1 months for the VA placement. Another method of analyzing this interval is to determine the percentage of shunt placements that required a revision within a given interval; 18% of the VP shunts and 12% of the VA shunts required revision within the first month. However, for those patients who required no revision at the end of 2 years follow-up, only 3% of the VP shunts went on to require revision as opposed to 10% of the VA shunts. This indicates that the complication rate is highest in the first month after shunting and that although the VP shunts required a slightly higher rate of revision in the first month, in the long run the VP shunts do better in terms of requiring less primary revision. The problem of infection was studied in detail. Eighteen infections developed from primary placements; Table 4 summarizes them according to initial placement and diagnosis. In cases that required a primary revision, infection was a cause in VA shunts 1.6 times as often as in VP shunts. The mean interval from initial placement to diagnosis of infection was 2.0 months; although no mortality was associated with the infections, the morbidity in terms of requiring further revision and ultimate poor neurological function was great. The causative organism was coagulase-negative Staphylococcus epidermidis in 16 of the 18 infections. Over the last 2 years prophylactic antibiotics were used in all of the cases, and a drop (not statistically significant) in the infection rate occurred. In most cases, prophylaxis consisted of methicillin 25 mg/kg every 6 hours, 24 hours preand postoperatively, and dicloxacillin, 6 rag/ kg every 6 hours beginning on the second postoperative day and for 4 more days. J. Neurosurg. / Volume 42 / June, 1975
VP and VA shunts for hydrocephalus Discussion
TABLE 3
Hydrocephalus has been described in medical writings since the time of Hippocrates. In 1914, Dandy and Blackfan 5 first established the pathology of hydrocephalus and developed sound anatomical, physiological, and surgical principles for its treatment. Nulsen and Spitz 11 proposed ventriculoatrial shunting and the Holter valve-regulated shunt system in 1952, and in 1955, Scott, et a l . ) 8 described the use of a ventriculoperitoneal shunt. Since that time much controversy has existed over which is the better placement, but few large series have compared their clinical results? ,14 In an excellent review of the treatment of hydrocephalus, Scarf~ 6 tabulated the published results of the different methods up to 1963; he found a strong tendency for the VP shunt to become obstructed at the distal end and a high incidence (30% to 50%) of permanent failures, and an equally high rate of complications with the VA shunting method. More recent reviews have indicated that the VP shunt may be superior overall to the VA shunt. Robertson, et al., TM reported a 44% revision rate, and Little, et al.,8 a 47% revision rate for the VP shunt. Revisions were due mainly to obstruction of the distal end by adhesions, in agreement with our data. Ascites, TM perforation of a viscous, ~9 intraabdominal cyst formation, 1 volvulus, 1~ and perforation of the scrotum or vagina, ~s are reported but exceedingly rare complications of VP shunts. Revision rates for the VA placement in recent large series range from 46% to 78%. 2,~2 Complications of the VA placement included pleural effusion, pulmonary emboli, inferior vena cava occlusions, superior vena cava occlusions, and septicemia. 7 Infection rate has been reported as high as 30% with this placement. ~ Our results indicate that although the VP and VA placements carried primary and total revision rates that were not significantly different statistically, the complications were less life-threatening, carried a lower morbidity, and were easier to correct in VP placements as compared to VA placements. It was also apparent that there was less need for elective revision due to body growth with the VP than with the VA placement. Our finding that the VA shunt carried a higher rate of infection than the VP shunt
Reason ]br revision of primary shunt
J. Neurosurg. / Volume 42 / J u n e , 1975
No. of Cases
Cause of Revision
VP: 67 Revisions, 58 % of initial placements
proximal: particulate matter wrong position other distal: omental adhesions peritoneal adhesions fibrin plug in catheter other infection
7 2
l0 (15%)
1
25 12
49 (71%)
9 4
7 (10.6%)
VA: 85 Revisions, 48 % of initial placements
proximal: particulate matter in ventricular catheter wrong position other distal : clot jugular thrombosis vena cava thrombosis other infection
27 (31%)
20 6 1 23 13 2 9
47 (55 %)
11 (13 %)
TABLE 4 Infection of initial shunt according to type of placemem and diagnosis
No. No. of Placed Infections type of placement:
vP VA
diagnos&:
114 177
aqueductal stenosis aqueductal stenosis and myelomeningocele tumor other
7 11
Vo
%
Revisions Due to Infection
6.1 6.2
10.6 13
4 4 4 6
agrees with other reports, s,* Staphylococcus epidermidis was the overwhelming cause of infection in other reported seriesJ ,9,17 The drop in infection rate over the last 2 years in our series since the introduction of prophylactic antibiotics, while not statistically significant, is in agreement with Yu and Patterson's work. 2~ Other factors that seem to be impor681
R. J. Ignelzi and W. M. Kitsch tant in the development of infection are the underlying disease process, the number and the type of preoperative procedures, the length of operation, and the number of postoperative diagnostic reservoir taps.
Summary Over the last 15 years, 300 patients with hydrocephalus were treated at the University of Colorado Medical Center and required a total of 664 operative procedures. When the effectiveness of the VA versus the VP placement was compared, no statistically significant difference as regards primary or total revision rates was found; however, VA shunts developed more serious complications, including infections, than did the VP shunts. VP shunts should be considered primarily because of the ease in placement, lower rate of serious complications, and ease of revision.
8.
9. 10. 11.
12.
13. 14.
Acknowledgment We would like to express appreciation to Franklin W. Briese, Ph.D., Department of Biometrics at the University of Colorado Medical Center, for his help in preparation of this manuscript.
References 1. Ames RH: Ventriculo-peritoneal shunts in the management of hydrocephalus. J Neurosurg
27:525-529, 1967 2. Becker DP, Nulsen FE: Control of hydrocephalus by valve-regulated venous shunt: avoidance of complications in prolonged shunt maintenance. J Neurosurg 28:215-226, 1968 3. Bruce AM, Lorber J, Shedden WIH, et al: Persistent bacteraemia following ventriculocaval shunt operations for hydrocephalus in infants. Dev Med Child Neurol 5:461-470, 1963 4. Callaghan RP, Cohen S J, Stewart GT: Septicaemia due to colonization of Spitz-Holter valves by staphylococci. Five cases treated with methicillin. Br Med J 1:860-863, 1961 5. Dandy WE, Blackfan KD: Internal hydrocephalus: an experimental, clinical and pathological study. Am J Dis Child 8:406-482, 1914 6. Fokes EC Jr: Occult infections of ventriculoatrial shunts. J Nenrosurg 33:517-523, 1970 7. Forrest DM, Cooper DGW: Complications of 682
15.
16.
17.
18.
19.
20.
ventriculo-atrial shunts. A review of 455 cases. J Neurosurg 29:506-512, 1968 Little JR, Rhoton AL Jr, Mellinger JF: Comparison of ventriculoperitoneal and ventriculoatrial shunts for hydrocephalus in children. Mayo Clin Proc 47:396-401, 1972 Luthard T: Bacterial infections in ventriculoatrial shunt systems. Dev Med Child Neurol (Suppl) 22:105-109, 1970 Murtagh F, Lehman R: Peritoneal shunts in the management of hydrocephalus. JAMA 202:1010-1014, 1967 Nulsen FE, Spitz EF: Treatment of hydrocephalus by direct shunt from ventricular to jugular vein. Surg Forum 2:399-403, 1952 Overton MC, Snodgrass SR: Ventriculovenous shunts for infantile hydrocephalus. A review of five year's experience with this method. J Neurosnrg 23:517-521, 1965 Ramani PS: Extrusion of abdominal catheter of ventriculoperitoneal shunt into the scrotum. Case report. J Neurosurg 40:772-773, 1974 Robertson JS, Maraqa MI, Jennett B: Ventriculoperitoneal shunting for hydrocephalus. Br Med J 1:289-292, 1973 Sakoda TH, Maxwell JA, Brackett CE Jr: Intestinal volvulus secondary to a ventriculoperitoneal shunt. Case report, d Neurosnrg 35:95-96, 1971 Scarff J: Treatment of hydrocephalus: an historical and critical review of methods and results. J Nenrol Neurosurg Psychiatry 26:1-26, ' 1963 Schimke RT, Black PH, Mark VH, et al: Indolent staphylococcus albus or aureus bacteremia after ventriculoatriostomy. Role of foreign body in its initiation and perpetuation. N Engl J Med 264:264-270, 1961 Scott M, Wycis HT, Murtagh F, et al: Observations on ventricular and lumbar subarachnoid shunts in hydrocephalus in infants. J Neurosurg 12:165-175, 1955 Wilson CB, Bertan V: Perforation of the bowel complicating peritoneal shunt for hydrocephalus. Report of two cases. Am Surg 32:601-603, 1966 Yu HC, Patterson RH: Prophylactic antimicrobial agents after ventriculo atriotomy for hydrocephalus. J Pediatr Surg 8:881-885, 1973
This research was supported by a National Institutes of Health Training Grant, 5T01, NS05564. Address reprint requests to: Ronald J. Ignelzi, M.D., Division of Neurosurgery, Veterans Administration Hospital, 3350 La Jolla Village Drive, San Diego, California 92161. J. Neurosurg. / Volume 42 / June, 1975
~" The authors report an analysis of 300 patients in whom primary treatment for hydrocephalus was either a ventriculoperitoneal (VP) or ventriculoatrial (VA) shunt. Although differences in primary and total revision rates between placements were not statistically significant, the VA shunt failures carried a higher morbidity than the VP placements. KEy WOADS 9 hydrocephalus ventriculoatrial shunt
9 ventriculoperitoneal shunt
9
REATMENT of hydrocephalus remains tion over 15 years in an attempt to limit or a major neurosurgical enigma. The standardize the variables. A total of 664 majority of current shunting procedures operative procedures were required to control involve either a ventriculoperitoneal (VP) or a the hydrocephalus in these patients. Table 1 ventriculoatrial (VA) placement. This report categorizes the patients by hydrocephalus analyzes the experience at the University of etiology. One-third of the patients had conColorado Medical Center over the last 15 genital hydrocephalus on the basis of aqueyears with regard to the effectiveness and ductal stenosis with or without myelomencomplications of these two placements. Our ingocele; the others had hydrocephalus analysis indicates that the peritoneal associated with subarachnoid hemorrhage, placements have less serious associated mor- trauma, porencephaly, meningitis, and the sobidity than the atrial placements. called syndrome of "normal-pressure hydrocephalus." Virtually all patients seen were treated with either VA or VP shunts (Table Materials and Methods 2); the placements were related to prevailing Inherent in all reviews of the treatment of practices rather than to patient preselection. hydrocephalus is a multitude of variables that The age of the patient when first receiving the make comparison difficult, such as the shunt ranged from 1 day to 90 years. At this etiology of the hydrocephalus, the type of time the current status is known in 82% of the valve system used, different surgeons and patients, 71% of whom have been followed for techniques, and methods of reporting follow- longer than 1 year. The initial shunt valve up. We conducted our own review of 300 systems were either the Holter, Pudenz, or cases of hydrocephalus treated at one institu- Denver design.
T
J. Neurosurg. / Volume 42 / June, 1975
679
R. J. Ignelzi and W. M. Kirsch TABLE 1
Etiology of hydrocephalus in 300 patients with shunts
Etiology obstructive hydrocephalus: aqueductal stenosis aqueductal stenosis and myelomeningocele tumor communicating hydrocephalus: subarachnoid hemorrhage, trauma porencephaly, normal-pressure hydrocephalus, meningitis total
No. of Cases
~o
54 43
18.0 14.3
55
18.4
148
49.3
300
100.0
TABLE 2 Type of initial shunt placements
Type of Shunt
No. Placed
VP VA other* total
114 177 9 300
7o 38 59 3 100
* Includes ventriculo-pleural,spinal-peritoneal, subdural peritoneal, subdural atrial, and ventriculocisternostomy.
Results
A comparison between thc VP and VA shunt revisions was made to ascertain the advantages and complications of each. It was found that 58% of the VP shunts and 48% of the VA shunts required at least one revision. This is not statistically different (p > .05). Table 3 summarizes the reasons for primary revisions of the VP and VA shunts regardless of the valve system used. Most revisions were distal, and related to the hazards of a foreign body in either the vascular system or peritoneal cavity that led to clotting in the shunt tubing or caused vascular occlusion of a major venous channel with atrial shunts and adhesions forming around the distal shunt tubing in the peritoneal placements. The VP distal revisions were often very minor and at times corrected under local anesthesia while the VA distal revisions usually required 680
general anesthesia and a search for a new portal because the vascular channel being used was thrombosed. Attention was then turned to an analysis of all the nonelective revisions to see if any consistent trends could be found. In the long run, the VP shunts required many more total revisions (58%) than did VA shunts (48%). However, the difference is not significant (p > .10). We next made an analysis of the time interval between the initial placement and first nonelective revision. Using the mean for comparison, we found that the interval between initial placement and primary revision was 4.8 months for the VP placement and 14.1 months for the VA placement. Another method of analyzing this interval is to determine the percentage of shunt placements that required a revision within a given interval; 18% of the VP shunts and 12% of the VA shunts required revision within the first month. However, for those patients who required no revision at the end of 2 years follow-up, only 3% of the VP shunts went on to require revision as opposed to 10% of the VA shunts. This indicates that the complication rate is highest in the first month after shunting and that although the VP shunts required a slightly higher rate of revision in the first month, in the long run the VP shunts do better in terms of requiring less primary revision. The problem of infection was studied in detail. Eighteen infections developed from primary placements; Table 4 summarizes them according to initial placement and diagnosis. In cases that required a primary revision, infection was a cause in VA shunts 1.6 times as often as in VP shunts. The mean interval from initial placement to diagnosis of infection was 2.0 months; although no mortality was associated with the infections, the morbidity in terms of requiring further revision and ultimate poor neurological function was great. The causative organism was coagulase-negative Staphylococcus epidermidis in 16 of the 18 infections. Over the last 2 years prophylactic antibiotics were used in all of the cases, and a drop (not statistically significant) in the infection rate occurred. In most cases, prophylaxis consisted of methicillin 25 mg/kg every 6 hours, 24 hours preand postoperatively, and dicloxacillin, 6 rag/ kg every 6 hours beginning on the second postoperative day and for 4 more days. J. Neurosurg. / Volume 42 / June, 1975
VP and VA shunts for hydrocephalus Discussion
TABLE 3
Hydrocephalus has been described in medical writings since the time of Hippocrates. In 1914, Dandy and Blackfan 5 first established the pathology of hydrocephalus and developed sound anatomical, physiological, and surgical principles for its treatment. Nulsen and Spitz 11 proposed ventriculoatrial shunting and the Holter valve-regulated shunt system in 1952, and in 1955, Scott, et a l . ) 8 described the use of a ventriculoperitoneal shunt. Since that time much controversy has existed over which is the better placement, but few large series have compared their clinical results? ,14 In an excellent review of the treatment of hydrocephalus, Scarf~ 6 tabulated the published results of the different methods up to 1963; he found a strong tendency for the VP shunt to become obstructed at the distal end and a high incidence (30% to 50%) of permanent failures, and an equally high rate of complications with the VA shunting method. More recent reviews have indicated that the VP shunt may be superior overall to the VA shunt. Robertson, et al., TM reported a 44% revision rate, and Little, et al.,8 a 47% revision rate for the VP shunt. Revisions were due mainly to obstruction of the distal end by adhesions, in agreement with our data. Ascites, TM perforation of a viscous, ~9 intraabdominal cyst formation, 1 volvulus, 1~ and perforation of the scrotum or vagina, ~s are reported but exceedingly rare complications of VP shunts. Revision rates for the VA placement in recent large series range from 46% to 78%. 2,~2 Complications of the VA placement included pleural effusion, pulmonary emboli, inferior vena cava occlusions, superior vena cava occlusions, and septicemia. 7 Infection rate has been reported as high as 30% with this placement. ~ Our results indicate that although the VP and VA placements carried primary and total revision rates that were not significantly different statistically, the complications were less life-threatening, carried a lower morbidity, and were easier to correct in VP placements as compared to VA placements. It was also apparent that there was less need for elective revision due to body growth with the VP than with the VA placement. Our finding that the VA shunt carried a higher rate of infection than the VP shunt
Reason ]br revision of primary shunt
J. Neurosurg. / Volume 42 / J u n e , 1975
No. of Cases
Cause of Revision
VP: 67 Revisions, 58 % of initial placements
proximal: particulate matter wrong position other distal: omental adhesions peritoneal adhesions fibrin plug in catheter other infection
7 2
l0 (15%)
1
25 12
49 (71%)
9 4
7 (10.6%)
VA: 85 Revisions, 48 % of initial placements
proximal: particulate matter in ventricular catheter wrong position other distal : clot jugular thrombosis vena cava thrombosis other infection
27 (31%)
20 6 1 23 13 2 9
47 (55 %)
11 (13 %)
TABLE 4 Infection of initial shunt according to type of placemem and diagnosis
No. No. of Placed Infections type of placement:
vP VA
diagnos&:
114 177
aqueductal stenosis aqueductal stenosis and myelomeningocele tumor other
7 11
Vo
%
Revisions Due to Infection
6.1 6.2
10.6 13
4 4 4 6
agrees with other reports, s,* Staphylococcus epidermidis was the overwhelming cause of infection in other reported seriesJ ,9,17 The drop in infection rate over the last 2 years in our series since the introduction of prophylactic antibiotics, while not statistically significant, is in agreement with Yu and Patterson's work. 2~ Other factors that seem to be impor681
R. J. Ignelzi and W. M. Kitsch tant in the development of infection are the underlying disease process, the number and the type of preoperative procedures, the length of operation, and the number of postoperative diagnostic reservoir taps.
Summary Over the last 15 years, 300 patients with hydrocephalus were treated at the University of Colorado Medical Center and required a total of 664 operative procedures. When the effectiveness of the VA versus the VP placement was compared, no statistically significant difference as regards primary or total revision rates was found; however, VA shunts developed more serious complications, including infections, than did the VP shunts. VP shunts should be considered primarily because of the ease in placement, lower rate of serious complications, and ease of revision.
8.
9. 10. 11.
12.
13. 14.
Acknowledgment We would like to express appreciation to Franklin W. Briese, Ph.D., Department of Biometrics at the University of Colorado Medical Center, for his help in preparation of this manuscript.
References 1. Ames RH: Ventriculo-peritoneal shunts in the management of hydrocephalus. J Neurosurg
27:525-529, 1967 2. Becker DP, Nulsen FE: Control of hydrocephalus by valve-regulated venous shunt: avoidance of complications in prolonged shunt maintenance. J Neurosurg 28:215-226, 1968 3. Bruce AM, Lorber J, Shedden WIH, et al: Persistent bacteraemia following ventriculocaval shunt operations for hydrocephalus in infants. Dev Med Child Neurol 5:461-470, 1963 4. Callaghan RP, Cohen S J, Stewart GT: Septicaemia due to colonization of Spitz-Holter valves by staphylococci. Five cases treated with methicillin. Br Med J 1:860-863, 1961 5. Dandy WE, Blackfan KD: Internal hydrocephalus: an experimental, clinical and pathological study. Am J Dis Child 8:406-482, 1914 6. Fokes EC Jr: Occult infections of ventriculoatrial shunts. J Nenrosurg 33:517-523, 1970 7. Forrest DM, Cooper DGW: Complications of 682
15.
16.
17.
18.
19.
20.
ventriculo-atrial shunts. A review of 455 cases. J Neurosurg 29:506-512, 1968 Little JR, Rhoton AL Jr, Mellinger JF: Comparison of ventriculoperitoneal and ventriculoatrial shunts for hydrocephalus in children. Mayo Clin Proc 47:396-401, 1972 Luthard T: Bacterial infections in ventriculoatrial shunt systems. Dev Med Child Neurol (Suppl) 22:105-109, 1970 Murtagh F, Lehman R: Peritoneal shunts in the management of hydrocephalus. JAMA 202:1010-1014, 1967 Nulsen FE, Spitz EF: Treatment of hydrocephalus by direct shunt from ventricular to jugular vein. Surg Forum 2:399-403, 1952 Overton MC, Snodgrass SR: Ventriculovenous shunts for infantile hydrocephalus. A review of five year's experience with this method. J Neurosnrg 23:517-521, 1965 Ramani PS: Extrusion of abdominal catheter of ventriculoperitoneal shunt into the scrotum. Case report. J Neurosurg 40:772-773, 1974 Robertson JS, Maraqa MI, Jennett B: Ventriculoperitoneal shunting for hydrocephalus. Br Med J 1:289-292, 1973 Sakoda TH, Maxwell JA, Brackett CE Jr: Intestinal volvulus secondary to a ventriculoperitoneal shunt. Case report, d Neurosnrg 35:95-96, 1971 Scarff J: Treatment of hydrocephalus: an historical and critical review of methods and results. J Nenrol Neurosurg Psychiatry 26:1-26, ' 1963 Schimke RT, Black PH, Mark VH, et al: Indolent staphylococcus albus or aureus bacteremia after ventriculoatriostomy. Role of foreign body in its initiation and perpetuation. N Engl J Med 264:264-270, 1961 Scott M, Wycis HT, Murtagh F, et al: Observations on ventricular and lumbar subarachnoid shunts in hydrocephalus in infants. J Neurosurg 12:165-175, 1955 Wilson CB, Bertan V: Perforation of the bowel complicating peritoneal shunt for hydrocephalus. Report of two cases. Am Surg 32:601-603, 1966 Yu HC, Patterson RH: Prophylactic antimicrobial agents after ventriculo atriotomy for hydrocephalus. J Pediatr Surg 8:881-885, 1973
This research was supported by a National Institutes of Health Training Grant, 5T01, NS05564. Address reprint requests to: Ronald J. Ignelzi, M.D., Division of Neurosurgery, Veterans Administration Hospital, 3350 La Jolla Village Drive, San Diego, California 92161. J. Neurosurg. / Volume 42 / June, 1975
