PEDIATRIC DENTISTRY/Copyright © 1979 by The AmericanAcademy of Pedodontics / Vol. 1. No. 2 / Printed in U.S.A. SCIENTIFIC

Antibiotic prophylaxis for the hydrocephalicdental patient with a shuntTheodore P. Croll, D.D.S.Donald G. Greiner, D.D.S.Luis Schut, M.D.

AbstractSurgically placed shunts have proven useful in divertingexcess accumulation of cerebrospinal fluid. The properlyfunctioning shunt prevents increased intracranialpressure, gross craniofacial distortion, and thedevelopment of mental retardation, all of which havealways been associated with the untreated hydrocephalicpatient. The postoperative complication of microbialcolonization of shunts has been greatly reduced byperioperative antibiotic coverage. The most frequentcolonizing bacteria are found in the flora of the mouth,oropharynx, and nasopharynx. The authors havepresented a protocol for prophylactic antibiotic coverageduring dental therapy.

Literature review

Hydrocephalus is a pathologic condition character-ized by abnormal accumulation of cerebrospinal fluid(CSF) within the brain fluid pathways.1

In the normal brain, the choroid plexus and certainlining cells secrete cerebrospinal fluid into the ventric-ular system. The fluid bathes the neural tissue of thebrain and cord and serves also to cushion and protectit. The ventricles are constantly being replenishedwith cerebrospinal fluid with a secretion rate of ap-proximately 30 ml every 90 min.1

Cerebrospinal fluid is drained into the circulatorysystem at the sites of the arachnoid villi of the thinmembranes covering the brain. The villi apparentlyfunction both as a result of pressure differentials be-tween the brain fluid and venous circulation and alsoby the relative pore size of the villi, facilitating fluiddrainage.1

When flow of CSF is increased or when normal flowpathways become altered or obstructed, abnormalfluid accumulation may result, swelling the ventricularsystem. Cerebral tissue may therefore be distorted,compressed, and/or displaced by the additional fluidpressure. Fig. 1 is a photograph of an untreated hydro-cephalic child.

According to Schut and Bruno,~ there are numerouscauses for hydrocephalus. These include overproduc-tion of CSF, obstruction in the aqueducts connectingventricles, passive ventricular enlargement due to lossof surrounding brain tissue from aging or cerebrovas-cular accident, Dandy-Walker syndrome, abnormaldevelopment of the cisterna magna and its connectionsover the brain surface, and insufficient CSF circulatoryabsorption due to failure of either villi function orvenous drainage of the arachnoidal spaces.

Hydrocephalus may be congenital or acquired. Ac-quired hydrocephalus may result from infection, hem-orrhage, or tumorous growth.

There are terms which further describe the natureof the abnormality. Obstructive or noncommunicatinghydrocephalus occurs when CSF is unable for somereason to flow from the ventricles into the cisterns.Nonobstructive or communicating hydrocephalus re-fers to abnormal fluid accumulation due to somedisorder in circulatory reabsorption of CSF. In thelatter type, there is free flow of fluid among the fourventricles, but normal CSF drainage is impeded.

There is no known way to prevent hydrocephalus,but great advances have been made in the last 20years in controlling the disorder. Shunting techniqueshave been developed by surgically placing a small tube

PEDIATRIC DENTISTRYVol. 1, No. 2 81

Fig. 1. An untreated hydrocephalic child is pictured. Notethe gross craniofacial distortion and enlarged cranial dimen-sions.

in a distended lateral ventricle and by diverting excessCSF through the tube into and through a valvedpumping chamber to a distant absorption site some-where in the body.2

The most common types of shunts are the ventric-ulo-peritoneal and ventriculo-venous shunts. Theshunt system consists of three components which areconnected at the time of surgical placement. Typicalshunts are pictured in Fig. 2.

The shunt operation consists of insertion of a ven-tricular catheter through a parietal bur hole into alateral ventricle, generally on the right side. The cath-eter's distal end is connected to a pumping chamberattached to the skull. The third component, a longdistal catheter attached to the valved pumping cham-ber, runs, in the case of one shunt type, subcutane-ously in the neck and enters venous circulation gen-erally by way of the common facial or internal jugularvein or enters directly into the right atrium.

In the ventriculo-peritoneal shunt, the distal cathe-ter continues on over the chest wall and enters theabdominal cavity through a subcostal incision.1"' Thedistal tip of the catheter ends in the pelvis where it is

Fig. 2. Two typical shunts are shown. Note the short proxi-mal catheters, the valved pumping chambers, and the longdistal catheters.

least likely to be obstructed by the omentum.2 Thediagrams in Fig. 3 illustrate the most commonly usedshunts.

Carefully placed shunts are known to give manyyears of excellent service. However, due to humangrowth and development and to various types of com-plications, shunts must periodically be revised. Com-plications of shunts include obstruction of the shuntcomponents by blood, debris, or anatomical meander-ing of the shunt tip;4 ' separation of shunt components;pressure necrosis of the tissues affected by the shunt;and an infection rate reported to be between 6% and23%.6

The incidence of hydrocephalic disorders is about1/750 births.1 It is very often associated with otherabnormalities of the spine and cord.1 Since the shuntplacement procedure is now the second must commonof neurosurgical operations and some 10,000 newshunts and 6,000 revisions are expected each year,prevention of shunt infection is of utmost concern.

Shunt infectionInfection of a shunt system results when microor-

ganisms "gain access at the time of valve insertion or

ANTIBIOTICS FOR HYDROCEPHALIC PATIENTS82 Croll, Greiner, and Schut

Fig. 3. The two most commonlyused shunts are depicted.On the left is pictured a ventric-ulo-peritoneal shunt, and on theright is a ventriculo-venousshunt.

during bacteremia at a later date.’’2 The bacteria col-onize the foreign material of the shunt, which leads tosepticemia and ventriculitis with potentially fatal con-sequences. It is generally felt that ventriculo-perito-neal shunts are less susceptible to infection from bac-teremia than are ventriculo-venous shunts, since theformer have no intimate circulatory communication.7

Ventriculo-peritoneal shunts, however, are not with-out risk of infection by staphylococci.8 There is con-troversy as to whether microorganisms which infectventriculo-peritoneal shunts are inoculated at the timeof shunt placement, or whether they gain access duringa subsequent bacteremia. This question is still open toacademic debate.

Once a shunt infection has been diagnosed, the mostwidely accepted treatment is antibiotic chemotherapyin conjunction with shunt revision in a new anatomicsite. Chemotherapy without shunt revision has hadonly limited success and is not a widespread mode oftreatment.2

Problem

Due to inevitable bacteremia from gingival hemor-rhage resulting from various dental procedures,9’ ,odental practitioners prophylactically cover at-risk car-diac patients with specific antibiotics to prevent sub-acute bacterial endocarditis.~l The antibiotic regimensare well documented in literature and have been re-cently altered to increase the antibiotic dosage anddecrease the duration of therapy.12

A comprehensive review of the medical and dentalliterature of the past 20 years reveals a conspicuousabsence of information regarding antibiotic coverage,for the hydrocephalic dental patient with a surgically

placed shunt in operation. No shunt infection has beendirectly related to dental procedures, but that doesnot preclude the possibility. Only Holt alludes to oraland nasopharyngeal sources of staphylococci andquestions their role in possible etiology of shunt infec-tions.’3

The purpose of this paper is to examine the problemof shunt infection, offering a protocol for special careof the hydrocephalic dental patient. Since the mostcommon shunts now being used are the ventriculo-peritoneal and ventriculo-venous shunts, the discus-sion will be limited to these types. Ventriculo-pleural,ventriculo-ureteral, lumbo-ureteral , and other shunttypes are rarely performed in the United States.~’ 2

Discussion

There are a number of important questions whoseanswers are crucial to the resolution of the problem athand. (1) Are the organisms most often implicated shunt infections available in the oral ecology for cir-culatory inoculation during a dental procedure? (2)Are there sufficient numbers of organisms to facilitateshunt colonization if bacteremia does occur? (3) Canan infective organism be transmitted to a patient fromthe hands of a dentist, hygienist, or dental assistant?(4} Does the severity of the dental procedure meritspecial attention? For example, should a dental pro-phylaxis be as much of a concern as multiple dentalextractions? (5) If antibiotic coverage is deemed ad-visable, what organisms should be the target of ther-apy, and what is the antibiotic of choice?

Without the benefit of data from a carefully de-signed animal study or from a long-term evaluation ofhuman subjects relating dental treatment history to

PEDIATRIC DENTISTRYVol. 1, No. 2 83

subsequent infection, we are compelled to conjecturesome of the answers guided by information presentlyavailable.

As previously noted, 6-23% of shunts become in-fected and require surgical replacement for this rea-son.~ The spectrum of microorganisms implicated iswide indeed. In a study of 289 human subjects, Schoen-baum et al. s demonstrated shunt infections with suchvaried flora as a- and fl-streptococcus, Neisseria men-ingitidis, a diphtheroid species, enterococcus, and oth-ers. s The research of Schoenbaum et al. agrees withthat of other researchers4’ ~3-~s that the most commoninfective microorganisms are Staphylococcus aureus(coagulase-positive) and Staphylococcus epiderrnidis(coagulase-negative).

Burnett and Scherp~9 report that both the staphy-lococci are generally present in the mouth even thoughthey constitute a minor fraction of the oral flora.

Peterson and Peacock9 tested 80 children betweenthe ages of 5-13 years who were undergoing dentalextractions. Thirty-nine of them gave positive bloodcultures after the surgical procedure. Of the 39 sub-jects, 18 had diseased teeth extracted while the other21 had clinically healthy teeth removed for orthodon-tic considerations. The organisms cultured from the39 patients included a-hemolytic streptococcus,peptostreptococcus, diphtheroids, coagulase-negativestaphylococcus, bacteroides, veillonella, neisseria, anda vibrio.9

Even though there are researchers who deny bac-teremia after oral manipulation in children, 2° datafrom Peterson’s study9 and from that of DeLeo et al.~°make it apparent that microorganisms capable ofshunt infection are available in the oral cavity forcirculatory inoculation and that microorganisms havebeen shown to be introduced into the blood duringdental treatment.

Presently, there is no way to be certain as to whetherS.. aureus or S. epidermidis exist in the mouth withsufficient number and virulence to colonize a shunt ifbacteria are introduced into the oral vascular supply.As previously mentioned, a strictly controlled animalexperiment may lend insight into this question. Wecan only surmise that this route of shunt colonizationcannot be ruled out. We therefore must weigh therisks of covering the hydrocephalic dental patient withantibiotics versus the risk of not doing so until exper-imental data give us a definitive answer.

It is doubtful that the dental practitioner, hygienist,or dental assistant may act as carriers for staphylococ-cus transmission if usual hand washing and steriliza-tion techniques are used. Shay and Clendenin2~

showed that in 19 attempts, a carrier dentist did notinoculate a noncarrier patient with S. aureus during a30-minute dental prophylaxis.

The question of severity of dental operation also

has no absolute answer. However, DeLeo’s group stud-ied 39 children who demonstrated a 5% positive bloodculture prior to oral prophylaxis and a 28% positiveblood culture after the procedure. The preoperativebacterial strains were identified as two anaerobic diph-theroids and two aerobic, coagulase-negative staphy-lococci. Twenty-eight separate strains were grownafter the dental procedure, including 7 aerobes (twodiphtheroids, two strains of Streptococcus viridans,and three coagulase-negative staphylococci); and 21anaerobes (nine diphtheroids, five strains of Veillo-nella alcalescens, three of Bacteroides rnelaninogen-icus, two of Peptostreptococcus anaerobius, one eu-bacterium, and one fusobacterium).~° It is importantto remember that these findings followed a procedureas innocuous as a routine dental prophylaxis.

It seems, there.fore, that any procedure more exten-sive than a tray application of fluoride gel or solutionmay result in some amount of gingival bleeding witha possibility of bacteremia. Endodontic or surgicalprocedures obviously have this potential.

Tinanoff et al. ~ have raised a question as to thebenefit of a pumice prophylaxis if only soft deposits ofplaque exist on the teeth. With the exception of cal-culus and gross extrinsic stain removal in a patientwith healthy gingiva, it is possible that a gentle, softtooth brush or tray application of a fluoride prepara-tion during each periodic visit may be the best servicefor the hydrocephalic patient not requiring other ther-apy involving gingival hemorrhage.

Prophylactic antibiotic coverage is used by manyneurosurgeons3, ~, s, ~5-~7 at the time of shunt placementto prevent infection by S. aureus and S. epiderrnidisintroduced during the operation. There is a generalconsensus that prophylactic coverage does decreasesubsequent infection. The antibiotic agents mostwidely employed are oxacillin, methicillin, cloxacillin,or dicloxacillin, all of which are effective against thetwo staphylococci strains.

RecommendationsWith the current state of knowledge concerning

possible oral etiology of shunt infection in the hydro-cephalic dental patient, the following recommenda-tions are made.1. The dentist should be familiar with the medical

history of the hydrocephalic patient, including thetype of shunt in place and its history. The patient’sphysician should be advised of the oral therapyrequired to keep all health care providers currenton the patient’s medical welfare.

2. As with all patients, meticulous hand washing andsterilization procedures are important.

3. If antibiotic coverage is to be employed, it is prac-tical to achieve as much treatment as possible pervisit. "Half mouth" or at least "quadrant" restora-

ANTIBIOTICS FOR HYDROCEPHALIC PATIENTS84 Croll, Greiner, and Schut

tive dentistry is advised, depending on the difficultyof the treatment and the comfort of the patient.

4. Extensive efforts by the dental practitioner shouldbe made in preventive dentistry education for thehydrocephalic patient and his/her parents. Fre-quent professional reinforcement of preventive in-formation and supervision of home care skills canbe valuable aids in eliminating the need for recur-rent corrective dental therapy.

5. For oral manipulative procedures that may causegingival hemmorhage, an antibiotic regimen similarto that of the perioperative period for shunt surgeryshould be employed.

a. For children weighing less than 40 kg (88 lb}, g dicloxacillin orally 30-60 min prior to the den-tal procedure and then 250 mg p.o. every 6 hr for8 doses.

b. For children over 40 kg (88 lb), an initial oraldose of 2 g dicloxacillin should be given 30-60min prior to the procedure followed by 500 mgp.o. every 6 hr for 8 doses.

c. For patients with a known allergy to the penicil-lins, we recommend the use of a cephalosporin.The dosage is 50 mg/kg of weight per day, di-vided into three doses. The regimen should bestarted one day prior to the procedure and con-tinue through two days postoperatively.

There are cases reported of cross-sensitivity be-tween the penicillins and the cephalosporins, andwhile this must be considered, the cephalosporins arestill the safest alternative at the present time. Somesynthetic cephalosporins are currently being testedwhich show great promise with no penicillin cross-sensitivity, but these products are not yet commer-cially available. In the extremely rare instance that across-allergenicity is known, further consideration ofother effective antibiotics is advisable.

Even though ventriculo-peritoneal shunts may notbe as susceptible to colonization from bacteremia asare ventriculo-venous shunts, it is unwise presently topresume that blood-borne bacteria are incapable ofinfecting the former type. Until research providesdefinitive data which prove that ventriculo-peritonealshunts cannot be infected by bacteremia unrelated toneurosurgical procedures, the above antibiotic prophy-lactic regimens should be employed for both shunttypes.

SummaryA cursory review of hydrocephalus and of surgically

placed shunts to control the disorder is presented. Asubjectively devised protocol is offered for prophylac-tic antibiotic coverage for the hydrocephalic dentalpatient with a shunt in place.

It is recognized that more definitive knowledge con-cerning this problem is needed, and research to answer

questions posed in this discourse is suggested by theauthors.

Acknowledgement

The authom would like to thank Dr. Eric B. Brooks for hisvaluable comments and suggestions.

References

1. Schut, L. and Bruno, L. A.: The Child With Hydrocephalus,Philadelphia: V. J. Sarte National Hydrocephalus Foundation,Inc., 1977.

2. Stark, G. D.: Spina Bifida, Problems and Management, Lon-don: Blackwell Scientific Publications, 1977, pp. 88-102.

3. Salmon, J. H.: "Surgical Treatment of Hydrocephalus in In-fancy," Hosp Pract, 13:80-88, 1978.

4. Forrest, D. M. and Cooper, D. G. W.: "Complications of Ventri-culo-atrial Shunts," A Review of 455 Cases," J Neurosurg, 29:506-512, 1968.

5. Brook, I., Johnson, N., Overtttrf, G. D., and Wilkens, J.: "MixedBacterial Meningitis: A Complication of Ventriculo- andLumbo-peritoneal Shunts," J Neurosurg, 47:961-964, 1977.

6. Yu, HC. and Patterson, R. H.: "Prophylactic AntimicrobialAgents After Ventriculoatriostomy for Hydrocephalus," J Pe-diat Surg, 8:881-885, 1973.

7. Bruce, D. A.: Personal communication, February, 1979.8. Schoenbaum, S. C., Gardner, P., and Shillito, J.: "Infections of

Cerebrospinal Fluid Shunts: Epidemiology, Clinical Manifesta-tions, and Therapy," J Infec Dis, 131:543-552, 1975.

9. Peterson, L. J. and Peacock, R.: "The Incidence of Bacteremiain Pediatric Patients Following Tooth Extraction," Circulation,53:676-679, 1976.

10. DeLeo, A. A., Schoenknecht, F. D., Anderson, M. W., andPeterson, J. C.: "Incidence of Bacteria Following Oral Prophy-laxis on Pediatric Patients," Oral Surg, 37:36-45, 1974.

11. American Heart Association: "Prevention of Bacterial Endocar-ditis," Circulation, 56:139, 1977.

12. American Heart Association: "Prevention of Bacterial Endocar-ditis: A Committee R.eport of the American Heart Association,"J Am Dent Assoc, 95:600-605, 1977.

13. Holt, R.: "The Classification of Staphylococci From ColonizedVentriculo-atrial Shunts," J Clin Pathol, 22:475-482, 1969.

14. Shurtleff, D. B., Christie, D., and Folitz, E. L.: "Ventriculo-auriculostomy-associated Infections, A 12 year Study," J Neu-rosurg, 35:686-694, 1971.

15. Fokes, E. C.: "Occult Infections of Ventriculoatrial Shunts," JNeurosurg, 33:517-523, 1970.

16. Venes, Jo L.: "Control of Shunt Infection," J Neurosurg, 45:311-314, 1976.

17. Perrin, J. C. S., and McLaurin, R. L.: "Infected Ventriculo-atrialShunts. A Method of Treatment," J Neurosurg, 27:21-26, 1967.

18. Shurtleff, D. B., Foltz, E. L., Weeks, R. D., and Loeser, J.:"Therapy of Staphylococcus Epidermidis: Infections Associatedwith Cerebrospinal Fluid Shunts," Pediatrics, 53:55-62, 1974.

19. Burnett, G. W. and Scherp, H. W.: Oral Microbiology andInfectious Disease, 3rd ed., Baltimore: The Williams & WilkinsCompany, 1968, p. 279.

20. Hurwitz, G. A., Speck, W. T., and Keller, G. B.: "Absence ofBacteremia in Children after Prophylaxis" Oral Surg, 32:891-894, 1971.

21. Shay, D. E. and Clendenin, G. G.: "Incidence of Coagulase-positive Staphylococci in the Upper Respiratory Tract of DentalStudents and a Study of Their Transmission During a RoutineDental Prophylaxis," J Dent Res, 42:110, 1963.

22. Tinanoff, N., Wei, S. H. Y., and Parkins-, F. M.: "Effect ofPumice Prophylaxis on Fluoride in Tooth Enamel," J AmDental Assoc, 88:384-389, 1974.

PEDIATRIC DENTISTRYVol. 1, NO. 2 85

Dr. Theodore P. Croll practices pedi-atric dentistry in Doylestown, Pennsyl-vania. He is a clinical associate in theDepartment of Pedodontics/Orthodon-tics at the University of PennsylvaniaSchool of Dental Medicine and a staffpedodontist in the Division of Dentistry,The Children's Hospital of Philadel-phia. Dr. Croll served from 1973-1976as a captain and a major in the U.S.Army Dental Corps in Hanau, FederalRepublic of Germany. Requests for re-prints may be addressed to: Dr. T. P.Croll, Division of Dentistry, The Chil-dren's Hospital of Philadelphia, 3401Civic Center Boulevard, Philadelphia,Pennsylvania 19104.

Dr. Donald G. Greiner is a 1963 grad-uate of Case Western Reserve DentalSchool. Prior to his pedodontic train-ing, he spent five years in the U.S.Army Dental Corps, which period in-cluded an internship at Walter ReedHospital. He received a pedodonticcertificate and a Master of Science de-gree from Ohio State University in1970. From 1970-1973, he was anassistant professor of pediatric den-tistry at the University of Connecticut.In 1973, he entered private practiceand was full-time in practice until 1976when he bacame Director of Dentistryat Newington Children's Hospital. Hecurrently spends half-time in practiceand half-time at the hospital.

Dr. Luis Schut is a 1954 graduate ofthe University of Buenos Aires MedicalSchool. He has had faculty appoint-ments at the University of PennsylvaniaSchool of Medicine since 1957 and iscurrently an associate professor ofneurosurgery there. He is a chartermember of both the American and In-ternational Associations of PediatricNeurosurgery. He currently serves onthe executive committee of the Ameri-can Society of Pediatric Neurosurgery.Dr. Schut has been the Chief of Neu-rological Services at Children's Hospi-tal of Philadelphia since 1969.

ANTIBIOTICS FOR HYDROCEPHALIC PATIENTS86 Croll, Greiner, and Schut