J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.67.4.468 on 1 October 1999. Downloaded from 468 J Neurol Neurosurg Psychiatry 1999;67:468–473 Bacterial meningitis associated with lumbar drains: a retrospective cohort study

William M Coplin, Anthony M Avellino, D K Kim, H Richard Winn, M Sean Grady

Abstract risk of meningitis to consider when using this Objectives—The infective potential of modality, and the infective potential of lumbar lumbar drainage is an important topic drainage is an important topic deserving deserving particular study. The aetiology, particular study. We conducted a retrospective incidence, and clinical findings associated cohort study of patients with positive CSF cul- with bacterial meningitis are described in tures after lumbar drain placement to ascertain patients having continuous lumbar CSF the possible aetiologies of meningitis. Our drainage to treat communicating hydro- objectives included defining the incidence of cephalus after subarachnoid haemor- meningitis after lumbar drain placement, rhage or CSF leaks after traumatic dural describing the association of positive CSF cul- rents. tures with the clinical signs and laboratory Methods—Retrospective review of the values considered diagnostic of meningitis, and records of patients with a positive CSF describing potential risk factors for contracting bacterial culture who underwent lumbar lumbar drain associated meningitis. drain placement over a 39 month period. Results—Thirteen cases of bacterial men- Patients and methods ingitis occurred subsequent to the use of PATIENT SELECTION 312 lumbar drain kits (4.2%). All menin- The method of patient selection was identifica- gitic patients had CSF pleocytosis, but not tion of all cases of bacterial meningitis. Cases all had peripheral leukocytosis. Fever, established by culture were identified from a peripheral leukocytosis, and CSF pleocy- list of all positive CSF cultures at Harborview tosis did not help to diVerentiate the pres- Medical Center, Seattle, WA, USA between ence of bacterial meningitis from other June 1992 and December 1995. Hospital infections. Eight patients had prior CSF charts were then screened for those who drainage procedures, including ventricu- received a lumbar drain. We reviewed in detail lostomy (n=5) or lumbar drain (n=5) the records of those patients who underwent placements; two patients received both lumbar drainage. Patients’ charts were used to procedures. Six of 13 patients developed confirm information from other data sources their CSF infection within 24 hours of (for example, microbiology laboratory lumbar drain insertion. Six of 13 patients records). Diagnoses treated by lumbar drains developed meningitis while receiving anti- included communicating hydrocephalus after http://jnnp.bmj.com/ biotics for other reasons. subarachnoid haemorrhage and traumatic cra- Conclusions—External lumbar drainage nial CSF leak. The degree of hydrocephalus or Departments of seems to carry a low risk of infectious CSF leak and the patient’s clinical condition Neurological Surgery meningitis and oVers a safe alternative to determined the duration of drainage. We and Neurology, ventriculostomy or serial lumbar punc- obtained our denominator (the number of Division of Pulmonary tures. Antibiotics do not seem to protect lumbar drain placements over this time) from and Critical Care completely against developing the infec- Medicine, Harborview intrahospital usage of lumbar drain kits. The tion. The infection happens most often number of kits used was likely higher than the Medical Center, on September 26, 2021 by guest. Protected copyright. University of with skin organisms. The meningitis often number of patients treated as there were Washington, Seattle, appears within 24 hours after lumbar patients who had more than one drain placed WA, USA drain placement. Daily CSF samples during their stays in hospital. The Human W M Coplin should include bacterial cultures but cell Subjects Review Committee of the University A M Avellino counts may not oVer any additional useful D K Kim of Washington approved this research activity. H R Winn information in diagnosing the complica- M S Grady tion. Lumbar drain insertion and man- LUMBAR DRAIN TECHNIQUE agement need not be confined to the After shave (if necessary), skin preparation Correspondence to: intensive care unit. (povidone-iodine solution), and draping of the Dr William M Coplin, (J Neurol Neurosurg Psychiatry 1999;67:468–473) Departments of Neurology lumbosacral area, 1% lidocaine was infiltrated and Neurological Surgery, between the palpated spinous processes of the Wayne State University, 4201 Keywords: cerebrospinal fluid; lumbar drainage; St Antoine, 8D-UHC, meningitis lower lumbar vertebrae. A 16G Tuohy spinal Detroit, MI 48201, USA. needle was advanced until it punctured the Telephone 001 313 577 meninges. A CSF sample was sent for routine 1242; fax 001 313 745 4216; email Continuous lumbar CSF drainage may serve as studies (cell count and diVerential, glucose, [email protected] a temporising measure in treating communi- protein, gram stain, and culture). The lumbar cating hydrocephalus caused by aneurysmal drain was threaded into the subarachnoid Received 6 July 1998 and in subarachnoid haemorrhage12and is one treat- space. The needle was withdrawn and the drain revised form 25 March 1999 ment for correcting CSF leaks (spinal or attached to a drainage bag. A silk tie secured Accepted 24 April 1999 cranial). Similarly to ventriculostomy, there is a the adapter connecting the intraspinal catheter J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.67.4.468 on 1 October 1999. Downloaded from Bacterial meningitis associated with lumbar drains 469

and the external drainage system. Sterile gauze protein, Gram stain, culture, and sensitivities of dressed the skin wound. A transparent adhesive organisms isolated (if any). dressing covered the external part of the intraspinal catheter to prevent dislodgment. DEFINING VARIABLES Patients did not receive specific antibiotic Meningitis associated with lumbar drains was prophylaxis for the procedure. defined before starting the review. Among Samples of CSF were obtained from a sam- other variables, the review sought to explore pling port along the drainage catheter. This the possible association of leukocytosis, fever, port lies just distal to the junction of the and pleocytosis (predominance of segmented indwelling catheter and its connection to the neutrophils) with meningitis diagnosed by remainder of the drainage system leading to the positive CSF culture. These definitions were collection bag. This site has a three way port derived from a prior study of infections related that was closed to the drainage bag side for to ventriculostomy.3 The incidence of ventricu- sampling. Also, before sampling, the mem- lostomy infections was beyond the scope of this brane over the sampling port was swabbed with study. a povidone-iodine solution that was allowed to The diagnosis of meningitis associated with dry completely before a sterile needle was lumbar drains required the following criteria: introduced through the membrane to obtain (1) no previous meningitis before lumbar drain the CSF sample. insertion; (2) sterile CSF culture at the time of lumbar drain insertion; (3) continuous lumbar drainage>24 hours before the positive culture CLINICAL AND EPIDEMIOLOGICAL DATA was aspirated; and (4) positive culture of a CSF ABSTRACTION specimen collected from the lumbar drain or The following demographic data came from from a lumbar puncture. the charts of patients who had a positive bacte- Fever was defined as a temperature>38.5°C. rial CSF culture and a lumbar drain: age, hos- Association of fever with meningitis required pital admission date, admitting neurological the following criteria: (1) fever within 48 hours diagnosis (for example, subarachnoid haemor- of the positive CSF culture; (2) no other rhage, traumatic CSF leak), and if the patient concurrent infection; (3) if the patient had had a history of diabetes mellitus. Operations fever within 48 hours before the positive CSF were recorded (including if the surgeon entered culture, then a temperature rise>0.6°C within the ventricles) as well as neurological instru- 48 hours of the positive CSF culture; or (4) mentation (for example, placement of intracra- fever lasting for 3 days. nial pressure monitors or ventriculostomies). Peripheral leukocytosis was defined as Other clinical data were gathered—namely, >10 000 white blood cells (WBC)/mm3 of other positive CSF cultures (lumbar puncture, blood. The attribution of the leukocytosis to ventriculostomy); other microbiological cul- meningitis required: (1) leukocytosis occurring ture results (for example, blood, urine, sputum, within 48 hours of the positive CSF culture; (2) wound); CSF cell counts for 48 hours before no other concurrent infection; (3) if the WBC and after drawing the culture positive sample; count was>10 000 48 hours before the positive temperatures for 48 hours before and after CSF culture, then an increase in WBC>1000/ 3

drawing the culture positive sample; signs of mm within 48 hours of the positive culture; or http://jnnp.bmj.com/ drain site infection (for example, leaking CSF, (4) leukocytosis lasting for 3 days. purulence, erythema, induration); duration of Pleocytosis in CSF was defined as a CSF lumbar drainage; peripheral white blood cell WBC count>11 cells/mm3. Attributing CSF count; intracranial and CSF opening pressures; pleocytosis to acute bacterial meningitis re- antibiotic use before identification of infection; quired: (1) >50% of the WBC were segmented CT findings (for example, presence of hydro- neutrophils; (2) pleocytosis occurring within cephalus or intraventricular or intraparenchy- 48 hours of the positive CSF culture; (3) if

mal haemorrhage); and most recent chest there was pleocytosis 48 hours before the posi- on September 26, 2021 by guest. Protected copyright. radiography findings. tive culture, then an increase of>50 WBC/mm3 The location in the hospital at which the within 48 hours of the positive culture; or (4) physician placed the lumbar drain (for exam- pleocytosis lasting at least 3 days. ple, intensive care unit, intermediate care unit, or ward), and on which ward the patient was at DATA ANALYSIS the time of positive CSF culture were noted. A Data were entered into a computerised data- nursing documentation sheet was used to base (SPSS for Macintosh, version 6.1.1, SPSS determine dressing and drainage system Inc, Chicago, IL, USA) University of Washing- change frequencies; the amount, colour (for ton computer resources were used to analyse example, xanthachromic v colourless), and the data. Analysis consisted of Fisher’s exact clarity (for example, cloudy v clear) of CSF test (two sided) along with the descriptive sta- drained; and the vertical position of the drain in tistics. relation to the patients’ head. This record included the presence of CSF leaking from the Results drain insertion site and drain manipulations. The initial review identified 24 patients with The profile of the culture positive CSF (cell positive CSF cultures and a case record of count and diVerential, glucose, protein, and lumbar drain insertion. Excluded from detailed gram stain) was recorded. Standard lumbar review were 11 of these 24 patients whose CSF drain protocol included daily collection of CSF cultures were considered to be plate contami- samples for cell count and diVerential, glucose, nation (for example, colony count of 1; n=8), J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.67.4.468 on 1 October 1999. Downloaded from 470 Coplin, Avellino, Kim, et al

Table 1 Clinical characteristics of 13 patients with lumbar drain-associated bacterial patient had positive cultures both from lumbar meningitis drain and lumbar puncture samples; the same organism grew from both CSF samples. Neurological Prior Drainage Concurrent Patient Age (y) problem surgeries duration* antibiotics Organism(s) Other infections (bacteraemia, pneumonia, or urinary tract or wound infections) occurred 1 47 SAH None 7 No K pneumoniae 2 53 SAH V, C 3 Yes CNS in five of 13 cases of lumbar drain associated 3 64 SAH V, C 4 Yes CNS meningitis. Five of the patients became bacter- 4 34 SAH C 1 No CNS aemic within 48 hours of the positive CSF cul- 5 69 see below† C, L 9 No P aeruginosa 6 53 SAH V, C 3 No Actinobacter calcoaceticus ture sample; only one of these patients 7 24 BSF None 1 No group A Streptococcus developed bacteraemia with an organism with 8 84 SAH None‡ 1 No CNS the same antibiotic sensitivity pattern as that 9 48 SAH V, C 1 No S aureus 10 61 SAH V, C§ 2 Yes CNS, Corynebacterium isolated from CSF (coagulase negative staphy- species lococci). Five patients had positive sputum cul- 11 25 BSF M 1 Yes Enterobacter cloacae 12 42 AVM R, L 1 Yes CNS tures within 48 hours of the positive CSF cul- 13 50 SAH None 6 Yes S aureus ture; three of these grew mixed flora and the other two grew organisms diVerent from the *This number does not count the day of lumbar drain placement but does count the day the cul- ture positive sample was taken. one causing meningitis. Only one of the 13 †Patient 5 had a CSF leak after sinus surgery. patients met clinical criteria for the diagnosis of ‡All patients had prior intracranial pressure monitors, except this one. pneumonia (leukocytosis, fever, purulent spu- §Two separate craniotomies. SAH=subarachnoid haemorrhage; BSF=basilar skull fracture; AVM=arteriovenous malforma- tum, and new progressive chest infiltrate on tion; V=ventriculostomy; C=aneurysm clipping; L=CSF leak repair; M=repair of mandibular radiography). Four patients had positive urine fracture; R=resection; CNS=coagulase negative staphylococci. cultures within 48 hours of drawing the Table 2 Relation of fever, peripheral leukocytosis, and CSF pleocytosis to lumbar drain positive CSF sample; in one of these cases the associated bacterial meningitis same organism grew from blood and from CSF (coagulase negative staphylococci). One patient Patients without other infection Patients with other infection had a positive culture taken from the lumbar (n=8) (n=5) drain insertion site; this sample grew two Sign present Sign absent Sign present Sign absent p Value* organisms, one of which was the same as that causing the meningitis (coagulase negative sta- Fever4 4410.56 Peripheral leukocytosis 6 2 3 2 1.00 phylococci). One patient had a positive culture CSF pleocytosis 8 0 5 0 1.00 of an intravenous catheter tip; this organism was diVerent from the one associated with *By Fisher’s exact test (two sided). meningitis. In one patient the culture of the and those who had a proved prior CSF intracutaneous part of the lumbar drain infection (n=3). By the study’s criteria, bacte- resulted in growth of the same organisms as rial meningitis complicated 13 of 312 lumbar that from CSF (Corynebacterium minutissimum drains (4.2%) inserted during the study period. and S epidermidis). The age range of patients with meningitis was 24 to 84 years (median 50). Table 1 presents ASSOCIATION OF MENINGITIS WITH CSF demographic data for the 13 patients. Five PLEOCYTOSIS, FEVER, AND PERIPHERAL patients had previous ventriculostomy. Twelve LEUKOCYTOSIS

patients had a prior intracranial pressure Table 2 shows the relation of fever, peripheral http://jnnp.bmj.com/ monitor. Three patients had subarachnoid leukocytosis, and CSF pleocytosis with 13 haemorrhage without angiographic identifica- cases of bacterial meningitis. Fever and periph- tion of a source. Eight patients had a total of 11 eral leukocytosis are not specific clinical signs intracranial operations; seven of these patients for bacterial meningitis associated with lumbar underwent eight aneurysm clippings after pre- drainage. All of the patients with meningitis senting with subarachnoid haemorrhage; two had CSF pleocytosis. Although the presence of patients required repair of CSF leaks; one CSF pleocytosis seemed to be an insensitive

patient had resection of an arteriovenous mal- marker for bacterial meningitis, a rising CSF on September 26, 2021 by guest. Protected copyright. formation. One patient’s only surgery was polymorph count may have heralded the infec- repair of a mandible fracture. Three patients tion. Of the seven patients with meningitis for had cranial CSF leaks before lumbar drain whom preinfection WBC counts in CSF were insertion. Not all patients had daily CSF sam- available, all seven had CSF WBC counts>11 ples sent from their lumbar drains. cells/mm3; these counts increased to as high as 2208 cells/mm3. CULTURE DATA A positive CSF culture aspirated from the POSSIBLE RISKS FOR CONTRACTING BACTERIAL lumbar drain confirmed all cases. Table 1 MENINGITIS ASSOCIATED WITH LUMBAR DRAINS shows the culture results. The most common Table 3 describes the association of several organisms isolated were coagulase negative sta- putative risks for contracting bacterial menin- phylococci (n=7/13). Eleven of the 15 isolates gitis associated with lumbar drains. A major found in the 13 patients were gram positive problem of this uncontrolled analysis is that organisms. Four patients had meningitis from a few infections occurred over these 39 months; gram negative bacillus. One patient had CSF therefore, the associations investigated were infection with multiple skin organisms. Two not found to be significant. patients had positive cultures from CSF speci- There were some potential associations mens collected over 3 consecutive days; one between continuous lumbar CSF drainage and patient had positive cultures from CSF speci- the subsequent development of bacterial men- mens collected over 4 consecutive days. One ingitis. Table 1 shows that bacterial meningitis J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.67.4.468 on 1 October 1999. Downloaded from Bacterial meningitis associated with lumbar drains 471

Table 3 Description of putative risks for developing rates per ventriculostomy placement (0% of lumbar drain associated bacterial meningitis 116 and 30,7 4.5% of 65,8 8.9% of 213,3 and 9 No of patients 10.3% of 29 ). Another series compared Risk (%) tunnelled ventriculostomy (10% infection rate Neurological diagnosis (based on CT) for 70 placements) with the standard Rickham Subarachnoid haemorrhage 9 (69) reservoir (27% infected after 66 placements).10 Intraparenchymal haemorrhage 3 (23) Intraventricular haemorrhage 7(54) In a comparison of the eVect of antibiotic Hydrocephalus 8 (61) prophylaxis, others found infection rates of 9% Underlying diabetes 1 (7) (44 placements with antibiotics) to 27% (26 Neurosurgical operation 8 (61) 11 Ventricles entered 0 (0) placements without antibiotics). A prospec- Drain placed out of intensive care unit 6 (46) tive study did not find a reduced risk with anti- Antibiotics 6 (46) 3 Opening CSF pressure>20 mm Hg 4 (57)* biotic use. Our rate may diVer from that of the Intracranial pressure>20 mm Hg 3 (23) ventricular CSF drainage studies because of Drain duration >5 days 3 (23) the diVerent technique or definitions. Other Problem/manipulation Leaks 5 (38) investigators have defined meningitis associ- Disconnections 1 (7) ated with CSF drainage diVerently in studying System changes 7 (54) 3 6 9–11 Dressing changes 13 (100) ventricular drainage systems. Gross blood 0 (0) As to possible risk factors, most patients had Prior ventriculostomy 5 (38) Prior intracranial pressure monitor 12 (92) subarachnoid haemorrhage (nine of 13) or intraventricular haemorrhage (seven of 13), Risk set=13. but not intraparenchymal haemorrhage *Data not available for 6 patients; this risk set=7. (present in three of 13). Although four occurred in 10 of our 13 cases by day 4, and all ventriculostomy series have discussed neuro- cases appeared by day 9 of continuous lumbar logical diagnosis381011 only one explored the drainage; however, six of 13 infections oc- association of intraventricular haemorrhage curred with catheters in place for 1 day. As with CSF infection.3 Blood products may be a outlined in table 3, intraventricular haemor- good medium for culture and may block CSF rhage and subarachnoid haemorrhage resorption, preventing clearance of debris. had>50% coincidence in our cohort, and these Patients with intracranial hypertension or may be important, but significance could not intraventricular haemorrhage may require pro- be determined because of our small sample longed CSF drainage. We found intracranial size. hypertension at hospital presentation in only The meningitis was treated in all patients, three of 13 patients; lumbar drain placement except for one whose family decided to withdraw supportive measures. This patient CSF opening pressure was raised in four of was the only one with meningitis associated seven patients for whom these data were avail- with lumbar drain as a cause of death. able. Others have noted an association between Otherwise, the drain was removed, appropriate intracranial hypertension and the development intravenous antibiotics started, and, later, a of ventriculostomy related meningitis, inde- decision made as to whether to reinstitute lum- pendent of drainage duration.3 Neurosurgical procedures are reported to bar drainage. http://jnnp.bmj.com/ precede ventriculostomy related meningitis Discussion cases,12 and intracranial surgery preceded lum- INCIDENCE AND RISK OF BACTERIAL MENINGITIS bar drain insertion in eight of our 13 patients, ASSOCIATED WITH LUMBAR DRAINS but none had their ventricles breached in that Smaller series have considered the meningitis surgery. risk with lumbar drain use. One reported 7 Ten of 13 patients contracted meningitis days of lumbar drainage in 39 patients to treat within 4 days of lumbar drain placement, and

CSF leaks after skull base surgeries; none six did so within the first day of lumbar drain- on September 26, 2021 by guest. Protected copyright. received antibiotic prophylaxis. Ten developed age. Five series have considered the association infectious meningitis (25.6%). The authors of duration of ventriculostomy with the inci- concluded that concomitant antibiotic prophy- 3 8 10–12 4 dence of meningitis, two describing a laxis is unnecessary. In a series reviewing seven direct relation between risk of infection and years of experience with continuous lumbar duration of drainage. The present data do not CSF drainage for 48 patients with CSF leaks at support an optimal interval for routine lumbar operative sites (11 of 48) or CSF rhinorrhea drain change. In patients requiring continued (37 of 48), none developed infectious meningitis.5 In our cohort, the incidence of drainage, a fresh drain can be inserted or serial lumbar drain bacterial meningitis was 4.2%, lumbar puncture can be used. Five of the 13 which included nine patients with subarach- patients underwent prior ventriculostomy; 12 noid haemorrhage and four patients with trau- received an intracranial pressure monitor, and matic or iatrogenic CSF leaks. From our five patients were treated with a previous cohort and others, it seems that placement of (uninfected) lumbar drain. lumbar drains is safe, with minimal risk of Only one patient had disconnection of the infection. external CSF drainage system. Others have The incidence of bacterial meningitis of described a relation between infection and 4.2% in this series is as might be predicted ventriculostomy irrigation.311Opening the sys- from the ventriculostomy infection data. Oth- tem can allow bacterial passage into the ers have found a wide range of CSF infection subarachnoid space. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.67.4.468 on 1 October 1999. Downloaded from 472 Coplin, Avellino, Kim, et al

ASSOCIATION WITH OTHER INFECTIONS concerning the incidence of meningitis per This study defined bacterial meningitis as a diagnostic CSF examination. positive culture from a CSF sample withdrawn A patient’s own infected blood contaminat- from a lumbar drain. Eight of the initially iden- ing a spinal needle can directly introduce tified 24 patients did not have lumbar drain bacteria.28 A retrospective series found an inci- associated meningitis, but had contaminated dence of meningitis of 2.1% (three of 140) in CSF samples. Additionally, there were three patients receiving lumbar puncture while patients with bacterial meningitis before inser- bacteraemic. This was not a significant diVer- tion of the lumbar drain. ence from the expected 0.8% (seven of 924) Five of 13 patients had an infection other incidence of meningitis in patients similarly than meningitis, as shown in table 2. The lack bacteraemic who did not receive lumbar of a relation between clinical findings and hav- puncture.29 ing meningitis versus another infection may reflect the small sample size. POTENTIAL ERROR Having pleocytosis was not statistically The denominator in our record review (drain related to having meningitis. All patients had kits used) may not accurately reflect the CSF pleocytosis in response to culture estab- number of patients receiving drains or the lished infections. Given the potentially serious actual number of drains placed. For practical consequences of delayed diagnosis, it is of con- reasons, we were unable to look at all the cern that CSF cell counts do not oVer a reliable patients treated with lumbar drainage. Our method of diagnosis before culture results are denominator assumes that hospital inventory available. We did not ascertain the likelihood of records are accurate and that all kits are the converse—that is, patients with pleocytosis actually used for the specific purpose (and without culture proved infection. Pleocytosis none lost or discarded because of contamina- could result from non-infectious meningeal tion or technical diYculties, etc). It is likely that inflammation, such as to a foreign body (the some patients had more than one kit used. This lumbar drain). Others have shown CSF might underestimate the incidence of infection; pleocytosis after ventriculostomy with sterile however, while the incidence seen in our review cultures.3813 Thus, pleocytosis alone may not may be lower than the actual incidence, it still suYce in diagnosing lumbar drain associated falls well within the reported range for bacterial meningitis, and the diagnostic “gold ventriculostomy related infections. standard” remains CSF culture obtained from Data for this retrospective cohort study came either the drain or from lumbar puncture. We from three related but diVerent sources (micro- cannot comment further on the subject of a biological records, patient charts, and hospital rising polymorph count in the CSF as a inventories for lumbar drain kits). We used the harbinger of culture proved meningitis. Clini- data from one source (the patient chart) to cal judgment must dictate individual decisions cross check the others. Data from the microbi- regarding drain removal and initiation of ology laboratory were used to select the study therapy before having culture results. cohort. This is by contrast with identifying patients RELEVANCE AND CLINICAL SIGNIFICANCE TO with lumbar drains and reviewing those records

LUMBAR DRAIN USE to calculate an infection rate. A prospective http://jnnp.bmj.com/ There are three main temporising measures for study could identify patients in whom meningi- patients requiring continual CSF drainage: tis was thought to have existed (either on clini- serial lumbar puncture, lumbar drainage, and cal grounds or CSF findings) and treatment ventriculostomy. Hydrocephalus occurs in instituted even in the absence of positive 20%-31% of patients with subarachnoid haem- cultures or where organisms were seen on CSF orrhage, 40% improving within 24 hours with- examination and the cultures failed to grow. out intervention.14 15 Others need their com- Many patients undergoing lumbar drainage

municating hydrocephalus treated only with receive antibiotic therapy for other reasons (as on September 26, 2021 by guest. Protected copyright. transient CSF drainage, either by continuous in the present study), possibly giving falsely external catheter or by serial lumbar puncture. negative cultures in the face of true CSF Using long tunnelled ventriculostomy, and giv- infection. ing only perioperative antibiotic prophylaxis, the incidence of infectious meningitis is Conclusions reported to be 4%, similar to the rate in the Continual external lumbar CSF drainage seems present study.16 Using serial lumbar puncture to carry a low risk of culture proved bacterial to treat symptomatic hydrocephalus after meningitis (4.2%). Lacking obstruction to CSF subarachnoid haemorrhage, none of 17 pa- outflow, a lumbar drain may pose less infectious tients developed infectious meningitis.15 risk than a ventriculostomy. Skin organisms Serial lumbar puncture carries a risk of most commonly cause this infection, which infection and causes repeated patient discom- often appears within 24 hours of drain place- fort. Infection complicating lumbar puncture ment. Concomitant antibiotic use does not seem can happen after diagnostic or therapeutic to oVer protection; however, we examined the drainage of CSF,17–24 epidural anaesthesia,25 concurrent use of antibiotics as a group of agents myelography,26 27 or intrathecal infiltration (for and not the particular agent used. Thus, we example, chemotherapy administration). cannot assert that what protection, if any, antibi- Causative organisms include Staphylococcus otics may oVer; such a determination awaits aureus, and streptococci (S pneumoniae, S prospective evaluation. From our findings, we salivarius, etc).17 18 24 26 There are no clear data suggest: (1) Using a sterile technique, lumbar J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.67.4.468 on 1 October 1999. Downloaded from Bacterial meningitis associated with lumbar drains 473

drains can be safely placed and maintained both 8 Smith RW, Alksne JF. Infections complicating the use of external ventriculostomy. J Neurosurg 1976;44:567–70. in and out of acute care units (the intensive care 9 Bering EA Jr. A simplified apparatus for constant ventricu- unit); and (2) as the presence of leukocytosis, lar drainage. J Neurosurg 1951;8:450–2. fever, and pleocytosis may have little predictive 10 Kim DK, Uttley D, Bell BA, et al. Comparison of rates of infection of two methods of emergency ventricular ability in heralding the appearance of lumbar drainage. J Neurol Neurosurg Psychiatry 1995;58:444–6. drain associated bacterial meningitis, this com- 11 Wyler AR, Kelly WA. Use of antibiotics with external plication of lumbar drainage should be firmly ventriculostomies. J Neurosurg 1972;37:185–7. 12 Sundbarg G, Kjallquist A, Lundberg N, et al. Complications diagnosed from routine CSF culture. Our find- due to prolonged ventricular fluid pressure recording in ings support the contention that the presence of clinical practice. In: Intracranial pressure: experimental and clinical aspects. Berlin:Springer-Verlag, 1972:348–52. pleocytosis alone is insuYcient and not specific 13 Lundberg N. Continuous recording and control of ventricu- for the diagnosis of lumbar drain associated lar fluid pressure in neurosurgical practice. Acta Psychiatr bacterial meningitis; however, increasing CSF Scand 1960;149(suppl):1–193. 14 Hasan D, Vermeulen M, Wijdicks EFM, et al. Management cell counts may oVer some information valuable problems in acute hydrocephalus after subarachnoid hem- for assessing management strategies. Case- orrhage. Stroke 1989;20:747–53. 15 Hasan D, Lindsay KW, Vermeulen M. Treatment of acute control and prospective observational studies hydrocephalus after subarachnoid hemorrhage with serial could clarify better potential risk factors for lumbar puncture. Stroke 1991;22:190–4. contracting lumbar drain associated bacterial 16 Khanna RK, Rosenblum ML, Rock JP, et al. Prolonged external ventricular drainage with percutaneous long- meningitis. tunnel ventriculostomies. J Neurosurg 1995;83:791–4. 17 Amsel S, Fried D. Pneumococcal meningitis following lum- bar puncture. Harefuah 1983;105:264–5. This work was supported in part by National Institutes of 18 Bergman I, Wald ER, Meyer JD, . Epidural abscess and Health Grant No NS-30305 (WMC, HRW). We thank Kevin J et al Anstrom, for assistance with statistical analyses, Heather M vertebral osteomyelitis following serial lumbar punctures. Bybee, Dolors T Jones, Rosemary L Dolph, neurological Pediatrics 1983;72:476–80. surgery research nurses, for their assistance in identification of 19 Bocker H, Franek A. Meningitis following lumbar puncture cases, and the clinical laboratory, especially the microbiology [letter]. Dtsch Med Wochenschr 1987;112:120. section, at Harborview Medical Center for its assistance in 20 Domingo P, Mancebo J, Blanch L, et al. Iatrogenic strepto- clinical laboratory data collection. coccal meningitis [letter]. Clin Infect Dis 1994;19:356–7. 21 Helwig H. Meningitis following lumbar puncture [letter]. 1986;111:1461. 1 Weir B. Physiology and pharmacology of aneurysmal Dtsch Med Wochenschr 22 Lanska DJ, Lanska MJ, Selman WR. Meningitis following rupture. In: Aneurysms aVecting the nervous system. Balti- more:Williams and Wilkins, 1987:287. spinal puncture in a patient with a CSF leak. Neurology 2 Sato J, Sato O, Kamitani H, et al. Intracranial surgery and 1989;39:306–7. postoperative management of patients with ruptured aneu- 23 Smith KM, Deddish RB, Ogata ES. Meningitis associated rysm in acute and subacute stage. Basal cisternal drainage with serial lumbar punctures and post-hemorrhagic hydro- and lumbar subarachnoid drainage. Neurol Med Chir 1979; cephalus. J Pediatr 1986;109:1057–60. 19:173–9. 24 Torres E, Alba D, Frank A, et al. Iatrogenic meningitis due 3 Mayhall CG, Archer NH, Lamb VA, et al. Ventriculostomy- to Streptococcus salivarius following a spinal tap [letter]. related infections. A prospective epidemiologic study. N Clin Infect Dis 1993;17:525–6. Engl J Med 1984;310:553–9. 25 Torres-Morera LM, Caballero-S’anchez V, Alonso-Blanco 4 Raquet F, Mann WJ. Lumbar cerebrospinal fluid drainage F, et al. Septic meningitis following intradural anesthesia. for prevention of cerebrospinal fluid fistulas. HNO Rev Esp Anestesiol Reanim 1988;35:46–8. 1993;41:335–8. 26 Bissagnene E, Kouassi YF, Ba ZV, et al. Streptococcus 5 Huang CI, Huang MC, Chen IH, et al. Diverse applications pneumoniae meningitis after myelography and intrathecal of continuous lumbar drainage of cerebrospinal fluid in infiltration. 1989;82:572–4. neurosurgical patients. 1993; Bull Soc Pathol Exot Filiales Ann Acad Med Singapore 27 Morparia HK, Vontivillu J. Case report: cerebrospinal fluid 22(suppl 3):456–8. 6 Kusske JA, Turner PT, Ojemann GA, et al. Ventriculostomy fistula. A rare complication of myelography. Clin Radiol for the treatment of acute hydrocephalus following 1991;44:205. subarachnoid hemorrhage. J Neurosurg 1973;38:591–5. 28 Dimitriu SM, Luca V, Turcu T, et al. Acute nosocomial bac- 7 Lundberg N, Troupp H, Lorin H. Continuous recording of terial meningitis. Apropos 27 cases. Rev Med Chir Soc Med the ventricular-fluid pressure in patient with severe acute Nat Iasi 1992;96:19–25. traumatic brain injury. A preliminary report. J Neurosurg 29 Eng RH, Seligman SJ. Lumbar puncture-induced meningi- 1965;22:581–90. tis. JAMA 1981;245:1456–9. http://jnnp.bmj.com/ on September 26, 2021 by guest. Protected copyright.