2003 advances in icp monitoring techniques

842019 2003 Advances in ICP Monitoring Techniques

httpslidepdfcomreaderfull2003-advances-in-icp-monitoring-techniques 112

Advances in ICP monitoring techniques

Jun Zhong Manuel Dujovny Hun K Park Eimir Perez Alfred R Perlinand Fernando G Diaz

Biomechanics Laboratory Department of Neurological Surgery Wayne State University School of M edicine Detroit MI USA

With the advent of newer devices for measuring intracranial pressure (ICP) and cerebral metabolism more alternativescontinue to rise aiming to control ICP This manuscript presents a proposed analysis of different ICP monitoring devices in order to make appropriate selection of them in our clinical setting including general and pediatric applications A systematic review of the literature was made analyzing the technical advances in ICP monitoring The recent in vitro and in vivo tests as well as mathematicalcomputer models were reviewed Practical applications of principles were discussed and compared based on the mode of pressure transformationA ventricular catheter connected to an external strain gauge transducer or catheter tip pressure transducer device is considered to be the most accurate method of monitoring ICP and enables therapeutic CSF drainage The signicant infections or hemorrhage associated with ICP devices causing patient morbidity are clinicallyrare and should not deter the decision to monitor ICP Parenchymal catheter tip pressure transducer devices are advantageous when ventricular ICP cannot be obtained or if there is an

obstruction in the uid couple though they have the potential for signicant measurement differences and drift due to the inability to recalibrate Subarachnoid or subdural uid-coupled devices and epidural ICP devices are currently less accurate With an increasing miniaturization of the transducers beroptic systems have been developed however there is a problem of measurement accuracy during the period of patient monitoring and external calibration should be performed frequently to ensure constant accuracyVentriculostomies continue to have a pivotal role in ICP control With a rational understanding of the applications and limitations of the different ICP monitoring devices the outcome for critically ill neurological patients is optimized [Neurol Res 2003 25 339ndash350]

Keywords Intracranial pressure monitoring sensor beroptic strain gauge pneumatic

INTRODUCTION

During the middle of the 1990s the measurement of intracranial pressure (ICP) was done indirectly by meansof lumbar cerebrospinal uid (CSF) pressure monitor-ing12 which was pioneered by Guillaume and Janny in19513 The advent of routine direct monitoring of ICPbegan in 19604 and is credited to Lundberg5 whodescribed the method of directly monitoring the uidpressure in the ventricle

The next milestone in ICP monitoring came in 1973with the introduction of the subarachnoid screw (bolt)6which led to the development of other methods of monitoring ICP including the subdural cup catheter7extradural monitors8 and the beroptic catheter Since

then due to the development of catheter-tip pressuretransducers parenchymal monitoring has also beenutilized9 Each type of device depending on theintracranial location and method of pressure transduc-tion has its advantages and disadvantages An optimaldevice would meet various specic requirementsaccuracy of absolute measurements (tolerance) constantvalues over timely measurements (drift) low depen-dency from previous or consecutive measurements(hysteresis) accuracy of repeated measurements (valid-

ity) and accuracy of the absolute value dependent on

the magnitude of the value (linearity)

INTRACRANIAL LOCATION OF THE ICP SENSORCurrent ICP devices can be placed in either epiduralsubdural subarachnoid parenchyma or ventricularlocations

IntraventricularThese are essentially open-ended conduits placed

directly into the ventricular CSF5 This method consist-ing of a uid-coupled device is considered the goldstandard for measuring ICP and to which all new

methods must be compared for reference Using a three-way stopcock one of the extracranial ends can be usedto intermittently drain CSF and the other can beattached to a pressure transducer via saline-lled tubingfor continuous ICP recording The catheter may also beused to instill selected medications into the CSF such asthrombolytic agents in cases of intraventricular hemor-rhages or clotting at the proximal catheter or the use of antibiotics in cases of ventriculitis However a ventri-culostomy has the disadvantage that it penetrates themeninges and brain which gives the risk of bacterialtransmission through the uid coupling The ventricu-lostomy catheter can be difcult to place when there is

compression or shift of the ventricles In such cases theICP waveform may be dampened and the values

Correspondence and reprint requests to Jun Zhong MD PhD 550 ECaneld Rm 38 Detroit MI 48201 USA

[Jzhongneurosurgerywayneedu]Accepted for publication February2003

Neurological Research 2003 Volume 25 J une 339 2003 Forefront Publishing Group0161ndash641203040339ndash12



recorded by the catheter may be low due to artifacteffect Fluid leaks may be located at the port of entry of the catheter in the skin through a stopcock or due to achink in the catheter resulting in erroneous and low ICPvalues Air bubbles blood clots and brain or otherdebris can all interfere with the pressure wave conduc-

tion from the ventricles to the external transducerresulting in incorrect ICP readings The transducerposition needs to be readjusted each time the level of the patientrsquos head is changed in order to ensure reliablepressure values Draining CSF and measuring ICP at thesame time has inaccurately shown low ICP measure-ments not reecting the actual ICP This functionrequires the two measurements to be performedseparately The potential risks of misplacement infec-tion hemorrhage and obstruction have led to ndingalternative intracranial sites and devices for ICP moni-toring We need to highlight that ventriculostomycontinues to be the gold standard modality and the

initial consideration in most circumstances Howeverother ICP measurement modalities such as intra-parenchymal transducers are gaining relative priority inselected cases They are especially helpful in thosesituations where ventricles are small or slit withsignicant brain swelling

SubarachnoidThese uid-coupled systems connect the intracranial

compartment to an external transducer via tubing Thesubarachnoid bolt is a hollow screw that goes into theskull abutting the dura The dural membrane is

perforated allowing CSF to ll the bolt and theirpressures to become equal Then the closed uid tubing

transmits the pressure in this space710 Even thoughinfection and hemorrhage risks are low these devicesare quite prone to errors including ICP underestimationmisplacement of the screw and occlusion by debris11

Epidural

ICP monitoring from the epidural space is an attractiveconcept as the dura does not have to be opened theplacement of the monitoring device is easy and theremay be a low incidence of serious infection andhemorrhages However these devices are prone tomalfunction misplacement and baseline drift after afew days of continuous usage The inaccuracy obtainedis the result of the relatively inelastic dura that transmitsthe pressure from the CSF to the sensor12 rather thaninaccuracy of the equipment

Intraparenchymal

There are currently two types ber-optic and wireThe rst one is made of thin ber optic cables with apressure transducer at the tip that requires a dedicatedmicroprocessor to interpret the signal which is alsoexpensive nonexible and liable to breakage The wiresystem contains a micro-transducer at the tip of a exiblewire This system is interfaced to be used with standardbedside monitors In both types the tip is introducedinto the parenchyma through a 4 mm hollow screwinserted into the skull13 Infection and hemorrhage ratesare quite low14 but these devices do not allow for CSFdrainage Accuracy is reliable and optimal second onlyafter intraventricular catheterization15 and the device

needs to be calibrated only once before insertionregardless of the head height level used

Table 1 Features of monitoring devices

Device Advantages Pitfalls

Intraventricularcatheter Gold standard of accuracyAllows drainage and sampling of CSFAllows ICP controlInexpensive

Most invasiveSometimes difcult to cannulate ventricleCatheter can be occluded by blood or tissueNeeds reposting of transducer level with c hange inhead p ositionPotential infection

Subarachnoid boltscrew Quickly and easily placedDoes not invade brainAllows sampling of CSFMay have lower infection rate

Blocked by swollen brainCatheter can be occluded by tissue or bloodMust be balanced and recalibrated frequently

Subdural epidural cathetersensor Least invasiveEasily and quickly placed

Increasing baseline drift over time accuracy andreliability are questionableDoes not provide CSF sampling

Fiberopitc probeCatheter tipstrain gauge

Can be placed in the subduralsubarachnoid intraventricular or intraparenchymalspacesEasily transportedMinimal artifact and driftHigh resolution of waveform

No irrigation less risk of infectionNo need to adjust for patient position

Cannot be recalibrated after it is placed unless aventriculostomy is used simultaneously for referenceBreakage of the beroptic cableHigh Cost

CSF cerebrospinal uid ICP intracranial pressure

ICP MonitoringJun Zhong et al

340 Neurological Research 2003 Volume 25 June



ICP monitored by means of parenchymal catheter-tipstrain gauge1617 or subdural catheter uid-coupleddevice18 has been reported to be similar to ventricularICP While parenchymal or subdural monitoring loca-tions have been supported in the literature to haveaccurate monitoring features uid-coupled epiduraldevices subarachnoid bolts and pneumatic epiduraldevices have been reported to have signicantly

different ICP readings as compared to ventricularICP1920 However other investigators21 have found thatsubdural and parenchymal beroptic catheter-tip pres-sure value does not always correlate well with ventri-cular ICP

SENSORSThe rst ICP microtransducers were introduced in the1980s Other than early laboratory prototypes the rstcommercial design was the Honeywell MicrotransducerCatheter MTC-P5F It was initially designed for intra-parenchymal pressure measurement13 The rst micro-

transducer to be widely used was the beroptic CaminoOLM ICP Monitor (Camino Laboratories San Diego CAUSA) It required a dedicated microprocessor-drivenamplier to provide a numerical value andor contin-uous waveform display However the high price of thesystem and its disposable features prevented manyneurosurgical centers from using this microtransducer-based ICP monitor

Strain gauge catheter tipThis transducer tip contains a silicon microchip with

diffuse resistant strain gauges The diaphragm is coatedwith a thin layer of silicone rubber to insulate thetransducer from the environment into which it is placedThe strain gauges are connected via tiny wires thatextend through the length of the exible nylon tube tocomplete a wheatstone bridge type circuit located in theconnector housing

Studies have found the Codman MicroSensor to beaccurate with an average difference of iexcl05 toDagger 26 mmHg between the MicroSensor reading and anintraventricular uid coupled external strain gaugetransducer It is stable with a daily drift that variesbetween iexcl013 to Dagger 011 mmHg per day The average

zero drift of the MicroSensor measured at air pressurewas 02sect 05 mmHg for the average duration of 38Dagger 16 days It has a high delity with a frequencyresponse greater than 10000 Hz which is achieved bythe elastic properties of silicon small overall size of thediaphragm and tiny volumetric displacement underpressure When this is incorporated into a ventricularcatheter the system allows simultaneous drainage of CSF and ICP recording It is exible and can betunneled beneath the scalp preventing it from beingeasily broken Its small size (a nominal outer diameter of 07 mm for the nylon vent tube and 12 mm for thetransducer case) is an additional advantage particularly

for pediatric patients The absence of a uid columnprecludes dampening by blood clots debris or air

bubbles Since irrigation is not needed there is a low riskof infection22

Fiberoptic catheter tipFiberoptic transducers can be easily placed through

the working channel of the endoscope and haveinterposed assembly23 It is characterized as having very

low zero drift over long periods of time very goodfrequency response and stable linearity24Crutcheld et al reported that the device had an

accuracy of sect 3 mmHg over a 0 to 30 mmHg range invitro The maximum daily drift was sect 25 mmHg withan average daily drift of sect 06 mmHg and the drift rateover a 5-day period was sect 21 mmHg In vivo thepressures and waveform characteristics were verysimilar between the beroptic device and a strain-gaugetransducer connected to a ventriculostomy25 A veryimportant advantage is that a pressure-sensitive elementis placed in the head so the risk of infection isminimized This gives it an advantage over prolonged

ICP monitoring which uses a uid-lled catheter andexternal transducer26 They allow for continuous record-ing and monitoring of ICP in different brain compart-ments and give accurate pressure readings makingpossible the analysis of waveforms in the compartmentwhere the probe is placed

Disadvantages commonly associated with externaltransducers such as the inuence of transducer dis-placement relative to the patientrsquos head measurementartifacts and damping of the high frequency compo-nents of the pressure waveform caused by the limiteddynamic response of uid-lled tubing are alsoimproved by the microtransducers

At present a new in vivo pressure sensor (FISOTechnologies Inc Quebec Canada) is available Thistransducer combines its well-recognized white lightinterferometric interrogation technique with provensilicon micromachined technology It delivers theadvantages of both tip catheter and external invasivepressure transducers Due to the optical nature of thetransducer and transmission line the system delivershigh delity pressure measurements even in the pre-sence of strong EM or RF elds or interference associatedwith those generated by electric scalpel MRI systems(httpwwwsocompage_fopmemhtm)

Implantable devicesFor long-term usage an ICP monitoring-implantable

system with no external connections has obviousadvantages These devices would still act as foreignbodies but the elimination of skin penetration wouldminimize the risk of infection The principal problemsassociated with these types of systems are difcult to berecalibrated after implantation increased drift requirescomplicated and expensive hardware and shows slowdynamic responses The Codman ICP Tele-Sensor is animplantable capacitance detector that is designed to be

placed inline with chronically implanted ventricularshunt systems27 Chapman et al 28 used a method of

ICP Monitoring Jun Zhong et al

Neurological Research 2003 Volume 25 J une 341



A

B

Figure 1 A Spiegelberg (Aesculap GmbH amp Co Center Valley PA USA) KG Pneumatic ICP monitor B The sensor of Aesculap-SpiegelbergICP monitor The Air-Pouch is an air-lled bag mounteda t the tip of a catheter ICP is transmittedthrough the air in the pouch and the tube to the Brain-Pressure Monitor





pressure monitoring in association with a ventricularcatheter and subcutaneous reservoir This has beenfound useful in eight patients without the disadvantagesinherent in other methods of management

Table 1 shows the advantages and disadvantage of different ICP monitors Figures 1ndash3 demonstrate threekinds of ICP monitors and their sensors

CEREBRAL PERFUSION PRESSURE CONCEPTS ANDMANAGEMENTStudies made by Rosner e t al 29 about cerebralperfusion pressure (CPP) ICP and systolic arterial blood

pressure (SABP) analyzed in response to differentvariables have shown that these parameters arecontrolled by autoregulatory mechanisms and positionalchanges Patients in whom CPP is greater or equal to 70torr respond relatively poorly to mannitol with small ICPdecline and those with CPP less than 70 torr respondwith adequate or more signicant decline in ICP Therationale for these results are based on the hypothesisbased on the association of low CPP with autoregulatoryvasodilation whereas high CPP is associated withvasoconstriction If mannitol should work by a vasocon-striction mechanism the ICP effects should be mostapparent under conditions of low CPP30 Head elevationhas mainly benecial effects over ICP with a grossreduction of 1 mmHg in ICP and a decrease of 2ndash3 mmHg in CPP for every 10 degrees of head elevationThus demonstrating that the CPP was not beneciallyaffected by elevation of the head off the bed

HARMONIC ANALYSIS OF INTRACRANIAL PRESSUREWAVEFORMSThe techniques of Fourier analysis have been utilized tocharacterize the wave components of the complex time-A

B

Figure 2 A Ventrix Fiberoptic ICP monitor (Model NL950-100 Camino NeuroCare Group San Diego CA USA)B The sensor of Ventrix ICP monitor





Table 2 Accuracy and stability of ICP monitoring devices

Year Author Description of Study Conclusion

1977 Salmon The applanation transducer was used to measure intracranialpressure (ICP) through the intact fontanel

The correlation coefcient with direct measurements of ICP was098

1979 Dietrich K The transducer was implanted automaticallycoplanar to the duravia a burr hole of 4 mm diameter in the conscious patient withoutspecial surgical equipment

At a full scale range of 100 mm of Hg the zero-point drift is less than1 mm of Hg24 h

1982 Weaver Comparison of ICP measurements betweentwo subrarachnod uid-coupled pressure transducers in the same patients of 20

More than 50 of patients demonstrated signicant differences inICP Patients harboring intracranial mass lesions showing cleardifferences

1983 Mendelow Simultaneous recordings of ICP using two types of subdural uid-coupled bolt devices and a ventricular catheter uid-coupledsystem in 31 patients

ICP recordings were within 10 mmHg of ventricularICO in 41 of the recordings using one type of bolt and 58 using the other kind

1984 Dearden Assessment of ICP measurement accuracy in a subarachnoid subdural uid-coupled bolt device using an infusion test in 18patients

Device read ICP accurately according to infusion test 48 of thetime

1985 Barlow Simultaneous recording of ventricular uid-coupled ICP comparedto a subdural uid-coupled catheter in 10 patients and a subduralcatheter tip pressure transducer device in another 10 patients

Compared to ventricular ICP 44 of the subdural uid coupleddevice measurementsand 72 of the subdural catheter tip pressuretransducer devices were within a 10 mmHg range

1985 Powell Simultaneous recording of ICP using an epidural pneumaticpressure transducer and ventricular uid-coupled catheter in 17patients

Marked differences in pressure up to 30 mmHg were recorded

1987 Ostrup Comparison of ICP readings between a parechymal beropticcatheter tip pressure transducer device and ventricular uid-coupled catheterof subarachnoid bolt in 15 adults and ve children

Measure drift up to 1 mmHg per day Parenchymal ICP readingswere generally within 2ndash5 mmHg of ventricular or aubarachoid ICPmeasurements

1987 Piek Simultaneousrecording of ICP using a parenchymalstrain gaugecatheter tip and a ventricular uid-coupled catheter in seven

patients

Parenchymal ICP was 4ndash12 mmHg lower than ventricular ICP butparallel changes in pressure were noted

1988 Mollman Simultaneous recordings of ICP with a subduralsubarachniduid-coupled and a ventricular uid-coupled catheter in 31 patients

The difference between the ICP readings was 012 mmHg with astandard deviation of 529 mmHg

1990 Chambers ICP recordings between a ventricular uid-coupled system in 10patients compared to a subdural beroptic catheter tip pressuretransducer and the same device situated in the ventricular catheterin another 10 patients

54 and 74 of the beroptic subdural and beroptic ventricularICP readings respectively were with 5 m mHg of the ventricularuid-coupled ICP measurements

1990 Piek In a seriesof 100 patients13 had simultaneousICP recordingsfroma parenchymal strain gauge catheter tip pressure transducer deviceand a ventricular iud-coupled catheter

An initial drift up to 4 mmHg in the rst day Parenchymal ICPmeasurements were generally 4ndash8 mmHg below ventricular ICP

1992 Artru A prospective study of parenchymal beroptic catheter tip ICPmonitors in 100 patients

Daily baseline drift of 03 mmHg

1992 Gambardella Comparison of a parenchymal beroptic catheter tip pressuredevice to ventricular uid-coupled ICP reading in 18 adult patients

55 of parenchymal beroptic ICP readings were 5 mmHg higheror lower than ventricular ICP measurements

1992 Schickner Comparison of ICP readings between a parenchymalberopticcatheter tip pressure transducer and ventricular uid-coupledcatheter in 10 patients

66 of beroptic measurements exceeded ventricular ICP and 21were lower Absolute pressure differences of up to 40 mmHg wererecorded

1992 Schwartz Comparison of ICP readings between an epidural pneumaticpressure device and a subdural strain gauge subdural beroptic orventricular uid-coupled catheter in six patients

ICP readings from the epidural device correlated with the otherdevice readings in only one case

1993 Chambers Simultaneous recording of ventricularuid-coupled ICP comparedto a beroptic catheter tip pressure at the tip of the ve ntricularcatheter in 10 patients

60 of the ICP readings with the beroptic device were within2 mmHg of the ventricular uid-coupled ICP readings

1993 Czech Comparison of simultaneous ICP recordings in 15 patients using aventricular uid-coupled ICP monitoring system and a epiduralpneumatic ICP monitoring device

In the majority of comparisons the epidural device ICPmeasurements were different from ventricular ICP recordings withdeviations between 20 and Dagger 12 mmHg

1993 Gambardella Comparison of parenchymalberoptic catheter tip pressuretransduction device to ventricular uid-coupled readings in 12pediatric patients

ICP values obtained by the parenchymal beroptic device were3sect2 mmHg lower than ventricular ICP readings

1993 Gopinath Six microsensortransducerswere monitored for drift at pressures of 10 mmHg and 20 mmHg

The maximal drift of any of the transducers was 1 mmHg over ninedays

1995 Gopinath Evaluationof the measurementaccuracy and drift of a new cathetertip straingauge ICP deviceThe device was placed in the lumen of aventricular catheter in 25 patients

No signicant measurementdrift was noted over an average of fourdays The device was 63 accurate (within 2 mmHg) compared toventricular ICP recordings

1999 Morgalla Seven transducers (HanniSet Camino Codman SpiegelbergMedex Epidyn and Gaeltec) were compared

Measurementaccuracy was best with HanniSet probes Spiegelberghad slightlylarger deviationsThe highesterror was noted in Epidynand Gaeltec The largest drifts were seen with Medex andSpiegelbergThe highest long-term drifts were f ound with Epidynand Gaeltec Drift did not exhibit a linear pattern

1999 Morgalla Tests were performed at six different pressure levels between0 mmHg and 50 mmHg with temperature varied between 20deg and45deg C

Very low drifts of less than 015 mmHg deg C for Codman Epidyn andNeurovent Gaeltec and Camino exhibited higher drifts of 018 mmHg and 02 mmHg deg C respectively

2000 Yau Comparison of Spiegelberg intraventricularICPintracranialcompliance monitoring device in ve sheep

The S piegelberg air-pouch ICPcompliance monitor provides ICPand compliance data that a re very similar to those obtained usingboth gold-standard methods and an intraparenchymal ICP monitor

2001 Morgalla Over a period of two years 40 ICP transducers were implanted in35 patients by one surgeon

From this study a malfunction was detected in two devices bytesting them in a water bath before insertionReadings for the probeby means of measurement in an open water bath just beforeinsertion are strongly recommended

2002 Vassilyadi Monitoring patientsrsquo ICP during endoscopic surgery The small size of the Microsensor ICP monitor enabled easyplacement through the working channel of the endoscope andinterposed assembly

Modied and updated from Guidelines for the Managements of Severe Head Injury brain Trauma Foundation 1995)





varying periodic waveforms of the human body Theperiodicity and linearity parameters from the signal canpropose the quality of the frequency response among thetransducers in the current clinical market The harmonicanalysis can dissect a complex waveform into an innitesummation of sine and cosine functions of properfrequency units

Neurophysiologists have employed Fourier analysis in

order to quantify the pressure and frequency response of the transducers used in ICP measurement The advan-tage of the technique lies in the quantitative representa-tion of pressure wave from the intracranial space and theeasiness to compare the frequency component of thewaveform

Czosnyka et al 31 measured frequency response fromthe ICP monitoring devices that are currently used in theclinical setting The relative changes in the readings of time versus frequency domain of sine waves werereported The range of frequency bandwidth varies from20 to 30 Hz It indicated that there is a difference inamplitude measurement by the different transducers

The report pointed out that there is a damping effect bythe inlet tubing and it made a frequency difference of above 7 Hz During the measurement the pressurechamber has two different static pressure levels 10 and50 mmHg The study compared the ratio of thefundamental harmonic amplitude among the trans-ducers and dened good frequency properties withinthe range of 0 to 50 Hz The dynamic property of testtransducer showed dampening of the pressure waveformstarted at a frequency above 10 Hz and continued up to50 Hz However there was no statistical difference inthe frequency characteristics of transducers at the twostatic pressures (10 and 50 mmHg) A

B

Figure 3 A Strain Gauge Monitor(ICP express TM Codman Johnson amp Johnson ProfessionalRaynhamMA USA) Thiskit consists of a 38 cm catheter with an integrated stylet and ICP sensor a catheter anchoring clip and a 7-gaugetunneling trocar B The sensor of Codman ICP monitor





COMPLICATIONSComplications of ICP devices include infection hemor-rhage malfunction and obstruction

Factors associated with ICP monitoring related com-plications include age (gt44 years old) prolongedmonitoring time (mean of 14 days or more) use of steroids prolonged hospitalization time and use inpatients with moribund conditions32

Colonization of ICP devices increases signicantlyafter ve days of implantation3334 Irrigation of uid-coupled ICP devices signicantly increases bacterialcolonization35 Excluding the higher rate in summariz-ing the range of infection for uid-coupled devices theaverage rate of bacterial colonization in deferentintracranial locations is 5 for ventricular36 5 forsubarachnoid37 4 for subdural38 and 14 inparenchymally placed catheter-tip strain gauge orberoptic devices39 Even though these studies docu-mented increasing bacterial colonization of all ICPdevices over time clinically signicant intracranialinfections are uncommon Holloway et al 40 studied

584 retrospectivelyexamined patients with head injuriesthat required ventriculostomy The incidence of ventri-culitis was 104 and a steadily increasing dailyincidence of infection was observed over the rst 10days of monitoring The authors also found thatprophylactic replacement of catheters at 5-day intervalsdid not reduce the frequency of infection Commonlyincidences of ventricular catheter infections lie in the10ndash17 range41 Randomized prospective studies areneeded to denitively conclude that routine catheterreplacement does not reduce the risk of infection Inmost centers where ventriculostomies are frequentlyused important factors for controlling CSF contamina-tion appears to be related to periodical CSF sampling(usually every 2ndash3 days) through the catheter port usingan aseptic technique as well as making a tunnel underthe scalp to exit the catheter away from the burr holeRecommendations of where to perform a ventricu-lostomy in the intensive care unit and emergency roomversus the operating room aiming to decrease contami-nation rates are based upon individual hospital condi-tions of infectious disease control Those centers withacceptable low rates of nosocomial infections would bebeneted placing ventriculostomies at the bedside

The incidence of hematomas with all ICP devices is

14 Signicant hematomas requiring surgical evacua-tion occur in 05 of patients in published reports of more than 200 patients requiring ICP monitoring42 Theincidence of fatal hemorrhage depends on the sensortype A 5 incidence of fatal hemorrhage in subduraldevices 4 in intraparenchymal and 11 in ventricu-lostomies have been reported43

Malfunction or obstruction has been reported as63 16 and 105 in uid-coupled ventricularcatheters subarachnoid bolts or subdural cathetersrespectively With ICP measurements higher than50 mmHg rates of obstruction and loss of signal arenoted17

In a series of 536 indwelling cerebral monitoringdevices from Detroit Receiving Hospital ventricu-

lostomies were associated with a 729 infection rateand 328 hemorrhage whereas the use of beropticdevices was associated with 087 hemorrhage rate andno apparent infections44

IN VITRO TESTSMorgalla and co-workers tested seven types of ICPtransducers (HanniSet Camino Codman Spiegelberg

Medex Epidyn and Gaeltec) by increasing pressurelevels up to 80 mmHg Measurement accuracy was bestwith HanniSet probes The maximum errors with thistransducer were 3 mmHg Camino and Codman showedsimilar results Spiegelberg had slightly larger deviationsWith Epidyn and Gaeltec the highest error was notedup to 10 mmHg in the high-pressure range The 24-hdrift was lowest with HanniSet (02 mmHg) and Camino(08 mmHg) The largest drifts were seen with MedexSpiegelberg and Gaeltec (18 mmHg) Ten-day drift waslowest with HanniSet (01 mmHg dayiexcl1) and Codman(02 mmHg dayiexcl1) The highest long-term drifts were

found with Epidyn and Gaeltec (15 mmHg day

iexcl1

) Driftdid not exhibit a linear pattern45The authors also found very low drifts of less than

015 mmHg deg C for Codman Epidyn and NeuroventGaeltec and Camino exhibited higher drifts of 018 mmHg and 02 mmHg deg C respectively Withinthe temperature range of 35deg C to 42deg C all probes testedshowed insignicant drift by the temperature46

ANIMAL STUDIESYau et al 47 compared the Spiegelberg intraventricularcatheter a standard intraventricular catheter and Cod-man intraparenchymal ICP microsensor in ve sheepICP measured using the Spiegelberg intraventricular air-pouch balloon catheter displayed a linear correlationwith ICP measured using the standard intraventricularuid-lled catheter (r 2 ˆ 09846) and the Codmanintraparenchymal strain-gauge sensor (r 2 ˆ 09778)respectively Automated measurements of intraventricu-lar compliance obtained using the Spiegelberg com-pliance device were compared with compliancemeasurements that were made using the gold-standardmanual cerebrospinal uid bolus injection technique atICPrsquos range from 5 to 50 mmHg A linear correlation wasdemonstrated between the two methods (r 2 ˆ 07752

p lt

0001 average biasiexcl

0019 ml mmHgiexcl1

)Kroin et al 48 implanted intracranial pressure sensorsin the frontal white matter and placed a uid-lledcatheter in the cisterna magna (CM) of four dogs tomeasure CSF pressure This was done in order to verifywhether the ICP sensors matched CSF pressure changesThe animals were tested using standard physiologicalmaneuvers such as jugular vein compression headelevation and CSF withdrawal from and saline injectioninto the CM The mean ICP pressure and CM pressurewere compared for months to demonstrate that thetransducer system produced minimal drift over time Theresults showed that the change in the ICP sensor closely

duplicated that of the CSF waveform in the CM inresponse to the well-known physiological stimuli More





important mean ICP pressure remained within 3 mmHgof CM pressure for months with a mean difference lessthan 03 mmHg

PEDIATRIC RESEARCHAlthough a variety of monitoring techniques and devicesare available nowadays there have been few studiesmade in the pediatric population We present here some

recent reports Jensen et al 49 studied 98 consecutive head injury

children requiring beropticndashICP monitoring The moni-toring devices were placed in the frontal parenchymaand no prophylactic antibiotics were used All cathetertips were cultured upon removal The usual duration of ICP monitoring was 3ndash15 days with the exception of one patient who underwent monitoring for 40 days Nocomplications occurred during insertion of the ICPmonitors Catheter tip cultures were positive for Staphy-lococcus epidermidis in 7 of the children but nonedeveloped clinical features of CNS infection The

location of placement or duration of ICP monitoringdid not affect the rate of catheter tips with positivecultures There was a 13 mechanical failure rateassociated with the beroptic device

Pople et al 50 studied 303 children requiring ICPmonitoring He found an infection rate of 03intracranial hemorrhage in 03 (result of low plateletcounts) displacement of monitor in 1 and malfunc-tion of the system in 26

According to Gambardella and co-workerrsquos study51the beroptic device and the ventricular catheter havethe same accuracy and reliabilitywhen used in childrenThe beroptic method correlates very closely with the

ventriculostomy method but the pressure values werealways 3sect 2 mmHg lower than those obtained with theconventional pressure transducer system especially incritically ill patients

MATHEMATICALCOMPUTER MODELAlthough invasive methods to measure ICP are the mostaccurate attempts have been made for a noninvasiveassessment of the ICP Here are some recent studies

Schmidt et al 52 introduced a method to continuouslysimulate ICP waveform In a system analysis approachthe intracranial compartment was viewed as a black box

with arterial blood pressure (ABP) as an input signal andICP as an output A weight function was used totransform the ABP curve into the ICP curve The outputICP waveform was generated using a weight functionderived from the transcranial Doppler blood owvelocity (V) and ABP curves In order to establish therelationship between TCD characteristics and weightfunctions the velocity ABP and ICP curves of a denedgroup of patients were simultaneously recorded A linearfunction between the TCD characteristicsand the weightfunctions was obtained by calculating a series of multiple regression analyses Given examples demon-strate the procedurersquos capabilities in predicting the mean

ICP the pulse and respiratory waveform modulationsand the trends of ICP changes

According to Schmidt et al 53 the noninvasive ICPwas calculated from the ABP waveform by using a linearsignal transformation which was dynamically modiedby the relationship between ABP and cerebral bloodow velocity In all simulations parallel increases in realICP and noninvasive ICP were evident The simulatedresistance to outow of cerebrospinal uid (Rcsf) wascomputed using nICP and then compared with Rcsf

computed from real ICP The mean absolute errorbetween real and simulated Rcsf was 41sect 22 mmHgminml By the construction of simulations specic todifferent subtypes of hydrocephalus arising from variouscauses the mean error decreased to 27sect 17 mmHgminml whereas the correlation between real andsimulated Rcsf increased from R 073 to R 089(p lt0001)

Ursino and Lodi54 presented a simple mathematicalmodel of ICP dynamics oriented to clinical practice Itincluded the hemodynamics of the arterialndasharteriolarcerebrovascular bed CSF production and reabsorptionprocesses the nonlinear pressurendashvolume relationship

of the craniospinal compartment and a Starling resistormechanism for the cerebral veins Moreover arteriolesare controlled by cerebral autoregulation mechanismswhich are simulated by means of a time constant and asigmoidal static characteristic The model is used tosimulate interactions between ICP cerebral bloodvolume and autoregulation Three different relatedphenomena are analyzed the generation of plateauwaves the effect of acute arterial hypotension on ICPand the role of cerebral hemodynamics during pressure-volume index (PVI) tests Simulation results suggest thefollowing

1 ICP dynamics may become unstable in patientswith elevated CSF outow resis tance anddecreased intracranial compliance provided cere-bral autoregulation is efcient Instability manifestsitself with the occurrence of self-sustained plateauwaves

2 Moderate acute arterial hypotension may havecompletely different effects on ICP depending onthe value of model parameters If physiologicalcompensatory mechanisms (CSF circulation andintracranial storage capacity) are efcient acutehypotension has only negligible effects on ICP andCBF If these compensatory mechanisms are poor

even modest hypotension may induce a largetransient increase in ICP and a signicant transientreduction in CBF with risks of secondary braindamage

3 The ICP response to a bolus injection (PVI test) issharply affected via cerebral blood volumechanges by cerebral hemodynamics and autore-gulation

ACCURACY AND STABILITY (TABLE 2 )The American National Standard for Intracranial Pres-sure Monitoring Devices has been developed by the

Association for the Advancement of Medical Instrumen-tation in association with the Neurosurgery Commit-





tee55 According to their standards an ICP device shouldhave the following specications

1 Pressure range from 0 to 20 mmHg2 Accuracy of sect 2 mmHg within the range of 0 to

20 mmHg3 Maximum error of 10 in the range of 20 to

100 mmHg

External strain gauge transducers are coupled to thepatientrsquos intracranial space via uid-lled lines whichare accurate and can be recalibrated but obstruction of the uid couple can cause inaccuracy Moreover theexternal transducer must be consistently maintained at axed reference point relative to the patientrsquos head toavoid measurement error whereas catheter-tip trans-ducers are placed intracranially Catheter-tip straingauge or berotic devices are calibrated before intra-cranial insertion and cannot be recalibrated onceinserted Consequently if the device measurement driftsand is not recalibrated there is potential for aninaccurate measurement especially if the ICP transducer

is used for several daysAlthough nonuid-coupled systems are easier to use

as the ICP measurement is independent of headelevation there are reports about signicant ICPmeasurement drift with beroptic39 pressure transduc-tion and strain gauge17 pressure transduction in theparenchymal space Fiberoptic cables are subject tobreakage if the cable is bent excessively The accuracyof a pressure transduction device can be assessed byplacing the device within the lumen of a ventricularcatheter and comparing the uid-coupled pressurereading to the device being tested Some authorsreported that beroptic and strain gauge catheter-tipdevices tested in this manner have shown differencescompared to the ventricular ICP readings5657

According to the literature38 a premature cessation of ICP recording occurred in 55 of cases using asubdural catheter 203 of cases involving a screwand 71 of cases with a ventriculostomy Thispremature cessation was usually due to a blockedcatheter in the screw that could not be cleared byirrigation with saline A signicantly dampened traceoccurred in 16 of cases with a Richmond screw in27 of cases using a subdural catheter and in 25 of cases using the intraventricular catheter Comparing

these three methods catheter dislodgement was notstatistically signicantThe use of a balloon to measure ICP in the epidural

space was found to be unreliable as the measuredpressure varied with the volume used to ll the balloonand whether the balloon was lled or emptied58

According to Martinez-Manas et al 59 ventricu-lostomy catheters need to be calibrated every 8 hbecause they have a mean drift of 5 mmHg every 8 hwith a maximum of 11 mmHg Fiberoptic devices have amean daily drift of 06ndash2 mmHg so that a signicantcumulative error in ICP after 3ndash4 days monitoring can berecorded25 Some authors stated that their tendency is to

drift towards positive values so when the results of monitoring are wrong these errors tend to overestimate

ICP60 On the other hand Bavetta et al 61 found amedian value for zero drift of iexcl3 mmHg in their studySuch a clear negative bias in zero drift had notpreviously been noted

INDIRECT MEASUREMENTS

Tympanic membrane displacement

Another method that may be used to indirectly measureICP is based on the measurement of the tympanicmembrane displacement6 2 ndash 6 4 S ince the CSF andperilymph may communicate through the cochlearaqueduct an increase in ICP will cause an increase inpressure on the oval window This pressure will then betransmitted to the tympanic membrane via the (con-tracted) ossicles of the middle ear An impedanceaudiometer placed in the external auditory canal emittedand detected sound waves bounced off the tympanicmembrane to calculate the membranersquos displacement If ICP were to affect the displacement in a consistent waythe ICP could be inferred A reliable system has yet to beproven

Transcranial DopplerTranscranial Doppler (TCD) ultrasonography

measures the velocity of blood ow in the basalintracranial arteries and is most often used to detectvessel narrowing Investigators have noted that char-acteristic changes in the blood ow waveforms occurwith rising ICP6566 TCD may be helpful as an adjuvantdiagnostic tool to indirectly determine elevations in ICPas it has been demonstrated that diastolic ow velocitydecreases systolic peaks become sharper and pulsatility

index increases as ICP rises Detecting changes in theseTCD-variables may help to more closely monitorpatients with subarachnoid hemorrhage (SAH) andstroke among others

Transcranial ultrasound propagationBy using bitemporal acoustic probes an ultrasonic

wave is transmitted through the head It is assumed thatelevatedICP and changes in intracranial tissue elastancewill alter the velocity of a sound wave In a pilot study of 10 healthy subjects and 11 patients with various types of neurological pathology it was demonstrated that

changes in wave velocity did appear simultaneouslywith changes in ICP (four patients had intraventricularcatheters)67 The correlation of these changes with ICPwas not made and needs to be worked out in the future

Jugular bulb monitoring Jugular venous oxygen saturation (SjvO2) monitoring

may be of value in patients with elevated ICP Thismeasurement requires the retrograde insertion of anoximeter-tipped catheter into the jugular bulb althoughsome preliminary work shows that noninvasivemeasurement may be possible using near-infrared

spectroscopy68

As cerebral blood ow (CBF) decreasesSjvO2 will also decrease in a non-linear fashion SjvO2





will fall if elevation of ICP results in a critical reductionof cerebral perfusion pressure (CPP) and CBF becausecerebral oxygen extraction is increased Similarly if thecerebral metabolic rate of oxygen (CMRO2 ) increasesand CBF remains constant SjvO2 will fall Converselyan increase in SjvO2 may occur with ICP elevationsrelated to cerebral hyperemia69

Visual evoked responseVork et al 70 have demonstrated a good relationship

between ICP elevation and a shift in latency of the N2

wave of the visual evoked response (VER) The N2 waveis normally found at 70 msec and is thought to be acortical phenomenon It is therefore likely to be sensitiveto potentially reversible cerebral cortical insults such asischemia or increased intracranial pressure

CONCLUSIONA ventricular catheter connected to an external straingauge transducer or catheter tip pressure transducer

device is considered as the most accurate method of monitoring ICP and enables therapeutic CSF drainageThe signicant infections or hemorrhage associated withICP devices causing patient morbidity are clinically rareand should not deter the decision to monitor ICPParenchymal catheter tip pressure transducer devicesare advantageous when ventricular ICP cannot beobtained or if there is an obstruction in the uid couplethough they have the potential for signicant measure-ment differences and drift due to the inability torecalibrate Subarachnoid or subdural uid-coupleddevices and epidural ICP devices are currently lessaccurate With an increasing miniaturization of the

transducers beroptic systems have been developedHowever there is a problem of measurement accuracyduring the period of patient monitoring Externalcalibration of the measurement accuracy should beperformed frequently to ensure constant quality

ACKNOWLEDGEMENTSWe are grateful for the contribution from the Great OmentumFoundation LA the critical editorial review by Joonsoo Park the helpof the Linn McDonald Codman Co Judith Ross Aesculap Co IntegraCo

REFERENCES1 Queckenstedt H Zur Diagnose der Ruckenmarkkompression

Deutsche Aeitschrift fur Nervenheikunde 1916 55 325ndash3332 QuinckeH Die Diagnostiche und Therapeutische Bedeutung der

Lumbalpunktion Klinischer Vortrag Deutsch Med izinische Wochens chif tr 1905 31 1825ndash1828 1862ndash1972

3 Guillaume J Janny P Manometrie intracranienne continue Interetde la metholde et premieres resultants Rew Neurol 1951 84131ndash142

4 Gilliam EE Intracranial hypertension advances in intracranialpressure monitoring Crit Care Nurs Clin N Am 1990 2 21ndash27

5 Lundberd N Continuous recording and control of ventricular uidpressure in neurosurgecal practice Acta Psychitary Neurol Scand 1960 36 (Suppl 144) 11ndash19

6 Vries JK Becker DP Young HF A subarachobnid screw formonitoring intracranial pressure J Neurosurg 1973 39 416ndash419

7 Wilkinson HA The intracranial pressure-monitoring cup catheterTechnical note Neurosurgery 1977 1 139ndash141

8 Wald A Post K Ransohoff J et al A new technique for monitoringepidural ICP Med Instrum 1977 11 352ndash354

9 Guyot LL Dowling C Diaz FG Michael DB Cerebral monitoringdevices Analysis of complications Acta Neurochir Suppl (Wien) 1998 71 47ndash49

10 Vries JK Becker DP Yang HF A subarachnoid screw formonitoring intracranialp ressure J Neurosurg 1973 39 416ndash441

11 Miller JD Bobo H Kapp JP Inaccurate pressure readings fromsubarachnoid bolts Neurosurgery 1986 19 253ndash255

12 Ream Ak Silverberg GD Corbin SD et al Epidural measurement

of intracranial pressure Neurosurgery 1979 5 36ndash4313 Ostrup RC Luerssen TG MarshallLF et al Continuous monitoringof intracranial pressure with a miniaturized beroptic device J Neurosurg 1987 67 206ndash209

14 Levin A The use of a beroptic intracranial pressure monitor inclinical practice Neurosurgery 1977 1 266ndash271

15 Sundbarg G Nordstrom C-H Messeter K et al A comparison of intraparenchymatous and intraventricular pressure recording inclinical practice J Neurosurg 1987 67 841ndash845

16 Piek J Kosub B Kuch F et al A practical technique for continuousmonitoring of cerebral tissure pressure in neurosurgical patientsPreliminary results Acta Neurochir (Wien) 1987 87 144ndash149

17 Piek J Bock WJ Continuous monitoring of cerebral tissue pressurein neurosurgical practice Experiences with 100 patients Intensive Care Med 1990 16 184ndash188

18 Mollman HD Rockswold GL Ford SE A clinical comparison of

subarachnoidcatheters to ventriculostomyand subarachnoidboltsA prospective study J Neurosurg 1988 68 731ndash741

19 Czech T Korn A Reinprecht A et al Clinical evaluationof a newepidural pressure monitor Acta Neurochir (Wien) 1993 125169ndash172

20 Yau YH Piper IR Clutton RE Whittle IR Experimental evaluationof the Spiegelberg intracranial pressure and intracranial compli-ance monitor Technical note J Neurosurg 2000 93 1072ndash1077

21 Gambardella G Zaccone D Cardia E et al Intracranial pressuremonitoring in children Comparison of external ventricular devicewith the beroptic system Child Nerv Syst 1993 9 470ndash473

22 Gray WP Palmer JD Gill J et al A clinical study of parenchymaland subdural minature strain-gauge transducer for monitoringintracranial pressure Neurosurgery 1996 39 927ndash931

23 Vassilyadi M Ventureyra EC Neuroendoscopic intracranial

pressure monitoring Child Nerv Syst 2002 18 147ndash14824 Bray RS Chodroff NG Narayan RK et al A new beropticmonitoringdeviceDevelopmentof the ventricularbolt In Hoff JTBetz AL eds IntracranialPressure VII New York Springer-Verlag1989 pp 45ndash47

25 Crutcheld JS Narayan RK Robertson CS Michael LH Evaluationof a beropticintracranial pressure monitor J Neurosurg 1990 72482ndash487

26 Poca MA Sahuquillo J Arribas M Baguena M Amoros S Rubio EFiberoptic intraparenchymal brain pressure monitoring with theCamino V420 monitor Reections on our experience in 163severely head-injured patients J Neurotrauma2002 19 439ndash448

27 Chapman PH Cosman ER Arnold MA The relationship betweenventricular uid pressure and body positionin normal subjects andsubjects with shunt A telemetric study Neurosurgery 1990 36181ndash189

28 Chapman PH Cosman E Arnold M Telemetric ICP monitoringafter surgery for posterior fossa and third ventricular tumorsTechnical note J Neurosurg 1984 60 649ndash651

29 Rosner MJ Coley I Cerebral perfusion pressure intracranialpressure and head elevation J Neurosurg 1986 65 636ndash641

30 Rosner MJ Coley I Cerebral Perfusion Pressure A hemodynamicmechanism of mannitol and the postmannitol hemogram Neuro-surgery 1987 21 147ndash156

31 Czosnyka M Harland S Piechnik S et al Systematic over-estimation of intracranial pressure measured using a Caminopressure monitor J Neurol Neurosurg Psychiatry 1996 61427ndash428

32 Rosner MJ Becker DP ICP monitoring Complications andassociated factors Clin Neurosurg 1976 23 494ndash519

33 Paramore CG Turner DA Relative risks of ventriculostomy

infection and morbidity Acta Neurochir (Wien) 1994 127 79ndash8434 Bekar A Goren S Korfali E Aksoy K Boyaci S Complications of





brain tissue pressure monitoring with a beroptic deviceNeurosurg Rev 1998 21 254ndash259

35 Mayall CG Archer NH Lamb VA et al Ventriculostomy-relatedinfectionsA prospective epidemiologic study N Engl J Med 1984310 553ndash559

36 Wineld JA Rosenthal P Kanter RK et al Duration of intracranialpressure monitoring does not predict daily risk of infectionscomplications Neurosurgery 1993 33 424ndash430

37 Chambers IR Mendelow AD Sinar EF et al A clinical evaluationof the camino subdural screw and ventricular monitoring kits

Neurosurgery 1990 26 421ndash42338 North B Reilly P Comparison among three mthods of intracranialpressure recording Neurosurgery 1986 18 730ndash732

39 Artru F Terrier A Gibert I et al Monitoringof intracranialpressurewith intraparenchymalberoptic transducer Technicalaspects andclinical reliability Ann Fr Anesth Reanim 1992 11 424ndash429

40 Holloway KL Barnes T Sung C et al Ventriculostomy infectionsThe effect of monitoring duration and catheter exchange in 584patients J Neurosurg 1996 85 419ndash424

41 Vinas FC Bedside invasive monitoring techniques in severe brain-injured patients Neurol Res 2001 23 157ndash166

42 Paramore CG Turner DA Relative risks of ventriculostomyinfection and morbidity Acta Neurochir (Wien) 1994 127 79ndash84

43 Bley A Olafsson S Webster S et al Complications of intracranialpressure monitoring in fulminant hepatic failureLancet 1993 341157ndash158

44 Guyot L Diaz FG Dowling C Michael DB Cerebral monitoringdevices Analysis of complications Acta Neurochirurgica 199871 47ndash49

45 Morgalla MH Mettenleiter H Bitzer M Fretschner R Grote EHICP measurement control Laboratory test of 7 types of intracranialpressure transducers J Med Eng Technol 1999 23 144ndash151

46 Morgalla MH Mettenleiter H Katzenberger T ICP measurementaccuracy The effect of temperature drift Design of a laboratorytest for assessment of ICP transducers J Med Eng Technol 1999 23 10ndash14

47 Yau YH Piper IR Clutton RE Whittle IR Experimentalevaluationof the Spiegelberg intracranial pressure and intracranial compli-ance monitor Technical note J Neurosurg 2000 93 1072ndash1077

48 Kroin JS McCarthyRJ StylosL Miesel K IvankovichAD Penn RDLong-term testing of an intracranial pressure monitoring device

J Neurosurg 2000 93 852ndash85849 Jensen RL Hahn YS Ciro E Risk factors of intracranial pressuremonitoring in children with beroptic devices A critical reviewSurg Neurol 1997 47 16ndash22

50 Pople IK Muhlbauer MS Sanford RA Kirk E Results andcomplications of intracranial pressure monitoring in 303 childrenPediatr Neurosurg 1995 23 64ndash67

51 Gambardella G Zaccone C Cardia E Tomasello F Intracranialpressure monitoringi n childrenComparison of external ventriculardevice with the beroptic system Child Nerv Syst 1993 9470ndash473

52 Schmidt B Schwarze JJ Czosnyka M Sander D Wittich IKlingelhoferJ A method for a simulationof continuous intracranialpressure curves Comput Biomed Res 1998 31 231ndash243

53 Schmidt B Czosnyka M Schwarze JJ Sander D Gerstner WLumenta CB KlingelhoferJ Evaluationof a method for noninvasiveintracranial pressure assessment during infusion studies in patientswith hydrocephalus J Neurosurg 2000 92 793ndash800

54 Ursino M Lodi CA A simple mathematical model of theinteraction between intracranial pressure and cerebral hemo-dynamics J Appl Physiol 1997 82 1256ndash1269

55 Brown E Intracranial Pressure Monitoring Devices Arlington VAAssociation for the Advancementof Medical Instrumentation1988

56 Gopinath SP Robertson CS Contant CF et al Clinical evaluation

of a miniature strain-gauge transducer for monitoring intracraialpressure Neurosurgery 1995 36 1137ndash114157 Holzschuh M Woertgen C Metz C Brawanski A Clinical

evaluation of the InnerSpace breoptic intracranial pressuremonitoring device Brain Inj 1998 12 191ndash198

58 Cornoneous NJ McDowall DG Gibson RM et al Measurementof extradural pressure and its relationship to other intracranialpressures An experimentaland clinical study J Neurol Neurosurg Psychiatry 1973 36 514ndash522

59 Martinez-Manas RM Santamarta D de Campos JM Ferrer ECamino intracranial pressure monitor Prospective study of accuracy and complications J Neurol Neurosurg Psychiatry 2001 70 138

60 Czosnika M Czosnika Z Pickard JD Laboratory testing of threeintracranial pressure microtransducers Technical report Neuro-surgery 1996 38 219ndash224

61 Bavetta S Norris JS Wyatt M et al Prospective study of zero driftin beroptic pressure monitors used in clinical practice J Neurosurg 1997 86 927ndash930

62 Dennis LJ Mayer SA Diagnosis and management of increasedintracranial pressure Neurol India 200149 (Suppl 1)S37ndash50

63 Kast R A new method for noninvasive measurement of short-termcerebrospinal uid pressure changes in humans J Neurol 1985232 260ndash261

64 Reid A Marchbanks RJ Bateman DE et al Mean intracranialpressure monitoring by a non-invasive audiological technique Apilot study J Neurol Neurosurg Psychiatry 1989 52 610ndash612

65 Hassler W Steinmetz H Gawlowski J Transcranial dopplerultrasonography in raised intracranial pressure and in intracranialcirculatory arrest J Neurosurg 1988 68 745ndash751

66 Aaslid R Markwalder T Nornes H Noninvasive transcranialdoppler ultrasound recording of ow velocities in the basalcerebral arteries J Neurosurg 1982 57 769ndash774

67 Pranevicius O Bertasius K Pranevicius M et al Non-invasivedynamic assessment of the elasticity of intracranial structures ActaNeurol Scand 1992 86 512ndash516

68 McCormick P Stewart M Boetting M et al Noninvasive cerebraloptical spectroscopy for monitoring cerebral oxygen delivery andhemodynamics Crit Care Med 1991 19 89ndash97

69 RobertsonCS GopinpathSP GoodmanJC et al SjvO2 monitoringin head-injured patients J Neurotrauma 1995 12 891ndash896

70 York DH Legan M Bennner S Further studies with non-invasivemethod of intracranialpressure estimationNeurosurgery 1984 14456ndash460




recorded by the catheter may be low due to artifacteffect Fluid leaks may be located at the port of entry of the catheter in the skin through a stopcock or due to achink in the catheter resulting in erroneous and low ICPvalues Air bubbles blood clots and brain or otherdebris can all interfere with the pressure wave conduc-

tion from the ventricles to the external transducerresulting in incorrect ICP readings The transducerposition needs to be readjusted each time the level of the patientrsquos head is changed in order to ensure reliablepressure values Draining CSF and measuring ICP at thesame time has inaccurately shown low ICP measure-ments not reecting the actual ICP This functionrequires the two measurements to be performedseparately The potential risks of misplacement infec-tion hemorrhage and obstruction have led to ndingalternative intracranial sites and devices for ICP moni-toring We need to highlight that ventriculostomycontinues to be the gold standard modality and the

initial consideration in most circumstances Howeverother ICP measurement modalities such as intra-parenchymal transducers are gaining relative priority inselected cases They are especially helpful in thosesituations where ventricles are small or slit withsignicant brain swelling

SubarachnoidThese uid-coupled systems connect the intracranial

compartment to an external transducer via tubing Thesubarachnoid bolt is a hollow screw that goes into theskull abutting the dura The dural membrane is

perforated allowing CSF to ll the bolt and theirpressures to become equal Then the closed uid tubing

transmits the pressure in this space710 Even thoughinfection and hemorrhage risks are low these devicesare quite prone to errors including ICP underestimationmisplacement of the screw and occlusion by debris11

Epidural

ICP monitoring from the epidural space is an attractiveconcept as the dura does not have to be opened theplacement of the monitoring device is easy and theremay be a low incidence of serious infection andhemorrhages However these devices are prone tomalfunction misplacement and baseline drift after afew days of continuous usage The inaccuracy obtainedis the result of the relatively inelastic dura that transmitsthe pressure from the CSF to the sensor12 rather thaninaccuracy of the equipment

Intraparenchymal

There are currently two types ber-optic and wireThe rst one is made of thin ber optic cables with apressure transducer at the tip that requires a dedicatedmicroprocessor to interpret the signal which is alsoexpensive nonexible and liable to breakage The wiresystem contains a micro-transducer at the tip of a exiblewire This system is interfaced to be used with standardbedside monitors In both types the tip is introducedinto the parenchyma through a 4 mm hollow screwinserted into the skull13 Infection and hemorrhage ratesare quite low14 but these devices do not allow for CSFdrainage Accuracy is reliable and optimal second onlyafter intraventricular catheterization15 and the device

needs to be calibrated only once before insertionregardless of the head height level used

Table 1 Features of monitoring devices

Device Advantages Pitfalls

Intraventricularcatheter Gold standard of accuracyAllows drainage and sampling of CSFAllows ICP controlInexpensive

Most invasiveSometimes difcult to cannulate ventricleCatheter can be occluded by blood or tissueNeeds reposting of transducer level with c hange inhead p ositionPotential infection

Subarachnoid boltscrew Quickly and easily placedDoes not invade brainAllows sampling of CSFMay have lower infection rate

Blocked by swollen brainCatheter can be occluded by tissue or bloodMust be balanced and recalibrated frequently

Subdural epidural cathetersensor Least invasiveEasily and quickly placed

Increasing baseline drift over time accuracy andreliability are questionableDoes not provide CSF sampling

Fiberopitc probeCatheter tipstrain gauge

Can be placed in the subduralsubarachnoid intraventricular or intraparenchymalspacesEasily transportedMinimal artifact and driftHigh resolution of waveform

No irrigation less risk of infectionNo need to adjust for patient position

Cannot be recalibrated after it is placed unless aventriculostomy is used simultaneously for referenceBreakage of the beroptic cableHigh Cost

CSF cerebrospinal uid ICP intracranial pressure





























A

B











B

























patients













































B


















iexcl1




p lt


0019 ml mmHgiexcl1




































100 mmHg





















spectroscopy68























































































































A

B











B

























patients













































B


















iexcl1




p lt


0019 ml mmHgiexcl1




































100 mmHg





















spectroscopy68




































































































B

























patients













































B


















iexcl1




p lt


0019 ml mmHgiexcl1




































100 mmHg





















spectroscopy68


















































































































patients













































B


















iexcl1




p lt


0019 ml mmHgiexcl1




































100 mmHg





















spectroscopy68




































































































B


















iexcl1




p lt


0019 ml mmHgiexcl1




































100 mmHg





















spectroscopy68











































































































iexcl1




p lt


0019 ml mmHgiexcl1




































100 mmHg





















spectroscopy68




























































































































100 mmHg





















spectroscopy68





























































































2003 advances in icp monitoring techniques

Documents