ORIGINAL ARTICLE - ABSTRACT
I can hear my shunt—audible noises associated with CSF shunts
in hydrocephalic patients
Received: 10 January 2017 /Accepted: 30 March 2017
#Springer-Verlag Wien 2017
Background Cerebrospinal fluid (CSF) shunts are life-long
implants, and patients have reported anecdotally on noises
associated with their shunts. There is, however, a marked lack
of information regarding acoustic phenomena related to CSF
Methods We identified all patients who had been treated or
followed in our neurosurgical department within a 15-year
period from January 2000 up to the end of 2014. After approv-
al of the local ethics committee all patients who were cogni-
tively intact were explored by a questionnaire and by personal
interview about acoustic phenomena related to their shunts.
Results Three hundred forty-seven patients were eligible for
the survey, and 260 patients completed the questionnaire.
Twenty-nine patients (11.2%) reported on noises raised by
their shunts. All of them experienced short-lasting noises
while changing body posture, mainly from a horizontal to an
upright position, or while reclining the head. Most of the pa-
tients reported on soft sounds, but loud and even very loud
noises occurred in some patients. Seventy-six percent of the
patients were not bothered by these noises as they considered
it as a normal part of the therapy or as proof that the shunt
device was functioning. Modern valves with gravitational
units are prone to produce noises in young adults, but nearly
all valve types can evoke noises.
Conclusions Noises caused by a shunt do occur in a consid-
erable number of patients with shunts. One should be aware of
this phenomenon, and these patients must be taken seriously.
Keywords CSF shunt .Noise .Shunt valve
A considerable amount of literature on CSF shunts has been
published in the last decades [1,2]. There is hardly a complica-
tion of shunt treatment that has not already been reported [7,10].
During follow-up, neurologically intact patients have anec-
dotally reported acoustic phenomena related to their shunts.
These reports have regularly been neglected or have not been
taken seriously by the patients’neurosurgeons. Although audi-
ble noises after mechanical heart valve replacement are a well-
known and adequately studied phenomenon [3,6], there is a
vast lack of data concerning shunt-related noises. Only three
case reports have so far been published [4,8,9]. The incidence
and clinical presentation of these noises remain unknown.
We evaluated our patients with shunts to determine the
frequency and nature of CSF shunt-related noises and their
association with other factors.
We identified all patients with an implanted CSF shunt device
who were treated and/or followed up in a tertiary hospital in
Germany from January 2000 until December 2014
(Department of Neurosurgery, Klinikum Kassel and Kassel
School of Medicine, University of Southampton).
Department of Neurosurgery, Klinikum Kassel, Moencheberg Str.
41-43, 34125 Kassel, Germany
Kassel School of Medicine, Universitiy of Southampton,
Empiric Educational Science FB 01, University of Kassel,
Department of Neurology, Klinikum Kassel, Kassel, Germany
Inclusion criteria were the presence of a CSF shunt device
at the time of the survey, implantation of the shunt for at least
12 months, and age above 10 years.
Exclusion criteria were death, pediatric patients under the
age of 10 years, the presence of severe communication prob-
lems, such as significant mental impairment or severe aphasia,
and a documented non-functioning shunt device at the time of
All subsequent patient charts were reviewed. The follow-
ing parameters were recorded: age at the time of the survey,
age at the time of shunt insertion, origin of hydrocephalus,
valve type, pressure setting, and shunt type (ventriculoatrial,
ventriculoperitoneal, others). If an X-ray in lateral view or a
CT scout scan was available, the distance between the valve
and the internal auditory canal was measured and documented.
Patients were examined via a questionnaire to determine
the presence of ear noises, loudness on a scale from 0 to 10
(0, no sound; 1, can barely be heard; 10, a starting jet plane),
the nature of the noise, confounding factors of occurrence,
subjectively assumed source of the noise, and the presence
of ear-nose-throat diseases or amblyacousia.
A questionnaire was sent to all eligible patients by mail.
The questionnaire is shown in Table 1(translated from
All patients who reported on noises in the questionnaire
were interviewed during an outpatient visit or by telephone.
A shunt-related ear noise was assumed when the sound
occurred immediately or within 2 weeks after shunt insertion
and when the sound could be provoked or stopped by posi-
tioning, movements of the body or the head, or other measures
such as the Valsalva maneuver or if the sound was audible to
the examining physician.
Tinnitus was defined as a noise originating from one or
both ears that was not evocable by or susceptible to move-
ment, noise, or other external stimuli and the first appearance
of which seemed incoherent with shunt insertion. The diagno-
sis of tinnitus and amblyacousia was confirmed by an ENT
When patients reported on noises that could not be classi-
fied as either a shunt-related noise or tinnitus, the noise was
described as unknown origin, for example, permanent noises
that can be altered by body posture.
Our study was approved by the local ethics committee of
the University of Marburg, Germany.
The following parameters were used for statistical analysis:
age at the time of shunt insertion and of the survey, duration of
the shunt treatment, gender, origin of the hydrocephalus, the type
of shunt (VP, VA), the valve type and pressure level, the presence
of amblyacousia or other ENT diseases, the distance between the
valve and the auditory canal, and the intensity of the noise.
Patients who reported no sounds served as a control group.
Statistical analysis was done by chi-square test or Fisher
exact test and ANOVA. Factors found to be significant (p
0.05) in univariate analysis were included for mutivariate
anlaysis (age at the time of the survey, valve type, shunt type,
shunt duration). The logistical regression model was used for
multivariate analysis. Analysis was performed with SPSS
software, version 11.
Utilizing the administrative database, 611 patients were iden-
tified as having a CSF shunt device. One hundred thirteen
patients had already died. One hundred eleven patients could
not be questioned because of severe communication problems
due to neurological impairment. Twenty-seven patients were
children under the age of 10 years. Thirteen patients had a
non-functioning shunt device. Of the remaining 347 patients
260 (69%) returned a completed questionnaire (Table 2).
Seventy-nine patients reported ear noises, and 29 patients
(11.2%) reported noises clinically related to their shunts.
Fifty-one of them were interviewed during an outpatient visit.
Twenty-eight patients were interviewed by telephone. Thirty-
Tabl e 1 Questionnaire translated
from German Do you have an ear noise? Yes No
When did you notice it the first time?
Do you remember the noise before your shunt insertion?
How would you describe the noise?
Is this noise permanent or intermittent?
In which situations does the noise occur? Does it occur in
special body postures? If yes please specify
How loud would you estimate that noise?
10 = a starting jet plane
Minimum 1 Maximum 10
Do you have difficulties with hearing? Which ear? Do you
have a hearing aid?
Do you have other ear diseases?
six could be identified as having tinnitus unrelated to the
shunt. In 14 patients the source of the noise remained unclear.
Clinical data of these subgroups are shown in Table 3.
Characteristics of audible noises
Shunt-related noises were reported as occurring while chang-
ing body posture, especially from a horizontal to an upright
position (13 patients), or could be provoked with head
reclination or other distinct movements of the head (7 pa-
tients). All seven patients with noises that could be provoked
by head movements were implanted with gravitational units.
Noises occurring during transition to an upright body position
were mainly reported by patients with differential valve types.
The shunt-related noise always occurred intermittently, and
lasted for only a few seconds in the majority of the cases
(69%), or could be stopped voluntarily by a specific move-
ment of the head.
About 90% of the patients described shunt-related noises as
an intermittent hissing or chattering sound, whereas tinnitus
patients described their noises as a continuous whistling and/
or hissing synchronous with the heartbeat. One patient expe-
rienced two different noises: a permanent whistling noise orig-
inating from the inner ear, and another intermittent hissing
related to the shunt that could be provoked by standing up
The majority of patients had gentle sounds with a reported
severity of 1 to 3 on a sound intensity scale from0 to 10, while
three patients (10%) reported significant chattering (7–8ona
sound intensity scale from 0 to 10). The median intensity of
the experienced noises was similar when comparing tinnitus,
shunt-related noises, and noises with unknown origin (4.0, 5,
and 4, respectively, p=0.927).
More than half of all tinnitus patients (21 patients, 57%) felt
disturbed by their shunt-related noise, but only a fourth (7
patients, 24%) of the shunt-related noisesbothered the patients
involved. This perception was not related to the volume of
In only three patients could the noise be verified by the
examining physician using a stethoscope placed over the
valve when the patient was asked to provoke the noise
All 29 patients with shunt-related noises believed that
the shunt system was the source of the noise. Twenty-
two of them indicated that they had gotten used to the
noise after some time. Four patients definitely felt com-
fortable with the noise and believed that it indicated a
correctly functioning shunt. However, four tinnitus pa-
tients believed that the noise that they perceived origi-
nated from the shunt.
Most of the patients (23 out of 29) indicated that the shunt-
related noise first occurred immediately after shunt insertion
or after revision surgery with insertion of a replacement valve.
In four patients the noise disappeared in due course (weeks or
months after the shunt surgery) without any evidence of shunt
Correlation of clinical parameters
One hundred twenty-three men and 137 women returned the
questionnaire; the occurrence of shunt noises was not related
Tabl e 2 Flow chart acquisition
of patients 611 patients 113 died
111 unable to communicate
27 children under the age of 10
13 with non-functioning
347 patients for
Tabl e 3 Patient characteristics
according to the different
Shunt-related noises 29 44.87 40.23 5.59 26/3/0 5 58
Tinnitus 37 56.97 42.17 10.34 22/15/0 4.0 57
No noise 181 51.97 40.12 8.81 149/28/4 0 54
Unclear noise 14 46.06 38.79 8.25 8/6/0 4 50
The mean age at shunt insertion was 38.8 years and did not
differ significantly betweenthe subgroups. Patients experienc-
ing shunt-related noise were significantly younger at the time
of the survey than patients not experiencing noise or with
tinnitus (44.87 years vs. 51.97 years vs. 56.97 years, respec-
tively; OR 0.009; 95% CI 0.957–0.992; p= 0.004). These
results demonstrated a shorter duration of shunt treatment in
patients with shunt-related noise compared to tinnitus patients
or patients with no perceived noises at all (median 5.59 years,
10.34 years and 8.81 years, respectively, p= 0.001). This
ceased to be significant in multivariate analysis (p=0.24).
Two hundred two patients received a VP shunt, while 54
patients were implanted with an atrial shunt device and 4
patients had cysto-peritoneal shunts. Patients with tinnitus
had an atrial shunt device significantly more often
(p= 0.005), but this observation lost its significance in multi-
variate analysis simply because it is a measure of age. Tinnitus
patients were significantly older than the other subgroups and
received their shunts on an average of 4.4 years earlier. In the
past, atrial shunts were used frequently, and age is a well-
known risk factor for the development of tinnitus .
All patients had their shunt valves located near the cranium.
There were no lumbo-peritoneal shunts or valves positioned
thoracally in this cohort. Radiographic images of the valves
were available for measurements in 239 patients. The distance
between the valve and the auditory canal did not vary signif-
icantly among shunt-related noises, tinnitus, or patients with-
out noises (median 58 mm, 57 mm, and 54 mm, respectively,
The different causes of hydrocephalus according to the
occurrence of noises are shown in Table 4. The cause of hy-
drocephalus did not influence the occurrence of any noise
Distributions of the different valve types are shown in
Table 5. Tinnitus patients showed the same distribution of
valves as patients without perceived noise. In the group with
shunt-related noises, Pro-GAV was significantly overrepre-
sented. This turned out to be significant even in the multivar-
iate analysis (OR 2.152; 95% CI 2.189–33.814; p=0.003).
Other valve types (fixed and adjustable valves, differential
pressure valves, and gravitational units) did generate audible
noises, but the relative risk for developing an audible shunt-
related noise with an adjustable gravitational valve was 0.216,
whereas the relative risk with an adjustable differential pres-
sure valve or other valve types was 0.08 and 0.04, respective-
ly. For a graphical illustration, see Fig. 1. Interestingly, the
pressure level or the valve setting in adjustable valves did
not influence the occurrence of audible noises (p=0.87).
In three cases, the presence of amblyacousia did not
negatively correlate to the patients experiencing shunt-
related noises, but amblyacousia was highly correlated to tin-
nitus (p < 0.0001).
Shunt-related noises are extensively neglected in the present
literature and publications, and indaily neurosurgical practice,
even though this survey shows that about 11% of all evaluated
shunt patients reported experiencing shunt-related noises.
Most patients did not report on these noises without
prompting because they assumed that the noise was a normal
part of their therapy. Patients rarely felt disturbed by the noise.
Some patients even felt comfortable with the noise, assuming
that it indicated a properly functioning shunt system.
Young adults with adjustable gravitational valves were es-
pecially prone to shunt-related noises. Many elderly patients
with NPH were implanted with the same valve type in our
cohort, but geriatric patients rarely experienced shunt-related
noises. This could correlate to their reduced physical activity.
With advancing age, rapid changes in body posture or pro-
nounced movements of the head are often avoided, or even
not possible, potentially preventing patients from perceiving
shunt-related noises. But the presence of amblyacousia, which
is common in elderly patients, did not protect an individual
from perceiving shunt-related noises. The skull may serve as a
transmitter of noises originating from the valve to the inner
ear, similar to a hearing aid, which is used in disturbances of
The source of the noise in Pro-GAV valves has already
been investigated by Stockhammer et al. . They showed
that vibrations of the ball-in-cone mechanism create the noise
at a flow of more than 200 ml/h. This corresponds to the
phenomenon that many patients experienced the noise while
changing body posture, mainly from a prone to an upright
position. This is a moment when peak flows in the shunt
system are expected. All patients reporting on noises related
to their head position, such as reclination, had shunts with a
gravitational unit. With distinct reclination the gravitational
Tabl e 4 Distribution of the origin of hydrocephalus according to the
Tinnitus Unclear No
Posthemorrhagic 8 14 4 64 90
NPH 5 7 3 44 59
Tumor 6 3 0 24 33
Spina bifida 1 3 1 16 20
Post-traumatic 1 2 1 6 10
Post-infectious 1 1 0 5 7
Pseudotumor 3 1 1 1 6
Unknown origin 2 3 1 12 18
Others 2 2 3 10 17
All 29 36 14 182 260
unit is deactivated, so during reclination of the head in an
upright position peak flow occurs because of siphoning,
resulting in an audible noise .
Patients in this cohort with other valve types without a ball-
in-cone mechanism (Spitz-Holter valve, Orbis Sigma) did not
report on noises, although the number of these valve types
included in our study is too small to draw reliable conclusions.
Termination of a shunt-related noise has been used as an
indicator for real shunt dysfunction in a case report .
However, we observed that shunt-related noises infrequently
disappear some weeks or months after shunt insertion, without
evidence of shunt malfunction in certain patients. The cause of
this this phenomenon is unclear. Perhaps the development
over time of a slight protein coating of the ball-in-cone
Fig. 1 Scatterplot describing the
correlation among the relative risk
of experiencing a shunt-related
noise, the valve used, and the
patient’s age at the time of the
Tabl e 5 Valve distribution in the
different subgroups Va l v e t y p e S h u nt - r e l a t e d
Pro-GAV (Miethke) 20 12 62 2
GAV (Miethke) 3 3 18 2
Hakim Medos programmable valve (Codman) 6 13 50 5
Spitz-Holter valve 0 7 13 2
Orbis Sigma (Integra) 0 0 4 1
Hakim Precision medium-low (Codman) 0 0 6 2
PaediGAV (Miethke) 0 0 14 0
Mono-Step valve (Miethke) 0 0 3 0
Strata valve (Medtronic) 0 0 3 0
Certas (Codman) 0 0 1 0
Unknown 0 1 8 0
mechanism protects the valve from resonance-like vibrations
without compromising proper valve function.
Most patients reported low-volume noises that they did not
perceive as disturbing. Even patients with loud and very loud
shunt-related noises often did not feel bothered, as the bruit
generally lasts for some seconds, and patients felt safe, often
indicating that: BAs long as I can hear my shunt, I know
everything is working well.^
Although it is not a surgical complication, for all intents
and purposes audible noises produced by a shunt system have
the ability to alter daily life for some patients. The need to
complement the informed consent with shunt-related noises
is worthy of discussion.
The incidence of acoustic phenomena related to CSF shunt
devices seems to be higher than previously thought. They
are rarely reported on the patient’s own initiative. Modern
ball-in-cone valves with gravitational units in young adults
are especially prone to evoking audible sounds.
These noises are usually gentle and well tolerated by pa-
tients. After an initial period of confusion, the majority of
patients get used to their noise and take it as an indicator of
a functioning shunt system.
ANOVA, Analysis of variance; CI, confidence interval;
CSF, cerebrospinal fluid; CT, computer tomography; ENT,
ear, nose, and throat; NPH, normal pressure hydrocephalus;
OR, odds ratio; VA, ventriculoatrial; VP, ventriculoperitoneal.
Compliance with ethical standards
Funding No funding was received for this research.
Conflict of interest All authors certify that they have no affiliations
with or involvement in any organization or entity with any financial
interest (such as honoraria; educational grants; participation in speakers’
bureaus; membership, employment, consultancies, stock ownership, or
other equity interest; and expert testimony or patent-licensing arrange-
ments), or non-financial interest (such as personal or professional rela-
tionships, affiliations, knowledge or beliefs) in the subject matter or ma-
terials discussed in this manuscript.
Ethical approval All procedures performed in studies involving hu-
man participants were in accordance with the ethical standards of the
institutional and/or national research committee and with the 1964
Helsinki Declaration and its later amendments or comparable ethical
Informed consent Informed consent was obtained from all individual
participants included in the study.
1. Flannery AMP, Mitchell L (2014) Pediatric hydrocephalus: system-
atic literature review and evidence-based guidelines. Part 1-10. J
Neurosurg Pediatr 14(Suppl 1):3–81
2. Jensen F, Jensen FT (1979) Acquired hydrocephalus. I. A clinical
analysis of 160 patients studied for hydrocephalus. Acta Neurochir
3. Koertke H, Hoffmann-Koch A, Boethig D, Minami K, Breymann
T, El-Arousy M (2003) Does the noise of mechanical heart valve
prostheses affect quality of life as measured by the SF-36 question-
naire? Eur J Cardiothorac Surg 24(1):52–57
4. Kolenda H, Schoener W (1995) Shunt-valve noise as an unusual
reason for shunt revision. Childs Nerv Syst 11:489–491
5. Martines F, Sireci F, Cannizzaro E, Constanzo R, Martines E,
Mucia M, Plescia F, Salvago P (2015) Clinical observations und
risk factors for tinnitus in a Sicilian cohort. Eur Arch
6. Moritz A, Steinseifer U, Kobinia G, Neuwirth-Riedl K, Wolters H,
Reul H (1992) Closing sounds and related complaints after heart
valve replacement with St Jude Medical, Duromedics Edwards,
Björk-Shiley Monostrut, and Carbomedics prostheses. Br Heart J
7. Nakano A, Tani E, Sato M, Shimizu Y (1994) Cerebrospinal fluid
leakage from the nipple after ventriculoperitoneal shunt: case re-
port. Surg Neurol 42(3):224–226
8. Ordia JI, Mortara RW, Spatz EL (1987) Audible cerebrospinal fluid
flow through a ventriculoperitoneal shunt. Report of two cases. J
9. Stockhammer F, Miethke C, Knitter T, Rohde V, Sprung C (2014)
Flow-related noise in patients with ventriculoperitoneal shunt using
gravitational adjustable valves. Acta Neurochir 156:761–765
10. Tan LA, Kasliwal MK, Moftakhar R, Munoz LF (2014)
Ventriculoperitoneal shunt with a rare twist: small-bowel ischemia
and necrosis secondary to knotting of peritoneal catheter. J
Neurosurg Pediatr 14(3):234–237
This is a relevant manuscript with the news that acoustic phenomena
related to CSF shunt devices are more frequent than previously thought.
Modern ball-in-cone valves with gravitational units in young adults are
especially prone to evoking audible sounds. These noises are usually
gentle and well tolerated by patients. Nevertheless, in their cohort of 29
patients with shunt valve noises, the authors found 57% of them were
disturbed and 24% bothered by the noise. There is a clear demand for
information from shunt patients to their neurosurgeons about this