Computerized Dynamic Posturography (CDP)

Timothy C. Hain, MD • Page last modified: October 24, 2021Return to testing index

cdp test Bertec CDP

Equitest dynamic posturography machine developed by Neurocom Inc. This device is the one we use in our dizzy clinic in Chicago.

Bertec CDP/IVR. This device is termed the "next generation" of dynamic posturography. (See Bertec web site).

Moving platform posturography, also called computerized dynamic posturography (CDP), or posturography for short, is a method of quantifying balance (although the definition of balance can be tricky). It is most applicable in situations where balance needs to be followed quantitatively, to determine whether a disorder is getting better or worse, or the response to treatment. There are several commercial vendors of posturography equipment.

Posturography is a neat and "easy" test. No wires, gel or cameras are needed. You just step on the plate, put on a safety harness, and try to stand up. A full "Equitest" takes only about 20 minutes. A technician can be trained to do an Equitest in about 20 minutes.

The best equipment is generally thought to be the "dynamic posturography" systems, which incorporate servomotors into their design. An example of this is shown above -- actually the "Equitest" apparatus from Neurocom (which no longer exists, being first sold to Natus and then to Bertec). A similar but less expensive machine is the "Smart-Balance Master". We will be mainly talking about the full "Equitest" type apparatus here. If it doesn't move the floor and visual surround, we are not talking about it's utility -- it should be safe to assume though that it is less than the full system.

Static posturography systems are much cheaper, but of course, have less general utility. Neurocom formerly sold one of these, called a "Basic Balance master". These can also be easily found from other vendors. We find generally reports generated by these devices of very little utility. We would just as soon have reported out a timed-tandem Romberg.

Obsolescence of the Neurocom CDP systems.

As of January 2019, Natus announced that Neurocom balance manager systems are no longer being sold. Support will be discontinued in 7 years (i.e. 2026). This means that these devices will eventually be gone from clinics. We were told that these devices will be replace by "Bertec" devices, which perform a similar function. According to the Bertec web site, these devices were developed in partnership with Dr. Lew Nashner, who was the developer of the Neurocom balance systems.

We have not used these devices as yet, and have no information about their utility to either diagnosticians or treaters. We did inquire about the pricing of one of their more elaborate systems, the CDP/IVR, and we thought it was out of the price range of most practices. We expect this will be mainly a "deep pocket" device -- and found in places like the VA hospital system. We expect that cheaper competitors will rise up once support for the Neurocom machines is pulled. One cheaper competitor is made by Biodex. We have had no experience with this device, but it doesn't look as if there is a visual component. There is also a company called "Boditrak" which offers software to do a CTSIB. Again, no experience. We are not even sure if this is a commercial product as boditrak does not respond to emails (as of late 2021).


There are 6 "sensory tests", which are arranged as follows:

Posturography conditions
Condition Vision Surface Visual Surround
1 Eyes Open Stable Stable
2 Eyes Closed Stable Stable
3 Eyes Open Stable Sway-Referenced
4 Eyes Open Sway-Referenced Stable
5 Eyes Closed Sway-Referenced Stable
6 Eyes Open Sway-Referenced Sway-Referenced

The general schema of the 6 sensory "conditions" is shown above. "Sway-referenced" is a term used both for the surface and for the visual surround. For the surface, sway referencing means that the signal from load-cells in the base of the platform is sampled by a computer, and then fed back to servo-motors that attempt to tilt the support surface so as to compensate for torque -- similar but much less effective than standing on a "teeter-totter". It is less effective because of imprecision in separating out sheer force from torque, and also because of limitations of the servo's. There are a number of insolvable biomechanical problems that are intrinsic to attempting to manipulate a multi-link body -- the trunk can go one way, the knees might bend, the head or torso might tilt. All of these present impossible problems to true "sway referencing".

Sway-referencing for vision means that signal from the load cells in the base is fed back instead to a tilting visual surround. Somewhat similar to wearing a lampshade on your head so that your vision moves with your body. This of course assumes that your head is stable on your trunk. The feedback system again has limitations from imprecise input and limitation of servo motor output. Again, there are similar insolvable biomechanical problems for a system like this. One might speculate that a VR system might be able to do a better job. The newer Bertec machines use a video display in an attempt to do this.

A reasonable summary would be that for these conditions in dynamic posturography, distorted signals related to torque at the ankles are fed back to the support surface and to the visual surround. This distorted feedback is destabilizing and makes it much harder for a person to control their center of gravity in space and remain upright. Because the conditions are cleverly arranged in such a way that the surface becomes less stable (4-6), and the visual input is also progressively distorted again between 4-6, the usual situation is that people sway more as the task becomes harder, and thus they have lower scores for the higher-numbered sensory conditions.

This is all somewhat artificial as it would be impossible to create another device that did the same as the branded posturography system, as the results are a result of a proprietary distortion algorithm. So these data are really just meaningful within the world defined by these machines.

Scores and Norms:

Typical Sensory Test
This is a typical "sensory" test graphic. The average score goes down with test #.


For the most useful variant of CDP, "sensory testing", the 6 subtests discussed above are progressively more difficult for a normal person. This means that normal people largely do worse (i.e. score lower), as the 6 subtests range from 1-6.

Scores for each trial of each subtest are computed from the angular difference between the patient's calculated maximum anterior to posterior displacements to the theoretical maximum displacement, and this is expressed as an inverse percentage between 0 and 100. In other words, 100 is no sway at all, and 0 means 12.5 deg of sway.

The composite score, which is computed as the weighted average of tests 3-6 combined with the average of tests 1 and 2. For example, if every test were done 3 times, there would be a total of 18 trials. The composite is the weighted average considering the average of test 1 to be one test, the average of test 2 to be one test, and then the other 12 tests. A total of 14 things to average. The graph here shows that composite scores decline with age in dizzy patients.

Thus, tests 1 and 2 are not weighted as heavily as 3-6, given that one administers a "full test", which includes 3 trials of 6 types of tests -- 18 overall. Of course, the meaning of the "composite" would change should one do less of one subtest and more of another.

Three subtests are "sway referenced", meaning that pressure is used to control the pitch angle of the platform with the goal of keeping ankle angle constant. This is not possible, but it does distort ankle angle input.

Sensory test scores ordinarily decline with age (Wolfson et al, 1992), reflecting the usual decline in balance with age. Repeated testing over 10 days may show a learning effect with better scores (Peterson et al, 2003).

There are also the "motor tests" of CDP, discussed here.

Typical Sensory Test
This is a typical "sensory" analysis graphic. This goes with the sensory tests above

Sensory Analysis:

The individual scores of test 1- 6 are composed of five different conditions involving vision and ankle feedback (vision: normal, distorted, no vision), (ankle: normal, sway referenced). Thus when one looks at the sensory test bar graph above, one is not very informed. One can see how reproducible the scores are, and the general trend, and the average.

An attempt to tease out the contributions of vision, somatosensation, and vestibular sensory input is reflected in the "sensory analysis" section of the report. This involves dividing one linear combination of scores by another, and then replotting them as shown above. There are four sensory analysis scores (as shown above), which are considered one by one in this discussion.

The Motor control test:

Less commonly done is the motor control test, which involves sudden translations or tilts. This is discussed here.

Clinical indications:

Posturography is insensitive to vestibular disorders, and normal posturography should not be considered indicative of normal vestibular function (Di Fabio, 1995). The author has had instances in which there is a severe disturbance of caloric testing and rotatory chair testing, accompanied by a normal CDP. CDP is therefore not an adequate test for vestibular disturbance, by itself.

In addition to the false negative problem (lack of sensitivity), CDP also has false positives - -it suggests that there is a vestibular problem when none exists. The "vestibular" pattern on CDP is actually not specific for vestibular disorders. This is particularly obvious these days when we have newer tests for vestibular (such as VHIT), that can be perfect even when the CDP is "vestibular". Some examples of false positives and negatives are shown here (and see above as well).

CDP may add value to a vestibular battery, when combined with other tests of vestibular function. Stewart et al (1999) suggested that audiometry combined with posturography was a cost-effective method of documenting a vestibular disorder. Obviously, we disagree that it is sufficient. Sataloff and others (2005) recently suggested that CDP adds value because it is abnormal in situations where ENG is normal. This is not necessarily a reason to use CDP, however. If we flipped a coin, and called heads abnormal, we could also make the same observation.

Posturography has been reported to be often abnormal in patients with chronic toxic encephalopathy due to solvent exposure. Serial follow up of these patients with CDP seems very logical.

Allum and others recently concluded that diagnosis of bilateral vestibular loss using posturography is best achieved using measure of trunk control following pure toe-up rotational perturbations under eyes-closed conditions (Allum et al, 2001). This is not a paradigm that is routinely available.

Posturography with the head held in different angles on the neck has been used in an attempt to diagnose cervical vertigo. Static posturography does not appear to be useful. Dynamic posturography, incorporating sway referencing, may be more sensitive (Alund et al, 1991).

Posturography is also very useful in medical legal situations where malingering is a possibility (see below).

Click here to see an example of a posturography output screen (courtesy of Neurocom, Inc). The main vendor of posturography equipment used in clinical contexts is Neurocom incorporated. Other vendors include Micromedical Technology, Metitur, and Vestibular Technologies (Tampa, maker of the "Balance Trak 500") and several makers of research balance equipment (e.g. AMTI, and Kistler).

Pitfalls of dynamic posturography

The main pitfall of Posturography is that it is misleading. A naive clinician might think that they can pop their patients into this box, push a few buttons, and out will come a diagnosis. This is not at all true - -this machine is not a diagnostic device.

While the "Sensory Analysis" suggests that there is a "SOM VIS VEST PREF" pattern, which a naive individual might think that this means that the patient has a proprioceptive, visual, vestibular or mechanism of imbalance sensitive to distortion of input, these are not at all accurate, and don't substitute for a physical examination..

Vestibular pattern in person with normal vestibular system
Patient with "Vest" pattern with 100% normal vestibular system.

The biggest problem is the "VEST" indicator, which one would think would suggest that the patient has a "vestibular" disorder. This is sometimes right and sometimes wrong. There are quite good examples where vestibular function can be shown to be perfect (with contemporary tests we can check inner ear rather thoroughly), but the "vest" pattern is indicated -- wrongly. Don't trust the "Sensory Analysis". To make diagnoses, you need a lot more data than a posturography machine.

Another problem is the "PREF" sensory pattern. This one is hard to figure out.

When combined with other information these patterns may be helpful. But CDP by itself -- it usually not for diagnosis.

Although one might think that the CDP machine is "foolproof", because you essentially just strap the patient in and operate a computer console, practically mistakes are easily made. Most of these problems arise from overly cautious operators (and perhaps patients).

Patient with bilateral vestibular loss, who fell on all but one sensory test. This patient is too unsteady to be tested. Note however that the system software produced strategy scores as well as a "very high Pref" score. The "Cevette" scoring suggested a vestibular etiology, based on a single score (test 1-1). This examples points out the pitfalls in blind reliance on computer analysis, as common sense tells one that if a person is unable to score higher than 2/100 -- they simply could not perform the protocol.


Not all patients with bilateral loss are too unsteady to be tested. Above is a patient who has complete bilateral vestibular loss -- nothing working at all, but who is good enough to play basketball.


As of July, 2002, there were 173 papers with the word "posturography" to be found indexed on Pubmed. This is substantial but not a big enough research effort to answer most of the questions posed below.

Cost effectiveness:

Several studies have suggested that posturography is a cost-effective and/or method of evaluating dizziness (Stewart, 1999; El-Kashian, 1998, Yardley et al, 1998). Our view is that the usefulness of posturography is a function of the type of patients that are being screened. Posturography might be of considerable utility in separating out patients with psychiatric disorders from vestibular or CNS disorders. We are less enthusiastic about the utility of posturography in a population with known inner ear disease, such as Meniere's or chronic ear disorders.

Use of a foam support surface provides a more challenging balance environment than the Equitest ankle-sway referencing system, and Allum has suggested that it may offer a good alternative.(Allum et al. 2002)


Fistula testing:

Perilymph fistulas are a rare cause of imbalance. Although a bedside fistula test procedure, involving aural pressure and sway, has been available using posturography, very little has emerged in the literature regarding success and failure.

Pressure TestPressure traces

The two traces above show a pressure test that was thought to be an indication for fistula surgery on the left ear. There is clearly more sway when pressure is placed into the left side. This equipment, produced by NeuroCom, is no longer commercially available, but a few units may still be in the field. We do not recommend this test for the reason developed below.

Ben-David and others (1997) suggested that posturography was effective in diagnosing Tullio's phenomenon (sound sensitivity, associated with fistula). This use appears reasonable to us, although it is puzzling why more has not been published as of this date (2002). It does seem to us that it might be difficult to distinguish findings related to suggestibility, from organic findings, as we see no practical way to prevent the patient from knowing that there is pressure or sound stimulation exerted in the ear in question.

Bilateral vestibular loss.

CDP in bilateral loss

This patient has complete bilateral vestibular loss, onset about 20 years prior to this test. Curiously, they had an "aphysiologic pattern" using the Cevette method of analyzing the conditions, showing that one cannot always believe the "aphysiologic" score.

Little has been published about CDP in bilateral loss to date. Sargent and others (1997) found that CDP was abnormal in bilateral loss. This is hardly suprising as bilateral loss usually causes substantial and significant effects on balance. On the other hand, the author has encountered patients with near complete bilateral vestibular loss, who performed normally on CDP. Thus, CDP is not 100% sensitive to bilateral loss. Baloh et al (1998) did not find CDP useful in separating patients with bilateral vestibular loss from cerebellar disease. At this writing (2002), more work is needed to calibrate CDP results to the degree of bilateral loss, and also in using CDP to differentiate among various other causes of imbalance.

Malingering and symptom exaggeration.

Follow the link above to find a discussion of CDP in several conditions with imbalance but no testing abnormalities.


Migraine associated vertigo.

One would not think that CDP would be very useful in episodic disorders such as migraine that are largely characterized by headache. The literature here is scanty and mixed. Cass (1997) reported that CDP was useful in identifying patients who could benefit from vestibular rehabilitation. Dimitri et al, in a careful study, found that CDP does not differentiate Migraine associated vertigo from Meniere's disease.

Movement Disorders

Only a handful of papers have been published regarding movement disorders such as Parkinsonism and PSP. More work is needed in this area.

Multiple Sclerosis

Posturography in person with MS and spastic gait.

We do not think that CDP is generally a reasonable procedure in persons with multiple sclerosis, as reported by Williams (1997), as MS is a multifocal disorder. We have tested occasional persons however. The picture above shows the result in one person, who had a spastic gait. Balance was poor. The "Cevette" analysis suggested an "aphysiologic" score. This is appropriate as this person was neither normal nor did he have a vestibular disorder.

Posturography in patient with SCA-3 (spinocerebellar atrophy type 3). This patient also had an aphysiologic score on their Cevette testing.

Cerebellar Disorders

These patients have a central ataxia, due to cerebellar damage. As posturography is mainly tuned to vestibular problems, one would think that it might show primarily a vestibular deficit (although this is really not true) but with an "aphysiologic" pattern. This is sometimes the case. A patient with a spinocerebellar ataxia (sca3) is shown above. They had what was mainly a vestibular pattern on the manufacturer's printout, but an "aphysiologic" pattern on the Cevette analysis. Not much different than bilateral vestibular patients. A patient with a paraneoplastic cerebellar degeneration is illustrated here.

From these cases, it seems plausible that a more sophisticated "Cevette" algorithm --- i.e. linear discriminant analysis -- might be able to differentiate between inner ear and cerebellar disorders to a greater extent. This could be a good project for future work.

Peripheral Neuropathy

In spite of a considerable research outflow, we do not think that posturography is a reasonable primary method of diagnosing peripheral neuropathy, as there are many more direct methods of doing this (such as measuring sensation). It might be a useful method of partitioning out how much imbalance is due to sensory disturbances to the feet, among other sensory impairments.

Orthostatic Hypotension

Posturography is generally normal in persons with orthostatic hypotension.

Posturography should be normal in persons with adequate blood pressure to stand upright during the test. An example of this is shown above.

Vestibular disorders

Most CDP research to date has been done on vestibular disorders. It appears clear right now that CDP is moderately sensitive to vestibular disorders.

On the other hand, the author has encountered patients with complete unilateral loss of vestibular function, who had normal results on CDP. An example is shown here, and another here. Thus CDP is not 100% sensitive. CDP is far more often false positive -- as nearly any type of imbalance will often be designated vestibular.

CDP results depend on the degree of the unilateral lesion, how long the person has had to compensate, the age of the person in which it has appeared, and their motivation to compensate. CDP does not appear to be very specific -- CDP measures something related to imbalance, which is found in a variety of sensory and central medical disorders.

Bottom line regarding utility:

At the present writing, we think dynamic CDP is useful in certain clinical situations. As we get more experience with CDP, it seems likely that it's indications will expand. Much more research is needed on CDP.

As dynamic CDP is not overwhelmingly useful (but does have a definite utility in malingering), it also follows that stripped-down versions of CDP that omit the servo-motor or the visual surround are rather marginal in utility.

Insurance coverage:

There are two CPT codes for CDP.

CDP is not uniformly covered by health insurance in the United States. Medicare covers it in some areas but not in others. In Illinois, Medicare does cover CDP 100% of the time. Rather peculiarly, in Illinois, Blue Cross Blue Shield generally doesn't cover posturography (about 19% of claims are paid -- depending on the type of BCBS). Cigna covers about 27% of the time, United Healthcare Medicare 100%, United Healthcare 0%. This seems mainly to be due to purposeful omissions of coverage by these insurances, although when questioned, Insurances sometimes claim that it is "not medically necessary'. It is difficult to see why different plans within BCBS consider CDP medically necessary and others do not. It would seem more accurate to simply state that some insurance covers more testing than others.

Fraud and questionable activities related to CDP insurance coverage. 92548 is for the "sensory tests", and 92549 is for both the "Sensory tests", and "motor control tests" (MCT). The 92549 code was introduced recently (i.e. about 2020). The "Sensory tests" can be (sort of) done by the "Foam and Dome" type low-budget equipment, but the MCT cannot be faked, and requires a rather expensive and complex machine. So perhaps the reason for the 92549 is to encourage use of the "real" CDP machine.

Nevertheless, in our opinion, dynamic CDP (namely Equitest and very close relatives, that explicitly include motorized visual surround and platform movement, such as "Smart Balance-Master" ) should have limited coverage by health insurance. CDP is FDA approved and has two CPT codes. It has one well documented diagnostic indication - detection of malingering (Cevette et al. 1995; Goebel et al. 1997; Krempl et al. 1998; Morgan et al. 2002). Dynamic CDP data can be pivotal in medicolegal situations as it can strongly suggest that individuals are feigning imbalance.

On the other side of the issue, the use of CDP in diagnostic contexts outside that of malingering is not well established. El-Kashlan and others found CDP to be more sensitive than physical examination in distinguishing patients with vestibular disturbances from normals (El-Kashlan et al. 1998). Stewart and others suggested that CDP is as cost-effective as audiometry and ENG testing for the evaluation of vertigo (Stewart et al. 1999). However, in the authors opinion, this conclusion differs from that of most specialists that evaluate vertigo. Dimitri and associates (Dimitri et al. 2001) found CDP of no utility in distinguishing migraine associated vertigo from Meniere's disease. Baloh and associates (Baloh et al. 1998) found CDP unable to distinguish between cerebellar and bilateral vestibular patients, a task that most dizziness specialists would find quite simple.To summarize, there is presently only a small amount of data about broader diagnostic usefulness of CDP, and it is conflicted.

Another potential indication for CDP is to guide physical therapy or to document the outcome of physical therapy. While several authors suggests that it is useful in this context, (Mirka et al. 1990; Shepard 1996; El-Kashlan et al. 1998), others find it of no use in documenting functional status (O'Neill et al. 1998). Outcome studies in which therapy guided by CDP is compared to therapy without CDP are presently lacking. We would like to see more work done here. We don't think that CDP should be billable by groups other than physicians.

We thank Neurocom Inc, for use of figures of their equipment to illustrate this page. (Neurocom was purchased by Natus in 2018, and no longer exists as a corporate entity).

CDP References: