For a variety of reasons, false positives and misinterpretations exist in impendence audiometry. So, no it is not 100% sensitive or accurate. Probably more of importance is the sensitivity and visual scaling that tells us the difference between a shallow A and a flacid A...the difference in auditory behavior and etiology is vast, but rarely considered in a world more concerned about the passing Bs and Cs (smile).

It is known that normal middle ear (empty of fluid) due to normal Eustachian tube dysfunction can be confirmed by tymanogram Type A i.e. true negative regarding  absent otitis media with effusion. In case of otitis media with effusion, the tympanogram will be Type B (with normal ECV), i.e. true positive regarding present otitis media with effusion. So, when a patient was diagnosed otitis media with effusion due to decrease hearing & presence of dull tympanic membrane but his tympanogram is Type A rather than B, this test will be considered false negative. Practically, i faced some cases like that. So, i think the sensitivity & accuracy of tympanometry is not 100%.  

Yes, otoscopy and other measures need to verify tympanometry--a collapsible ear canal and obstructions in the probe can render a useless reading. Another consideration is case history--age, ear infection history, allergies, post op, etc. 

Thanks Max for your comment.

Tympanometry is a useful test.However is not 100% accurate as false positive and negatives occur. Best to corelate with clinical history and examination.

I agree with Nadarajah.  Tympanometry is a test measuring tympanic membrane compliance which is nil during OME or presence of any middle ear fluid. Of my experience, tympanometry has no 100% sensitivity because sometimes i meet some pediatric patients with decrease hearing & tympanometry is normal (Type A) but clinically their tympanic membranes were dull, so clinically they were considered case of OME & due to 3 months F/U with no improvement despite proper medical management, patients underwent surgery (myringotomy & ventilation tube insertion) which confirmed the diagnosis.

According to the report AHRQ Publication No. 03-E023 page 64, which can be found here: http://archive.ahrq.gov/downloads/pub/evidence/pdf/otdiag/otdiag.pdf,

the diagnostic test with the highest specificity (95%) is tympanometry performed by a proffessional. The sensitivity was 93,8%. However, the article suggested that pneumatic otoscopy had the best balance between sensitivity and specificity.

Hope this helps.

Reagards to all

Thanks Panagiotis for your valuable comment & attached pdf file.

Good question! Tympanometry is certainly not 100% sensitive or specific. AR Maw wrote an excellent summary of this in the paper referenced below(1). He cites Brooks’s work (2) on the issue and mentions a sensitivity of ~90% and a specificity of ~75%. Of course there are more complicating factors surrounding the use of tympanometry to diagnose OME - the tracing is often challenging to interpret - is it a Type B or a Type C? Was there a technical issue during the test? Was there an adequate seal? etc. Another study showed that 34% of children undergoing surgery for otitis media with

effusion were found not to have middle ear fluid(3).

Having said this about tympanometry, the ability to identify OME by otoscopy alone is disturbingly poor (4). So it may remain the most reliable practicable test that we have for now. The issue of a 'false negative' tympanogram in the presence of reduced hearing and abnormal otoscopy should be seen in this context.

1) Maw AR. Using tympanometry to detect glue ear in general practice. BMJ. 1992

Jan 11;304(6819):67-8.

2) I Brooks DN. Acoustic impedance studies on otitis media with effusion. Int] Pediatr Otorhinolaryngol

1982;4:89-94. 3) Black N, Crowther J, Freeland A. The effectiveness of adenoidectomy in the treatment of glue ear: a randomized controlled trial. Clin Otolarvngol 1986;11:149-55. 4) Buchanan CM, Pothier DD. Recognition of paediatric otopathology by General

Practitioners. Int J Pediatr Otorhinolaryngol. 2008 May;72(5):669-73.

Thanks David for you valuable sharing.

Dear David regarding "children undergoing surgery for otitis media with effusion were found not to have middle ear fluid", this is another important issue. Sometimes, i meet some pediatric patients diagnosed OME due to reduced hearing & dull TM on otoscopy with tympanogram Type B (ECV is normal) and these all findings continue after medical management  with 3 months F/U. When these children underwent surgery, TM looks not dull & even no fluid present after myringotomy & i discussed this with more senior otologists, some advised to not put VT & others advised to put VT particulary if there was previous recurrent AOM for 3 months. In these cases there might be no false positive Tympanogram but true positive one (Type B with normal ECV) which accorded with preop clinical diagnosis of OME (reduced hearing & dull TM on otoscopy) and the possible explanation of disappearance of middle ear fluid during surgery for OME is forced open of Eustachian tube by forced air pumping respiration of patient, performed by aneasthist during GA leading to drainage of ME fluid via momentary opened Eustachian tube. What do you think?   

h have seen one case with normal tympanic membrane on cl.exam.,bil.CHL,in male 60 years diagnosed as otosclerosis by one senior,on tympanometry it was type B[ome],on surgery fluid was found....

Thanks so much for all.

Of course, the fluid aspect is a transient one--often going from fluid to adhesive to residue over time. Lack of Eustachian tube patency and resulting own voice occlusion may be the only distrubing clue to the sufferer as so many OME cases are allergy related with nonsuppurative presentation. I cannot count the number of times patients have asked me to clean out their ear canals, as they were positive it was plugged with cerument--only to find a clear canal--but upon cone of light otoscopy fluid or at least negative pressure visible. Pseudoephedrine--so hard to come by today in most countries because of the inordinate reguations that have all but obliterated it to those who need it while not making a dent in the illegal concoctions it is also used in--still is the best gentle approach to restoring mild Eustachian function without the depressant effects from antihistamines (which I think are far overused).  But in many of the allergy cases a mere 60 mg of Pseudoephedrine will open the Eustrachian tube and allow equalization of outer and middle ears in a short time. Longer standing and more involved cases, of course, will need whatever the physician deems necessary, even equalization tubes in chronic cases.     

Thanks Max for your valuable comment. During attending the otological conferences, a medical devices company presented an automated modified politzer apparatus for in-office or home treatment of ETD & OME without need of antibiotherapy or surgical VT insertion. Dear MAX, What do you think about this new therapeutic option? Are there any study comparing this new therapeutic procedure into the old classic maneuver like blowing while nose & mouth are closed. Are there  evidence based experiences regarding this new therapeutic option of OME? 

Hazem, each case will depend on the particulars: age of patient, etilology (purulent/non-purulent-suppurative/non-suppurative), culture results, rest of body diagnosis (secondary to something else?), inhalent allergy. I would say that at least 90% of OME cases are non-bacterial and of course not amenable with antibiotic therapy. Is the fluid adhesive, and is this a chronic situation that could delay development of the child or interfere with work etc. in an adult? So many variables, and my greatest concern when I teach ENTs and audiology teams is if they spending the time to sort them out and taking the short cut by automatically reaching for the script pad and ventilation tubes. More important, are they looking for tooth/jaw sepsis, which is a major cause of ear problems? We have determined from years of research that chronic bacterial infections (EO, OME, etc.) arise from throat and jaw issues more than from the ear itself. Of course, infants who are fed a bottle lying with the heads backwards will get formulae into their middle ear via Eustachian tube. So there is much to look at and taking the time to handle it intelligently is the key to appropriate treatment.

Thanks Max for your re-comment. 

I think tympanometry combined with audiometry is 100% sensitive for detecting middle ear effusion.  See my audiological algorithm for this in J Laryngol Otol 1976;90:141 which was tested by immediate myringotomy in children.  (Note missing plus sign in Fig 7).

Effusions were found in 154 ears, and in all of these fluid had been predicted by the algorithm.  The experienced registrar, inspecting the ears under the operating microscope immediately pre-op,  had predicted fluid in 132 ears, possible fluid in 19, fluid unlikely in 3.  

There were, however, audiological "false positives".  In 20 cases, ears predicted by the algorithm to contain fluid were found to be dry.   Follow up or reassessment of these cases and consideration of many points or arguments (Dry myringotomies and impedance testing  Pediatrics 1978;61:152) led to the clear conclusion that these ears had indeed contained fluid, and that it was myringotomy that was the fly in the ointment, ie false negative myringotomies, not false positive audiological prediction.  To answer the question posed, you need a Gold Standard criterion for fluid presence, and this is undoubtedly tympanometry.

Middle ear pressures in those with fluid at operation ranged from -50 to -400 mm water.  In those with normal gradients and sharp peaks on the curve (the Type C population), fluid was present if the average 4 and 8 kHz pure tone hearing was 15dB or worse.

Thanks Anthony for your valuable comment. Dear Anthony you said "ie false negative myringotomies", i think there is no false or true myringotimies because myringotomy is a procedure & not a test. So, we can say that myringotomy revealed presence or absence of middle ear fluid & this accord or dis-accord with preoperative tympanometry findings. Hence, tymp will be false negative if clinically there is hearing loss with dull TM in otoscopy while tymp is type A, dis-according with finding of ME fluid during myringotomy. On the other hand, tymp will be false positive if clinically there is no hearing loss with no dull TM in otoscopy while tymp is type B, dis-according with finding of absent ME fluid during myringotomy. So, false negative tymp is an issue of decrease sensitivity but false positive is an issue of decrease specificity. What do you think?

"myringotomy is a procedure & not a test"

It is both, surely.   A myringotomy is done to remove fluid, to improve the hearing, to confirm a diagnosis, to clarify a hearing loss,  to provide a definitive explanation for a middle ear hearing loss.

If fluid is found at operation, this speaks for itself.  If no fluid is found, it is clear that in some cases this had been present immediately before operation.  So operative findings are not fit as an absolute criterion, a Gold Standard.  Test accuracy cannot be calibrated against an unreliable outcome.  The best criterion for fluid is probably overall weighted clinical impression, amalgamating results from myringotomy, otoscopy and audiology,  taking account of inherent unreliability of individual components.  If forced to rely on one of these 3 independent indicators, audiology/tympanometry is undoubtedly the best.

Myrigotomy with VT is a therapeutic procedure for a case of unimproved OME after proper management & 3 months follow up, which is dependent on preoperative optimal clinical examination  & hearing tests (PTA & Tympanometry).  

quite a few cases where impedacne is type A and middle is full of serous fluid.

Insertion of tympanostomy should not be performed purely on tympanometry finding.

Clinical history and examination should take precedence. Tympanometry is useful in confirming the clinical diagnosis.

Abdul raised the point that a serous filled middle ear cavity can yield what appears a Type A--in my experience, that is a transient reading and if retested a short interval later will yield the expected impendence pattern. Ascertaining with otoscopy and patient report of occlusion or own voice loudness is also necessary.

Agree with Max. The basic point is findings on tympanometry alone should not be the only indication for insertion of tympanostomy tubes.

Thanks for all who shared my topic and i can conclude that the principal guide to do or not VT after proper management & watchable follow up for 3 months, are the history of disease, clinical findings, otoscopic findings & only according tympanometry (Type B + normal ECV in case of decrease hearing & dull TM in otoscopy, or Type A in case of return of hearing to normal & intact normal TM in otoscopy) with PTA (CHL in case of decrease hearing & dull TM in otoscopy or normal audiogram in case of return of hearing to normal & intact normal TM in otoscopy). If tympanometry does not accord with specialized doctor' diagnosis, then VT decision depends on continued non improved complain of patient & positive clinicootoscopic findings. So i totally agree with Nadarajah & Max "The basic point is findings on tympanometry alone should not be the only indication for insertion of tympanostomy tubes".

Thank you Hazem.

Very nicely summarized.

I also agree with Hazem that tympanometry result should not be the only basis for tympanostomy tube placement

My experience is that tympanometric findings should not be relied upon in 100% cases. On many occasions,  I have found middle ear was full of fluid whereas the tympanogram was of type A. Clinical judgement should be given more importance. 

Thanks Saileswar for your comment. Could you mention the percentage of your cases with middle ear full of fluid whereas the tympanogram was of type A.

No it is not. otoscopy is the best choice.

Thanks so much Dr. Leonardo for your kind sharing. Do you agree with my last conclusion?

I agree we start with otoscopy, and might add the better lighting and close-up view of video otoscopy. It is possible to obtain a false negative (Type A, but more likely Shallow A) as OME is a transitory condition and fluctuates throughout the day. We might catch the condition at any point of the cycle and not realize its longer term implications.  

Thanks so much Prof. Max for your valuable re-comment on my topic.

ONE CASE WAS DIAGNOSED AS OTOSCLEROSIS IN ADULT,BUT ON TYMPANOMETRY WAS OME, confirmed at surgery.

What were the test results?

Tympanometry can be normal in the presence of OME. Equally abnormal tympanogram does not always mean presence of OME.

OME is best diagnosed by taking number of factors into consideration. History, presence of hearing loss, otoscopic or microscopic examination and tympanometry.

Otoscopy is best way for diagnosis 

Otoscopy via red reflex and cone of light otoscopy is helpful in confirming or delineating function, but gives little actual middle ear information from which to know the status of the middle ear. Tympanometry combined with advanced video otoscopy provide, in my opinion, the most objective and definitive information from which to assess middle ear function.   

I think the diagnosis of otitis media with effusion is mostly depends on the collection of data from the clinical assessment which is very important as well as the diagnostic tools such as tympanometry .,I mean the tympanometry sometimes gives false negative .

I agree with Rafid. OME is diagnosed on history, clinical examination and tympanometry. Diagnosis should not be made solely on tympanometry.

N.Manoharan