1.    The Workers' Compensation Division is directed to establish appropriate guidelines for the specific manner in which audiograms should be administered. In developing these guidelines, the Division should consider, inter alia: (1) whether all audiograms should be administered using a uniform brand and model of audiometer; (2) whether guidelines should be adopted for when and how audiometers should be uniformly calibrated; (3) establishing a definitive margin of error; (4) requiring audiologists to perform specific reliability and validity checks during the course of an audiogram; (5) modifying the existing WC-123HL form to allow for the reporting of any required reliability and validity checks; (6) whether the speech discrimination portion of all audiograms should be conducted using only a recorded voice; and (7) what method should be used to select an appropriate audiogram when two audiograms that are both rated “good” differ by more than the established margin of error.

    2.    Until such time as the Workers' Compensation Division has promulgated additional rules for administering audiograms in workers' compensation hearing loss cases, when two valid audiograms that have both been performed after the claimant's date of last exposure to occupational noise are within a margin of error of plus or minus ten decibels, and do not differ by the same amount or in the same direction at all frequencies, the rule of liberality should be applied, and the claimant should be granted a permanent partial disability award based upon the audiogram demonstrating a higher level of hearing loss.

    3.    Until such time as the Workers' Compensation Division has promulgated additional rules for administering audiograms in workers' compensation hearing loss cases, when two valid audiograms that have both been performed after the claimant's date of last exposure to occupational noise fall outside a margin of error of plus or minus ten decibels, or are within a margin of error of plus or minus ten decibels but differ in the same amount or in the same direction at all frequencies, then the claimant should undergo additional audiometric testing. Presumably the third audiogram will be within a margin of error of ten decibels of one of the existing two audiograms, and will not differ by the same amount or in the same direction at all frequencies, so that the rule of liberality may be applied to the two audiograms falling within this criteria, and the claimant may be granted a permanent partial disability award based upon the audiogram demonstrating a higher level of hearing loss. If a claimant chooses to have the additional testing performed by a physician of his or her choosing, then the claimant shall pay the expense of the testing. If, however, the claimant chooses to have the Division select the examining physician, then the Division shall be responsible for such cost.

Davis, Chief Justice:

    This appeal presents a challenge to the way permanent partial disability awards have been determined in Workers' Compensation hearing loss cases since this Court handed down its decision in Bilbrey v. Workers' Compensation Commissioner, 186 W. Va. 319, 412 S.E.2d 513 (1991). Particularly, Mr. Jasper Blackburn challenges the practice of automatically basing a disability award on the audiogram demonstrating the lowest level of hearing loss when there is a discrepancy between audiograms that exceeds the margin of error. We find that additional rules should be promulgated to create uniformity in the way audiograms are conducted and to establish a method for selecting the best valid audiogram. Consequently, we direct the Division to promulgate such rules. Until such time as these rules are in place, we find that where two valid audiograms are within a margin of error of plus or minus ten decibels, the liberality rule should be applied, and the claimant should be given the benefit of the audiogram demonstrating a higher level of hearing loss. Where two valid audiograms differ by a margin greater than plus or minus ten decibels, then an additional audiogram should be performed.

    Nearly two years later, by order entered March 4, 1998, the Division granted Mr. Blackburn a 10.65% permanent partial disability (hereinafter “PPD”) award based upon Dr. Touma's recommendations. Both parties protested the order and it was referred to the Workers' Compensation Office of Judges (hereinafter “the OOJ”). In connection with the protest, Dr. Touma was deposed on November 3, 1998. He testified that Dr. Hatfield's audiogram revealed better thresholds than his own, and that Dr. Hatfield's audiogram had also produced a more accurate representation of Mr. Blackburn's true hearing loss impairment. Thereafter, by order entered June 8, 1999, the OOJ affirmed the Division's award of 10.65% PPD. However, the order stated that because Dr. Hatfield's report had not been included in the record, it had not been considered by the OOJ in reaching its decision in this case. Marrowbone subsequently filed a motion for reconsideration based upon the Division's failure to include Dr. Hatfield's report in the record submitted to the OOJ. Marrowbone's motion was granted, and the OOJ subsequently issued an order finding that the audiogram obtained by Dr. Hatfield was the most reliable. Based upon Dr. Hatfield's audiogram, the OOJ reduced Mr. Blackburn's PPD award to .73%. On appeal, the Workers' Compensation Appeal Board (hereinafter “the WCAB”) affirmed the .73% PPD award by order entered March 31, 2000. Mr. Blackburn then appealed the WCAB order to this Court, and oral argument was had on June 4, 2002. Thereafter, on June 13, 2002, this Court, on its own motion, scheduled this case for re-argument and directed the parties to file briefs addressing specific questions posed by the Court. Specifically, the Court asked the parties to address the following issues:
    (1) are all tests being done at the level as specified in West Virginia law; (2) is there a standardized system of determining reliability of the tests; (3) set forth reasons why this Court should retain its holding in James Bilbrey vs. WCC and Ranger Fuel Corporation, 412 S.E.2d 513 (W. Va. 1991), adopt the rule of liberality, or adopt an alternative, and if so, what alternative; (4) explain the methods used to ascertain whether tests are being administered in accordance with West Virginia law and rules and regulations; and discuss the impact of a margin of error.

        The parties are hereby directed to inform the Court of any differing expert opinions which have come about since 1991 when the Court decided Bilbrey[ v. Workers' Comp. Comm'r, 186 W. Va. 319, 412 S.E.2d 513 (1991)], regarding how the “best valid audiogram” is to be determined when all audiograms are reliable and within the margin of error.

        Further, the parties are hereby directed to advance to the Court their own recommendations for resolving conflicts in the evidence in work-related hearing loss claims.

        Finally, the parties are hereby directed to relay to the Court the differing schools of thought, if any, as to whether work-related hearing loss is progressive or static.

The case was re-argued and submitted for decision on October 8, 2002.

    Marrowbone responds that this Court in Bilbrey correctly recognized the universally accepted medical principle that any deterioration in an individual's hearing after the date of last exposure is due to some factor other than the prior occupational noise exposure. Following this principle, Marrowbone notes, Drs. Hatfield and Touma both agreed that Dr. Hatfield's audiogram should be used to determine the degree of whole person impairment suffered by Mr. Blackburn. Marrowbone further submits that the rule of liberality is intended to resolve conflicts in the evidence. Marrowbone points out that there is no conflict in the evidence presented in this case due to the agreement of the examining physicians that Dr. Hatfield's test represents the most accurate measure of Mr. Blackburn's impairment. Moreover, Marrowbone argues that hearing loss claims are unique by nature, as recognized by this Court in Bilbrey, and, due to the inherent differences between hearing loss claims and other types of claims (particularly in the fact that noise induced hearing loss is not a progressive condition _ as are other types of occupational conditions), the principles this Court established in Bilbrey were correct and should not be repudiated. Marrowbone finally observes the liberality rule was well established at the time Bilbrey was decided, and thus it is not a valid basis upon which to alter Bilbrey.

    The Division responds that in view of the concurring opinions of Drs. Touma and Hatfield, the only evaluating otolaryngologists of record in this claim, the reliable, probative and substantial evidence of record overwhelmingly supports the decision of the OOJ. Hence, the WCAB order affirming the OOJ was not plainly wrong. Additionally, the Division contends that the treatment of hearing loss claims pursuant to Bilbrey is not inconsistent with the liberality rule. The Division directs this Court to the well-established principle that the liberality rule does not relieve claimants of the burden of substantiating their claims. In this case, the Division maintains, both evaluating physicians agree that the proper measure of Mr. Blackburn's impairment is .73%. Thus, there is no proper evidence supporting a higher award.

     1. Are all tests being done at the level specified in West Virginia law? The parties agree that all audiometric tests currently being relied upon to either support a diagnosis of noise-induced hearing loss or to ascertain the appropriate level of disability resulting therefrom are being done at the levels specified in West Virginia law, specifically, W. Va. Code § 23-4-6b (1986) (Repl. Vol. 2002).   (See footnote 5)  The parties further agree that there are no requirements as to the order in which the tests are performed, and no uniformity as to the manner in which the tests are being conducted. The Division additionally explains that since at least 1958, it has been recognized that certain methods of presenting the pure tones used in an audiogram can affect the patient's responses.

     2. Is there a standardized system of determining reliability of the tests? The parties' responses to this particular issue differ somewhat. Mr. Blackburn contends that there is no uniform system for determining the reliability of audiometric testing as the outcome of audiometric testing is essentially subjective. Marrowbone, on the other hand, asserts that the audiologist's rating of a test as “good,” “fair,” or “poor” provides one method of determining the reliability of the tests. However, Marrowbone suggests that any rating other than “good” by an audiologist should be accompanied by a specific explanation of the reason for the ranking. Marrowbone also states that there are other reliability indicators included on the WC-123HL form,   (See footnote 6)  but in its experience the Division ignores these factors and relies solely on the audiologist's ranking. The Division agrees that the audiologist's rating is a method of determining reliability, and submits that there are a few other standardized methods for assessing the reliability of an audiogram. For instance, audiologists/ontologists typically conduct what is referred to as an SRT/PTA Comparison   (See footnote 7)  to gauge the reliability of a test while it is being performed. Comparing two audiograms provides another measure of reliability. A third method of assessing reliability is to repeat the same pure tones at different points during an evaluation, as it is virtually impossible for an individual to remember whether or not he responded to a particular tone after having heard other tones in the meantime. The Division however, does not, assert that this method is used regularly, and we are aware of no requirement for its employment. Finally, the Division asserts that an indirect test of reliability arises from the education and training required in order to become an audiologist under W. Va. Code §§ 30-32-1 to -23 (Repl. Vol. 2002).

     3. Should the Court to retain its holding in Bilbrey, adopt the rule of liberality in hearing loss cases, or adopt an alternative rule? Mr. Blackburn again argues that Bilbrey should be abandoned because the measurement of noise-induced sensorineural hearing loss is subjective and lacks any standardized method of evaluating reliability. Mr. Blackburn submits that all evidence should be presumed reliable unless it can be proven otherwise and the claimant should be entitled to the benefit of the most favorable results obtained. Both Marrowbone and the Division contend this Court's holding in Bilbrey should be retained as the medical foundation upon which Bilbrey is based has not changed. Occupational hearing loss is a non-progressive occupational disease. This unique medical fact makes it inappropriate to use the rule of liberality to choose the higher of two widely- varying audiograms when attempting to ascertain the proper amount of impairment sustained by a claimant whose exposure to noise ended prior to the audiograms being administered. The Division additionally asserts that to use the rule of liberality in such cases would allow a judicially-created rule to take the place of proven scientific fact and would relieve claimants of their burden of establishing their claims.

     4. What are the methods used to ascertain whether tests are being administered in accordance with West Virginia law and rules and regulations? In an answer that is somewhat unresponsive to the specific question asked, Mr. Blackburn submits that West Virginia laws, rules, and regulations pertaining to how tests are being administered are inadequate. For example, he states that there are no statutes or regulations providing standards for the calibration of audiometers, for testing environments, or for any other equipment that may be used. Acknowledging that physicians are permitted to consider only noise-induced hearing loss in determining the degree of a claimant's impairment, Mr. Blackburn complains that it has become a widespread practice among physicians to adjust the threshold hearing shifts measured at each frequency to reflect other factors, particularly the effects of the natural aging process, even in the absence of any evidence that a particular claimant's hearing has actually been diminished by such factors. Mr. Blackburn contends that this practice is inconsistent with legislative intent as neither the statute nor the regulations provide for any “adjustment” of test results based upon statistical probability or mere speculation by an evaluating physician. Marrowbone contends that the only method utilized by the Division to ensure that testing is being conducted in accordance with West Virginia law and rules is a letter to its evaluating professionals titled “Report Outline For Permanent Partial Disability Evaluation (Noise- Induced Hearing Loss Only),” which outlines the information sought by the Division and the procedure the Division requests the specialists to follow. The Division explains that, at this time, the only method in use to ascertain whether tests are being administered in accordance with West Virginia law, rules, and regulations is a visual inspection by its claims managers of the reports received from physicians and/or audiologists.   (See footnote 8) 

     5. What is the impact of the margin of error? Before relating the parties' comments regarding the impact of the margin of error, it should be noted that there presently is no law or regulation in place establishing the acceptable margin of error for hearing loss cases in this State. Mr. Blackburn and the Division are both supportive of using a margin of error of plus or minus five decibels when comparing audiograms by the same audiologist on the same machine that were administered on two different occasions. They also agree that a margin of error of plus or minus ten decibels is acceptable when comparing audiograms by different audiologists on different machines, due to possible differences in earphone placement and calibration. Marrowbone, however, considers a difference of plus or minus ten decibels to be too great, and urges that the margin of error should be set at plus or minus five decibels even when two audiograms have been administered at different clinics. Marrowbone and the Division effectively agree that the margin of error should not differ by the same amount, or in the same direction, at all frequencies. In other words, there should be interweaving between two valid audiograms.   (See footnote 9) 

In the absence of interweaving, they argue, then the audiogram results should then be considered suspicious. Using the instant case as an example, Marrowbone notes that Mr. Blackburn's two audiograms showed worse results at every one of the eight relevant frequencies. Thus, Marrowbone suggests, the differing results were not simply due to the “margin of error.”

    With regard to how the margin of error should be applied once it has been established, Mr. Blackburn and Marrowbone agree that when two tests fall within the appropriate margin of error, the results more favorable to the claimant should be used to determine his or her level of impairment. Where the difference between two audiograms exceeds the margin of error, Mr. Blackburn suggests that a third test should be performed by an evaluator of the claimant's or the Division's choosing, while Marrowbone suggests that the Division should be required to refer the claimant for a third audiometric evaluation that would include all tests necessary to ascertain the claimant's true level of hearing loss, including, but not limited to, brain stem audiometric testing and acoustic reflex testing.   (See footnote 10) 

The Division opines that when two audiograms differ by more than the accepted margin or error, and both have been performed after the date of last exposure, the evaluator needs to look for a cause other than occupational noise for the decreased hearing levels. Because noise-induced hearing loss is a static condition, the Division maintains that it is inappropriate to use the rule of liberality to choose the higher of two widely-varying audiograms in assigning the level of impairment.

     6. Are there any differing expert opinions that have come about since Bilbrey regarding how the “best valid audiogram” is to be determined when all audiograms are reliable and within the margin of error? Mr. Blackburn, Marrowbone and the Division all agree that the medical foundation upon which Bilbrey is based has not changed, and they are aware of no expert opinions contrary to Bilbrey regarding how the “best valid audiogram” should be determined.

     7. What are your recommendations for resolving conflicts in the evidence in work-related hearing loss claims? Mr. Blackburn adopts the recommendation advanced by Dr. Gary D. Harris in his deposition in this case. Dr. Harris' recommendation assumes that all audiograms under consideration are rated reliable and valid, and that they were performed after the date of last exposure to occupational noise. The recommendation also considers each ear separately and uses the four-frequency decibel sum for 500, 1000, 2000 and 3000 hertz, which are the levels used to calculate whole person impairment under the West Virginia Workers' Compensation laws. See W. Va. Code §§ 23-4-6b(b)(1) and (c)(1). Specifically, Dr. Harris suggests:

    If the four frequency decibel sum differs by 20 dB or less, then such audiograms are probably within the range of normal multiple-audiogram variability and both may represent a claimant's permanent hearing levels. Using the “worse” audiogram (the one with the highest four-frequency decibel sums) would give the claimant the benefit of the difference.

    If the four-frequency decibel sum differs by 40 dB or more on two different audiograms, then either some individual frequencies have differed by more than 10 dB, or there has been no interweaving^[   (See footnote 11)  ^] in the two audiograms. Two audiograms with four-frequency decibel sums that vary by 40 dB or more in either ear are significantly different, regardless of the reason for the difference. The better of the two tests, assuming it met all the criteria previously discussed, is a better representation of the claimant's permanent hearing loss.

    This leaves then audiograms with four-frequency decibels sums [sic] that differ by more than 20 dB, but less than 40 dB. In such instances it would be reasonable to obtain a third audiogram to see if the most accurate representation of a claimant's permanent occupationally related hearing levels can be determined.

    Stated simply, Marrowbone suggests that where two audiograms are within a margin of error of plus or minus five decibels, regardless of whether there is interweaving,   (See footnote 12)  the claimant's award should be calculated from the audiogram that is most favorable to him or her. Where the difference between two audiograms is between five and ten decibels per frequency, Marrowbone proposes that additional testing should be performed to obtain an accurate depiction of the claimant's true hearing level. Finally, where the difference between two audiograms represents a difference that is greater than an average of ten decibels per frequency for the eight measured frequencies in the whole person impairment calculation, then the audiogram demonstrating the greater degree of impairment should be discounted so long as: (1) the claimant is no longer exposed to occupational noise; (2) both audiograms were obtained after the claimant's exposure to occupational noise had ceased; and (3) the audiogram showing the lower degree of whole person impairment was unequivocally of “good” reliability.

    Finally, the Division recommends that when two audiograms that are both deemed reliable differ by an amount that is within the accepted margin of error, either audiogram would be appropriate upon which to base a compensation award. In these circumstances, the rule of liberality could be applied to select the audiogram that would provide a higher award of compensation. When two reliable audiograms differ by an amount greater than the margin of error, then the audiogram demonstrating the best level of hearing should be used. However, when two audiograms differ by an amount considerably greater than the margin of error, the prudent course would be to perform a third audiogram in an effort to determine the most accurate representation of a claimant's hearing loss.

     8. Is noise induced hearing loss progressive or static? The parties agree that once exposure to occupational noise has stopped, occupational hearing loss becomes a non-progressive condition.

    Prior to Bilbrey, there had been “little or no consistency in the manner in which the Commissioner grant[ed] permanent partial disability awards for noise-induced hearing impairment or in the tests that [were] required in order to determine what percentage of loss [was] actually due to noise.” 186 W. Va. at 324, 412 S.E.2d at 518. Bilbrey established numerous guidelines in an attempt to correct these inadequacies and to provide a system under which the Division could reach consistent results while also providing a record that would permit this Court a meaningful review. However, Bilbrey stopped short of directing how certain required tests should be performed. 186 W. Va. at 323, 412 S.E.2d at 517 (“[O]ur opinion today does not instruct the physicians how to perform the tests discussed, but instead, advises as to what tests must be performed in order for this Court to reach an informed decision on appeal.”).

    Through this appeal, however, it has become apparent that perhaps the Bilbrey decision should not have been so limited. As this case and Bilbrey exemplify, often two audiograms performed on the same patient will obtain significantly different results. Such differences may be the result of a variety of factors, not the least of which is the fact that an audiogram is a very subjective test. In Bilbrey, we opined that “if one audiogram shows a substantially worse four frequency total than a second audiogram, the expert must work with the premise that since a noise-induced loss is static, some other factor must be responsible for the difference between the two audiograms, such as a sinus or eustachian tube problem.” 186 W. Va. at 324, 412 S.E.2d at 18.

    Notwithstanding the Bilbrey Court's explanation that “experts have pointed out that if there is a fluctuation in the hearing loss between audiograms which is greater than the margin of error, then the audiogram which shows the least amount of hearing loss should be used to determine the hearing loss due to noise exposure,” 186 W. Va. at 323-24, 412 S.E.2d at 517-18 (emphasis added), the Court nevertheless went on to state, in dicta, that where the difference between two audiograms was greater than the margin of error, “the better audiogram of the two should be used as the audiogram most representative of the sensorineural loss.” Id. at 324, 412 S.E.2d at 518. Thus, by using the phrase “better audiogram” as opposed to a more specific phrase referring to “the audiogram which shows the least amount of hearing loss,” the Bilbrey Court appears to have rejected the expert view in favor of a more neutral procedure. Unfortunately, the Court failed to elaborate on just what it meant by the term “better.” It is also important to note that, while the Division, and to some degree the bar, have apparently interpreted Bilbrey's explanation of the expert view as a holding, the Bilbrey Court stopped short of creating any new principle of law regarding how to determine which of two widely varying, yet apparently valid, audiograms most accurately represents a claimant's true level of hearing loss. Therefore, we endeavor to do so here.

    Without minimizing the impact that the subjectivity of audiograms has on the results obtained, the evidence presently before this Court suggests that a variety of factors unrelated to the claimant's condition or cooperation with the testing process may also impact those results. Such factors may include the manner in which the testing is conducted, whether the speech discrimination portion of an audiogram is administered using a monitored live voice or a recorded voice, the methods used to calibrate the machinery, or perhaps even the type or brand of machinery used. We believe that establishing uniformity with regard to factors such as these would further the goals Bilbrey initially identified, namely obtaining consistent results in hearing loss cases and providing for a more meaningful review. Another problem that has come to light in this appeal is the absence of any established margin of error. Without a definitively set margin or error, there can be no assurance of consistency in the application of any rule pertaining to hearing loss claims that utilizes the margin of error.

    Additionally, we commented in Bilbrey that, because of the subjectivity of audiograms, the reliability and validity of the test results become important factors. 186 W. Va. at 323, 412 S.E.2d at 517 (“The audiogram is a subjective test, as it measures a subject's response to noise. Thus, the reliability of the test and the validity of the results are important factors.”). The parties to this case, and the experts providing evidence on their behalf, have indicated that while there are numerous methods for judging reliability and validity that are commonly used within the profession, there exist no specific laws or rules mandating their use, and there is no place provided on the workers' compensation hearing loss forms to indicate or ascertain their use.   (See footnote 13)  Requiring that specific reliability and validity measures be utilized and recorded, the audiograms would presumptively be more accurate, their reliability could be more effectively monitored by the Division, and the review process would be enhanced. Thus, for the reasons heretofore mentioned, we hold that the Workers' Compensation Division is directed to establish appropriate guidelines for the specific manner in which audiograms should be administered.   (See footnote 14)  In developing these guidelines, the Division should consider, inter alia: (1) whether all audiograms should be administered using a uniform brand and model of audiometer; (2) whether guidelines should be adopted for when and how audiometers should be uniformly calibrated; (3) establishing a definitive margin of error; (4) requiring audiologists to perform specific reliability and validity checks during the course of an audiogram; (5) modifying the existing WC-123HL form to allow for the reporting of any required reliability and validity checks; (6) whether the speech discrimination portion of all audiograms should be conducted using only a recorded voice; and (7) what method should be used to select an appropriate audiogram when two audiograms that are both rated “good” differ by more than the established margin of error.   (See footnote 15) 

Footnote: 1

    different audiograms with thresholds that vary no more than 10 dB, usually suggest unchanged hearing. However, these different audiograms, in the absence of any real change in hearing, may vary by as much as 10 dB at some frequencies, but not at all the frequencies tested, and not in the same

direction.

time such tests are needed to reach an informed decision. Brain stem audiometric testing should be performed only when there is suspicion of an acoustic neuroma.

(Emphasis added).

        Test/retest threshold reliability has been discussed and studied since the inception of the electronic pure tone test. It has been known since at least 1958 that certain methods of presenting the pure tones to the patient can affect the patient's responses. Specifically, it has been known since 1958 that comparing thresholds obtained using an ascending stimulus presentation, to thresholds obtained using a descending stimulus presentation can provide an internal reliability check.”

(Citation omitted). Dr. Harris further explained that

        Commonly, 1 kHz thresholds are obtained twice with the 1 kHz test threshold expected to be within 5 dB of the 1kHz retest threshold for the ear under test. . . .

    Providing a place on the workers' compensation audiogram form for recording the test/retest thresholds at 1 kHz will help insure that such is done.

        Having one threshold sought in ascending (soft to loud) trials and then the other sought with descending (loud to soft) trials, will enhance the sensitivity of this comparison in determining reliability (Woodford C, et al. A screening test for pseudohypacusis. The Hearing Review, No. 1977).

Dr. Harris suggests that this information could be elicited from the test audiologist by including a form similar to the following on the audiogram:

ascending 1 kHz threshold right ear __         descending 1 kHz threshold right ear __

        To facilitate the adoption of such rules and regulations for disability determinations, the Legislature authorized the Commissioner to create the Health Care Advisory Panel to assist with the “[e]stablish[ment of] protocols and procedures for the performance of examinations or evaluations performed by physicians or medical examiners[.]” W. Va. Code § 23-4- 3b(b) (1990) (Repl. Vol. 1998). Similarly, the Legislature established the Compensation Programs Performance Council, W. Va. Code § 21A-3-1 (1993) (Repl. Vol. 1996), [hereinafter

charged with the administration and enforcement of the workers' compensation statutory law of this State, pursuant to W. Va. Code § 23-1-1 (1997) (Repl. Vol.1998), should be accorded deference if such interpretations are consistent with the legislation's plain meaning and ordinary construction.”).