Comment: We received seven comments about the 3-month guidelines in proposed 4.00A, 4.00C2d, and 4.00D. Four of the commenters thought that this was a "blanket rule" which would require delay in the adjudication of a class of claims or unnecessary delay of many claims. Three of the commenters pointed out that the deferral would have no practical effect in many SSI claims because many SSI claimants will receive little or no treatment; therefore, there will be little or no additional evidence even if we were to wait 3 months. One commenter said that delay in some cases would be unwarranted because 3 months will have already elapsed before the date of adjudication. The same commenter pointed out that deferral would not be necessary in some of the most serious cases because it will be clear that the claimant is disabled. This commenter suggested that we revise the rules to provide that deferral of adjudication is never appropriate without a specific reason particular to the individual's case. One medical organization supported this guideline, stating that it was welcome to "those of us in the field of cardiac rehabilitation."

Response: We have not deleted the 3-month guidelines, but we have clarified the rules in response to the comments. We never intended to create a "class" of claimants who would be treated differently from other claimants. The requirement for a longitudinal record in cardiovascular cases is consistent with our rules for evaluating all types of impairments, not only cardiovascular impairments. We need a longitudinal record to establish the severity and duration of most impairments, especially those that may be amenable to treatment.

We provided a specific time guideline in the cardiovascular rules because many individuals with cardiovascular impairments have sudden onset of their impairments, have undergone surgery, or have impairments that will improve with treatment; in all three instances, they may recover in a relatively short time. We, therefore, must have some indication that the individual will be disabled for 12 months, and we believe that a 3-month recovery period is the minimum amount of time needed to demonstrate whether there will be a significant recovery. We also did not intend to have the 3-month guideline be an invariable rule and accordingly we used qualifying phrases, such as "usually" and "whenever there is such evidence," in the proposed rules.

However, in response to the comments we have clarified 4.00A and 104.00A by stating specifically that the reason we require 3 months of evidence is to establish a longitudinal picture of the individual's impairment(s), in terms of medical severity, functioning, and symptomatology. We also explain that, for the same reasons, a longitudinal record is important even when the individual has not received ongoing treatment. We have also added explicit statements to final 4.00A, 4.00C, and 4.00D in part A, and 104.00A and 104.00C in part B, that it is not necessary to defer issuing a determination or decision on a claim when the available evidence establishes that the claimant is disabled.

The fact that an individual may have no treating source does not mean that we cannot establish a longitudinal clinical record. If necessary, we may purchase a consultative examination for comparison with earlier evidence. However, in response to the comments, we have added a new third paragraph in final 4.00A to address this concern.

Finally, the commenter who noted that many claimants will have already established a 3-month history was correct. For exactly that reason, we do not believe that the 3-month guideline is especially onerous or that it will result in many delays in adjudication. The final rules do not say (nor did the NPRM) that adjudicators should defer claims for an additional 3 months, only that they should establish a 3-month record. Indeed, we expect that even when deferral is necessary, we will rarely have to defer the claim for a full 3 months because some time will necessarily have passed by the time of adjudication.

Comment: We received several comments about our use of the New York Heart Association (NYHA) functional criteria in proposed adult listings 4.02, 4.04, and 4.06. One commenter said that the criteria are vague and believed that we would misapply them to the detriment of claimants. Another commenter said that, in the majority of cases, treating sources do not use the NYHA classifications. Another commenter was concerned about how to use the classifications when the individual's impairment fluctuates, saying that it is unrealistic to presume that some patients will remain in one classification. Two commenters were concerned about how we would develop evidence and weigh treating source opinions about the classification.

One commenter said that the criteria were in a publication that has long been out-of-print. This commenter, and several others, thought that, at a minimum, we should include the definitions of the NYHA classification levels in the listings. One commenter, however, pointed out that if the NYHA alters its classification standards in the future, it may become necessary for us to revise the listings that employ those standards.

Finally, two commenters were concerned that the proposed rules failed to include the NYHA "therapeutic classifications." One of these commenters said that the final regulations must at least recognize the concept behind the therapeutic classifications, which is that some individuals must restrict their activities for reasons of medical safety, even though they are technically capable of performing more exercise than is safe or medically appropriate.

Response: We adopted the comments by removing the references to the NYHA functional classifications from final listings 4.02 and 4.04 and replacing them with explicit, descriptive functional criteria based on the NYHA criteria, and by removing all functional criteria from final listing 4.06. In this way, we have avoided the pitfalls pointed out by the commenters, although we do not believe that our proposed use of the classifications was as problematic as the commenters did.

Cardiologists have used the NYHA functional classifications for many years, and the criteria are well understood by people in the profession; this undoubtedly accounts for the fact that we did not receive any comments from physicians or medical organizations about our proposal to use the criteria. Nevertheless, we agree with the commenters that there are a number of real problems that could arise in an adjudicatory context and, on balance, believe that the most straightforward solution is to simply state exactly what we require in the listings themselves.

As we have explained in the first section of this preamble, the deletion of the references to the NYHA functional criteria also necessitated a reorganization of the rules in final listings 4.02 and 4.04. This is not a substantive change; we merely reorganized the criteria of the proposed rules to place the medical criteria describing the listed medically determinable impairments first—as in all other listings—followed by the explicit functional criteria. This organization is also consistent with the organization of the mental listings in 12.00 and 112.00, as well as several other listings that include disabling functional consequences among their criteria.

We decided to delete the proposed functional criteria from final listing 4.06 for several reasons, chiefly because they were unnecessary. Individuals whose impairments demonstrate the findings in final listings 4.06A, 4.06B, and 4.06E have impairments that will cause at least NYHA functional class III limitations. Individuals whose impairments satisfy the criteria in the reference listings, final listings 4.06C and 4.06D, already have impairments that meet the requirements of other listings. Therefore, the proposed functional requirement for listing 4.06 was at best redundant. Also, the proposed requirement for functional class III "or class IV" was unnecessary for logical reasons: Since functional class III would have been sufficient to establish disability under proposed listing 4.06, a person who had functional limitations equivalent to class IV (i.e., symptoms at rest) would have had an impairment that was more severe than was necessary to meet the requirements of this listing. Finally, some of the provisions of final listing 4.06 are the same as provisions in final childhood listing 104.06, which does not include functional criteria; deletion of the functional criteria in the adult listing makes the corresponding part A and part B listings consistent with each other.

We did not include the NYHA therapeutic classifications in the listings, or explicit criteria based on the therapeutic classifications, because we believe that such considerations are generally more suitable to the assessment of residual functional capacity. However, we want to assure the commenters that the principle illustrated by the therapeutic classifications is a basic principle in all of our disability evaluations. It has long been our policy that valid medical reasons for limiting an individual's level of exertion (or any functioning, including mental functioning) can be sufficient to establish that the individual is in fact limited in that capacity; the individual need not demonstrate this limitation by putting himself or herself at risk in order to demonstrate that there is a limitation.

As to the comment about individuals whose functional status changes and who, thus, might properly be classified in more than one of the NYHA classes at different times, it is our policy that an individual need not be continuously limited to the degree specified in order to be found disabled. This is one of the reasons we repeatedly stress the need for longitudinal evidence, which will permit our adjudicators to draw conclusions about how the individual generally functions, over time. Practically speaking, the fact that an individual's ability to function may briefly improve, only to worsen again, is immaterial to the determination that the individual is unable to work.

Although we have deleted all references to the NYHA classifications as listings criteria, we have retained one reference in the preface to the listings. This is in final 4.00C2c of the preface, in which we list NYHA class IV heart failure as a risk factor militating against the purchase of an exercise test. We believe that in this narrow context the use of the term is sufficiently clear.

Comment: Ten commenters strongly opposed our proposals in the NPRM for using exercise testing. In particular, the commenters opposed proposed 4.00D2, which would have required treadmill exercise testing in all cases in which such testing was not medically contraindicated, and proposed in 4.00C2a, which stated that valid exercise tests, although not the exclusive means for assessing the severity of heart disease, should be the "primary basis" for evaluating functional capacity under the listing. One commenter said that the proposed rules would result in exercise testing that was unnecessary, expensive, and dangerous. One commenter said that our statement in proposed 4.00C2a that the utility of exercise testing is "well recognized" would further improperly underscore the importance of exercise testing.

Response: We adopted the comments asking us to delete the reference to the primacy of exercise testing in 4.00C2a and all of proposed 4.00D2. As we have explained above in the explanation of the final rules, final 4.00C2a is now confined to a discussion of when to purchase exercise testing, and limits the circumstances in which purchase will be appropriate.

The commenters offered many arguments for removing the above-mentioned references in proposed 4.00C2a and 4.00D2. Because we have adopted the comments, we have not summarized all of the comments and arguments here. Most of those who commented on this aspect of the final rule pointed out that our statement in proposed 4.00C2c could have been misunderstood, as indeed it was by most of these commenters. The commenters assumed that our statement that treadmill exercise testing would be the "primary basis for evaluating functional capacity under this listing" meant that treadmill testing would be the primary basis for assessing functioning at all steps of the sequential evaluation process. The plain language of the proposed rule, of course, did not say this, but was confined to a discussion of whether a claim could be allowed at the listings step. Nevertheless, the comments do demonstrate that there were misunderstandings, especially because we also proposed to require the purchase of exercise testing in most cases.

We were also persuaded by the arguments that there will be many cases in which it will be evident that the claimant is disabled, obviating the need for—and expense of—exercise testing. We also agreed with the commenter who pointed out that, even though we require consideration of exercise testing in the context of the entire record, the use of the word "primary" in the proposed rules could have been interpreted as an instruction to our adjudicators to automatically give more weight to the results of such testing than to any other evidence; such an instruction would be contrary to our policy.

In related comments, several commenters suggested that we emphasize in the final rules that exercise testing is merely one tool that physicians use to evaluate the severity of cardiac disease, and that the proper evaluation of a claim for Social Security or SSI benefits based on disability requires consideration of the test results in the context of all the relevant evidence. As we have explained above in the explanation of the final rules, we have adopted the comment by adding a statement to this effect in 4.00A and 4.00C2e(1) of the final rules. Final 4.00A emphasizes the need to consider all of the evidence; the fact that no claim may be denied solely because the claimant's impairment(s) does not meet or equal in severity the requirements of a listing; and the need to provide an individualized assessment in every case. Final 4.00C2e(1) emphasizes the need to consider exercise test results in the context of all of the relevant evidence. We did not add similar statements in 4.00C2a because we revised it to change its focus to the purchase of exercise testing, and there was no longer a context for such a discussion in that section.

We did not agree with the commenter who said that our purchase of exercise tests would be dangerous. We have many safeguards to ensure proper review by an expert physician before we will purchase such testing. We have also provided a detailed methodology for performing the tests safely and in approved facilities, including a requirement for supervision of the test by a physician.

We also did not agree with the comment about our statement of the utility of exercise testing in the first sentence of proposed 4.00C2a; although, we did revise the statement. We agreed with those commenters who referred to the information supplied by the American Medical Association and the American College of Cardiology in the joint amicus curiae brief these organizations filed with the U.S. Court of Appeals in New York in the case of the State of New York v. Sullivan, 906 F.2d 910 (2d Cir. 1990), that treadmill exercise testing does not assess all types of functioning or functioning in different environments. We, therefore, revised 4.00C2a to state more accurately that it is well recognized by medical experts that exercise testing is the best tool currently available for estimating aerobic capacity.

Comment: Four commenters, quoting or paraphrasing the district court's decision in 1989 in the State of New York case, stated that the exercise test is not a valuable tool for assessing disability because the exercise test results in a misdiagnosis of ischemic heart disease more than one-third of the time. They also cited the part of the court's decision that noted that, in certain circumstances, other widely used procedures (e.g., the exercise thallium test, the equilibrium radionuclide angiogram, and angiography) are more reliable in measuring the severity of ischemic heart disease.

Three of these commenters, in identical or similar language, also asserted that the proposed listings authorized adjudicators to deny benefits based on the results of suboptimal exercise tests. Another commenter stated that many exercise tests purchased on a consultative basis are terminated due to detoning or hyperventilation before cardiovascular challenge can occur, thus rendering the exercise test useless for evaluation of cardiac functioning.

Response: We do not use treadmill exercise tests to diagnose the existence of impairments; we use them to estimate the aerobic capacity of individuals and, hence, as a test of a certain kind of functional capacity. Indeed, in the preamble in the NPRM, we stated that we proposed to remove the requirement for a targeted heart rate because "the proposed listings do not use the exercise test for diagnosis but rather for functional evaluation." We made this statement in part because we agree with the commenters that the failure to demonstrate ischemia on an ECG does not by itself establish that the individual's heart is normal; however, it is evidence of the exercise test level that can be safely performed without triggering ischemic ST changes and, hence, is some evidence of aerobic capacity. Even when an exercise test is negative, it still provides some information about the individual's ability to exercise to the level achieved on the test.

Although we agree with the commenters' opinion that the exercise thallium test is the best method for diagnosing the existence of ischemic heart disease, we repeat that the issue in these listings is not one of diagnosis but of functioning. We also agree that the other tests may be useful for evaluating the severity of ischemic heart disease in some circumstances; however, they are not generally more reliable than the exercise test for this purpose. For this reason, we do not agree that they belong in our listings as alternative criteria. The listings are only examples. We do not attempt to state every possible fact pattern in the listings; that is the province of our findings that claimants may have an impairment that is equal in severity to a listed impairment.

In response to the three commenters who asserted that the proposed rules would permit adjudicators to deny benefits based on suboptimal exercise tests, we have stated above that the failure to meet or equal the requirements of a listing is never a basis for denying a claim. We can find nothing in the proposed rules, or in the preamble to the proposed rules, that could have led the commenters to draw this conclusion. The results of suboptimal exercise testing are only one piece of evidence that we must consider in the context of the entire record.

Comment: Another commenter was concerned that the proposed rules would "no longer prohibit adjudicators from relying upon suboptimal exercise tests in determining that the listing for ischemic heart disease is not met." The commenter asserted that the use of suboptimal tests would allow adjudicators to conclude "erroneously" that a listing was not met, and that our statement in the preamble that suboptimal tests are a useful measure of function (56 FR 31267) is contradicted by listing 4.04, which does not permit adjudicators to consider the test as a sign of the claimant's limitations. The commenter said that the listings should be changed to specifically allow adjudicators to consider the inability to complete a test as a ground for establishing that an impairment is of listing-level severity, as is done in listing 4.02, which recognizes that markedly symptomatic exercise intolerance may meet that listing. Alternatively, the commenter recommended that the listing should explicitly state that the inability to complete a test is not evidence that a person is free from ischemic heart disease. A similar comment added that our assurance that those cases that do not meet the listings could be allowed under the rules for equivalence was not persuasive because the commenter had seen many cases in which this had not occurred.

Response: We believe that the commenter who made the first comment misunderstood our prior rules and the statement we made in the preamble to the proposed rules about why we deleted the requirement for attainment of a targeted heart rate. An impairment meets a listing in only one way: by matching exactly the criteria in that listing. Because the results of suboptimal exercise tests have never been criteria in our listings, they cannot be—and never could have been—a per se reason for finding whether a person's impairment met or did not meet a listing. Thus, the listings did not "prohibit" us from using the results of suboptimal exercise tests to find that an impairment did not meet a listing. The only question in any "meets" finding is whether the person has presented evidence that satisfies the specific criteria of the listing.

As we have explained in the preceding response, the failure of a suboptimal test to diagnose coronary disease does not mean that the test results are without use; and it is not a "false negative" in the scientific sense, because we are not using the exercise test for diagnosis. This is not a new policy. We have always required our adjudicators to consider at all steps in the sequential evaluation process all relevant evidence of an individual's functioning, which includes failure to complete an exercise test.

Nor does listing 4.04 prohibit adjudicators from using the results of suboptimal or negative exercise tests as evidence of a person's aerobic functional capacity. It is faulty reasoning to conclude that we do not consider the implications of a person's ability to function on a suboptimal or negative exercise test because we do not include a criterion within the listing for a test that is not "positive." All this means is that a claimant with a suboptimal or negative exercise test cannot have, on the basis of that test, an impairment that "meets" the requirements of those listings that include exercise tests among their criteria.

We did not adopt the suggestion to add a criterion to final listing 4.04 which would permit a finding of "meets" the listing based on markedly symptomatic exercise intolerance. The commenter's comparison with listing 4.02 was inapt. In proposed listing 4.02B, the requirement for "markedly symptomatic exercise intolerance" (the language of which we have now deleted in our clarification of the final rule) had to be associated with objective, demonstrable chronic heart failure, with cardiac enlargement or ventricular dysfunction. In final listing 4.04A, the listing for angina demonstrated by exercise testing, the objective requirement is the ECG abnormality or inability to maintain adequate blood pressure on exercise to support the inference that symptoms alleged at a low level of exercise may be attributed to ischemia; if we were to replace it with an allegation of exercise intolerance, the listing would not have any objective criteria. Again, this does not mean that an individual who is actually limited by his or her ischemic heart disease despite inability to "meet" this listing cannot be found disabled. The person may present evidence demonstrating equivalence to a listing, or by establishing a residual functional capacity which prevents him or her from working at past work and other work. Also, exercise intolerance may be indicative of a noncardiac condition that may alone or in conjunction with a cardiac impairment render a person disabled.

Nevertheless, we find reasonable and helpful the commenter's suggestion that we add a statement to the listing that the ability or inability to complete an exercise test is not evidence that a person is free from ischemic heart disease. We believe, however, that the statement is more appropriate in 4.00C2e(1) of the preface, where we discuss evaluation of the results of exercise testing. We have, therefore, added the statement as a new second sentence in final 4.00C2e.

Comment: Four commenters referred to the amicus curiae brief filed in the case of State of New York v. Sullivan by the American Medical Association and the American College of Cardiology, in which it was argued that the treadmill exercise test is not a fair measure of stress from heat, cold, humidity, pollution, altitude, psychological pressures, or other work-related stresses in the workplace. The brief also pointed out that in the workplace a person may, for example, be required to perform tasks that require substantial use of the arms, isometric exertion and physical effort that is sudden or prolonged, and these things are not measured by the treadmill exercise test. In identical language, two of the commenters said that the proposed rules demonstrated that we "continue to ignore this medical reality."

Response: These final rules are consistent with the facts presented in the amicus curiae brief. As the commenters noted in their quotes from the brief, the argument presented by the two medical associations discussed the efficacy of the treadmill stress test as the sole or primary basis for establishing a person's ability to work, not the propriety of our use of the treadmill exercise test in our listings. Indeed, we received comments on the NPRM from both medical organizations, and neither of them objected to the proposed rules in this regard—in fact, no physician advised us that he or she objected to this aspect of the proposed rules.

However, as we have explained above, we have revised the final rules to lessen further the likelihood that they can be misinterpreted to mean that the treadmill exercise test is the "primary" basis for assessing "functioning" throughout the sequential evaluation process. We have also revised the final rules to underscore our longstanding policy that when an individual's impairment does not meet a listing, all relevant evidence must be considered to determine whether the individual is disabled. It is in the residual functional capacity assessment that factors are considered such as the effect of heat, cold, humidity, pollution, altitude, psychological pressures, and other work-related stresses.

Comment: Four commenters also opposed the rule that listing 4.04, for cardiac ischemia, could not be "met" without the results of exercise testing unless the individual was at risk for exercise. The commenters said that this rule prevents consideration of other test results, such as echocardiograms and angiograms, under the listings. One of the commenters said that the listings must state that the criteria in listings 4.04B and C (the listings that include echocardiography and angiography among their criteria) can be met even if the claimant can exercise on a treadmill or even if an exercise treadmill test is negative. The same commenter said that 4.00A "instructs adjudicators to disregard" evidence of other tests, the opinions of treating physicians and the claimant's pain, and that the listings do not permit signs or symptoms (such as pain) to be considered as the basis for proving that an impairment actually meets a listing.

Response: We did not adopt the comments, except to the extent that we included language in the new fourth paragraph of final 4.00A which reminds adjudicators that the listings are only examples of common cardiovascular disorders that are disabling. The listings are intended to be a screening device by which we can identify and allow claims filed by the most obviously disabled individuals; they are not an all-inclusive list of disabilities under which all individuals must be found disabled. Even though echocardiography can demonstrate abnormality of cardiac functioning, and coronary angiography is the "gold standard" for ascertaining the presence of coronary atherosclerosis, neither of these tests provides information about whether any discovered disease is directly related to or predictive of functional status in the way that a positive exercise test does.

This does not mean, however, that individuals who are unable to exercise for the requisite time on a treadmill, or individuals who have negative exercise tests will be found not disabled. All it means is that we cannot find that they have impairments that "meet" the listings. We believe that our revisions to the preface have clarified that such individuals will always receive the benefit of an individualized assessment that considers all of the evidence—including the results of any other tests they may have undergone, their symptoms, and their treating sources' opinions.

With respect to the last comment, we could not find any text in the NPRM in which we instructed our adjudicators to disregard any evidence, nor would we have intentionally written such a rule because it would have been contrary to our policy. Indeed, subsequent to the publication of the NPRM, we published final rules that explicitly address the importance of the role of pain and treating source opinions in our adjudications (see "Standards for Consultative Examinations and Existing Medical Evidence," 56 FR 36932, August 1, 1991; and "Evaluation of Symptoms, Including Pain," 56 FR 57928, November 14, 1991). Again, if the evidence of record, including evidence from a treating physician, and including evidence of pain or other symptoms, establishes that an individual is disabled, the fact that the evidence fails to show that a listing is met or equaled will not prevent a favorable determination. Moreover, listing 4.04 is a listing for cardiac ischemia, which ordinarily describes a kind of chest pain; we do not even consider using this listing unless the individual has the symptom of "chest discomfort," as we state in the opening sentence of the listing.

Comment: One commenter criticized the example of equivalence we had provided in the preamble to the proposed rules (56 FR 31267). The commenter also said that the listings as proposed did not set forth any substantive method for considering nonexercise test evidence in determining equivalence, and the general language in the preface about equaling any listing was insufficient.

Response: We disagree. We did not include the example in the final rules because it is virtually impossible to devise a single example, or even two or three examples, that would illustrate such complex concepts as, for instance, equivalence. Moreover, any example we devised would have to be so clear and unambiguous that it would not provide useful guidance; indeed, this could be a valid criticism of the example we provided in the preamble to the NPRM.

We do not provide methods for determining equivalence in any of the listings sections in part A or part B. We have separate rules in §§ 404.1526, 416.926, and 416.926a for that purpose.

Comment: In addition to those commenters who mentioned that an evaluation of pain should be included as part of the listings, two commenters thought that the proposed rules allowed exercise testing to be used to overcome an individual's allegations of pain. One commenter said that lack of adequate consideration of the limitations imposed by pain associated with ischemia will result in unrealistic functional assessments. The commenter referred to the decision in Avery v. Secretary of HHS, 797 F.2d 19 (1st Cir., 1987), which requires an individualized assessment of pain in the event of a medically determinable impairment that could cause pain.

Response: We did not intend to give the impression that a single exercise test in and of itself would be sufficient evidence to overcome an allegation of pain. As noted above, in November 1991, we published final rules regarding the evaluation of pain (56 FR 57928). Among other things, these rules require consideration of all relevant evidence, including a number of specific factors that must be considered whenever we consider an allegation of a symptom, such as pain. Our rules also require an individualized assessment of each person's allegations of pain, consistent with the Avery decision.

Comment: Several commenters raised issues of treating physician opinion and its relationship to the rules on exercise testing. A number of the commenters questioned the proposed rule in 4.00C2c that permitted a program physician to override a treating physician's opinion as to the advisability of conducting an exercise test under some circumstances. Most stated that we should never permit a program physician to require a claimant to submit to an exercise test against the recommendation of a treating physician. Several commenters thought that the program physician's decision to override a treating source's opinion would be limited to a review of the medical evidence in the claim file without giving proper weight to the treating source's opinion. Three commenters, citing the proposed rule in 4.00C2f (final 4.00C2e(1)) that required us to obtain statements from treating sources before we would decide whether to override their opinions, felt that the rule would create professional relations problems which will be magnified when the treating physician's opinion is overridden. One person suggested that the rules should require the program physician to ask the treating physician his or her opinion on whether the exercise test should be performed and to ask whether an exercise test had been performed in the past year. One commenter felt this policy could interfere with patient/doctor relationships. Another questioned if a statement should be made in the regulations that the facility performing the test should decide whether the test is contraindicated prior to its performance. Three commenters wanted to know what we would do if there is no treating source.

One commenter asked whether our use of the term "examining source" meant that we would contact every physician who had examined the claimant, even if the physician was not a treating physician.

Several commenters stated that we were not giving proper weight to the opinion of treating physicians because proposed 4.00D2 required the purchase of an exercise test if a timely exercise test was not available and posed no significant risk to the individual.

Response: In response to the comments, we have made a number of revisions in the preface; however, we did not adopt the comments that asked us to prohibit program physicians from ever overriding the opinions of treating sources. In 4.00C2c of the NPRM we stated that we "will generally give great weight to the treating physicians' opinions and will generally not override them." We also characterized as "limited" the situations in which we would override such an opinion and require a written statement of explanation whenever a source's opinion was overridden. Clearly, therefore, we did not expect program physicians to override the opinions of treating sources except in the rarest of cases.

To make clear what those cases are, and in response to the comments, we first deleted proposed 4.00D2, which required exercise tests in all cases in which such testing posed no significant risk to the individual, thereby sharply limiting the number of cases in which the issue of whether to purchase an exercise test will even arise. Second, we revised final 4.00C2a to provide guidance for when purchase of an exercise test may be appropriate: when there is a question whether an impairment meets or is equivalent in severity to one of the listings, or when there is insufficient evidence in the record to evaluate aerobic capacity, and the claim cannot otherwise be favorably decided. Third, we revised 4.00C2c to state plainly that it will be a rare case in which the program physician would override the opinion of a treating source with respect to the purchase of an exercise test. If the treating source's opinion is overridden, the program physician must document the file with a rationale setting forth his or her reasons for doing so. We also expanded final 4.00C2c to indicate that an exercise test will only be purchased if it poses no significant risk to the individual and the claim cannot otherwise be favorably decided. Fourth, we revised final 4.00C2e(1) to state even more clearly the rule from the NPRM that, if purchase of an exercise test is contemplated, and the claimant has a treating physician who has not ordered an exercise test, and there is no apparent reason why the person would be at risk, contact must be made with the source to determine the source's opinion whether exercise testing poses a significant risk, or whether there is some other reason why the treating physician did not have the test performed, such as because of a financial bar.

We do not believe that there will be significant professional relations problems from these rules. We believe that physicians will understand that we are contacting them out of concern for their patients' well-being, inasmuch as we will have to contact them when we need to purchase an exercise test and it is not clear why they have not ordered one, or why they think an exercise test poses a significant risk. Also, again, it should be a rare instance when the treating physician's opinion will be overridden.

It has always been the responsibility of the provider performing the exercise test to determine whether there is any risk to the claimant. Also, if there is no treating source, the program physician will be responsible for assessing the risk of exercise to the claimant, as has always been the case in the past. Neither of these policies is new, nor are they unique to the cardiovascular listings. However, in response to the comment that asked whether all "examining" but nontreating sources should be contacted, we have revised the section by deleting the phrase "examining physician." Under the aforementioned "Standards for Consultative Examinations and Existing Medical Evidence," we no longer use the term "examining physician," and our rules requiring recontact with physicians are clearly set out. Whether it will be necessary to contact other sources who may have examined the claimant but who are not treating sources will vary from case to case.

Comment: With regard to general policy on treating source opinions, several commenters said that the rules demonstrated that we do not give appropriate weight to such opinions. One commenter said that we did not provide any instruction in the listings on how to evaluate treating physician opinions about the nature and severity of the claimant's impairment, or which demonstrate that particular exercise test results are questionable, or provide some other basis for establishing disability. Another commenter said that the final rule must make clear that the results of an exercise test do not constitute substantial evidence to override a treating physician's opinion on the severity of the claimant's impairment. One commenter from a State located within the Second Circuit said that an exercise test should not be purchased if the treating physician can determine disability without it. The commenter said that, by purchasing an exercise test, we would not be according proper weight to the opinion of the claimant's treating physician if the treating physician has determined that the test is not necessary to evaluate disability.

Response: We added the clause, "in addition to information about the nature and severity of the impairment," to the end of the second sentence of the second paragraph of final 4.00A (which was the third sentence in proposed 4.00A). This revision reflects §§ 404.1527 and 416.927 of our rules. Otherwise, we did not adopt the comments. We would not provide rules for evaluating treating source opinions in this listing or any other because there are separate rules (§§ 404.1527 and 416.927) for considering the opinions of treating physicians. (See 56 FR 36932, "Standards for Consultative Examinations and Existing Medical Evidence," August 1, 1991.) The rules apply to all impairments.

When a treating physician provides us with evidence that demonstrates that the results of a particular exercise test are questionable, or a treating source provides evidence upon which we could base an allowance, our general rules provide that we will have to determine whether there is a conflict in the evidence. If there is a conflict in the evidence, we are required to resolve it, giving more weight to the treating source's opinion than we would if the opinion had come from a source who was not a treating physician (see §§ 404.1527(d) and 416.927(d)). If necessary to resolve the conflict, we may recontact the treating physician for additional information. Of course, we may be able to resolve the conflict based on the existing evidence, such as when we agree with the treating source that the results of a test are questionable or that the claimant is disabled. Our rules also explain how we decide whether there is substantial evidence to overcome a treating source's opinion. If the treating physician's opinion that a claimant is disabled is well supported by the evidence of record and is not contradicted by the other substantial evidence, we will give it controlling weight.

In response to the last comment, the Act provides that we must determine whether a claimant is disabled, not the treating physician. We are not aware of any decision of the Second Circuit, or any other Federal court, that holds that a treating source may make the determination of disability or decide the sufficiency of the evidence for SSA under the Act.

Comment: Several attorneys opposed the proposed rules, asserting that they were in conflict with the Social Security Act and the decision in the State of New York case in which the United States Court of Appeals for the Second Circuit ruled that all relevant evidence must be considered in determining disability and that the results of exercise tests cannot be used to the exclusion of other evidence to deny disability benefits. The commenters offered a number of arguments addressing the proposed policies both in general and as they pertained specifically to the "meets," "equals," and residual functional capacity aspects of our decisions. Two of the commenters cited the United States Supreme Court's decision in Sullivan v. Zebley, 493 U.S. 521 (1990), in support of the notion that all disability determinations must be based on an individualized consideration of each claim and reflect all relevant medical evidence.

Response: We believe that the revisions we have already described address most of the commenters' concerns, and that these rules are not contrary to the Second Circuit's holding in State of New York. We also believe that we have the statutory authority under section 205(a) of the Act (made applicable to title XVI by section 1631(d)(1)) to issue these rules, which have national applicability and which the Second Circuit (or any other court) has not reviewed or found to be invalid.

We also want to emphasize that we consider all of the medical evidence of record at step 3 of the sequential evaluation process, i.e., the listings step. We do this to be certain that the evidence is consistent, to resolve any apparent conflicts in the record, and, if we ultimately rely on a treadmill exercise test to find a claimant disabled under the listings, to be certain that it is a valid test and not a "false-positive" test. Our policy does not permit a finding of "meets" or "equals" in the face of a record that casts doubt on the validity of the test results; plainly, we must consider the entire record to determine whether the test raises such a doubt. Moreover, at the equivalence stage of step 3, evidence that may not play a role in a finding of "meets" may be decisive in our determination that the claimant is disabled; and, as we have repeatedly stated throughout this preamble, if a claimant is unable to establish that he or she has an impairment that meets or equals the severity of a listing, we will continue the sequential evaluation to determine whether, based on all the evidence of record, the claimant is disabled.

Beyond that, we believe that the Supreme Court ruled in Zebley that it is not possible to provide individualized assessments to all claimants under the listings, and that the listings could never be revised to adequately provide for such assessments. For example, in its detailed analysis of the listings as a sole test of disability for children claiming benefits under the SSI program, the Court stated that "the listings obviously do not cover all illnesses and abnormalities that actually can be disabling," that "the listings * * * exclude any claimant whose impairment would not prevent any and all persons from doing any kind of work, but which actually precludes the particular claimant from working, given its actual effects on him—such as pain, consequences of medication, and other symptoms that vary greatly with the individual," that "the equivalence analysis excludes" many claimants, and that "there are several obvious categories of claimants who would not qualify under the listings but who nonetheless would meet the statutory standard." 493 U.S. at 533-34. Even more to the point, the Court stated:

Even if the listings were set at the same level of severity as the statute, and expanded to cover many more * * * impairments, no set of listings could ensure that * * * claimants would receive benefits * * * under the individualized, functional analysis contemplated by the statute * * *. No decision process restricted to comparing claimants' medical evidence to a fixed, finite set of medical criteria can respond adequately to the infinite variety of medical conditions and combinations thereof, the varying impact of such conditions due to the claimant's individual characteristics and the constant evolution of medical diagnostic techniques.

Id. at 539, emphasis in original. The Court's holding in Zebley, therefore, was to require an additional step beyond the listings step, at which we would be required to satisfy the statutory criterion in title XVI of the Act of providing an individualized analysis for children based on all the evidence on a case-by-case basis; it was not that we must expand the listings to provide individualized assessments at the listings step. Therefore, we believe that, far from supporting the commenters' contentions, the Zebley decision refutes them.

Again, all claimants who are not found disabled at the listings step receive the kind of individualized assessment which considers the effects of pain, the side effects of medication, and the myriad other factors that describe their particular responses to their particular impairments, as required by the law. Much of the new text we have added to the prefaces to Parts A and B of these final rules is intended to underscore these important, longstanding policy principles.

Moreover, the listings step, which is only an allowance mechanism, is not required by the Social Security Act. If we did not have a listings step or if we were precluded from using our listings, it would only work to the detriment of claimants in terms of delaying some allowances, not to their advantage.

Information in the Preamble

Comment: One commenter expressed the view that in a number of instances, and especially in the childhood listings, we had proposed significant changes to our rules that were not mentioned or explained in the preamble. The commenter suggested that this would preclude the public from having a meaningful opportunity to comment on proposed rule changes.

Response: We believe the NPRM did give the public a meaningful opportunity to comment on the proposed rules, and that the detailed nature of the comments we received, which are discussed in this preamble to our final rules, attests to this fact.

The New Listings Will Result in More Denials

Comment: Three commenters thought that the proposed listings would result in our denying more claims for benefits. One commenter said that the new listings would result in denials of benefits to "tens of thousands of seriously impaired adult cardiac disease claimants." The other two commenters said that the proposed listings would result in more denials by the State agencies, and that this would result in more requests for hearings before an administrative law judge. One of these commenters also thought that the proposed listings provided less discretion to the State agencies than the former listings did.

Response: We use the listings only to find people disabled. Failure to have an impairment that meets or equals in severity a listing is never a basis for denying any claim, whether the claim is filed for an adult or a child.

However, it is true that some people whose impairments would have met the criteria of our prior listings will not be found to have impairments that meet these final listings. It is also true that some of these people will not be found disabled. For example, under 4.04B of the prior listings, individuals with chest pain of cardiac origin and an ejection fraction of 30 percent had impairments that met the listing. We now know that an ejection fraction of 30 percent does not reliably establish the severity or functional effects of an impairment; indeed, many people with ejection fractions of 30 percent function normally.

It is, of course, not true that the listings will result in the denial of individuals who are seriously impaired, nor was that our intent. Our only aim in these revisions was to fairly and accurately update the listings consistent with current medical knowledge.

Nevertheless, in response to these and several other comments discussed below, we have revised 4.00A and 104.00A to clarify that the guidance in the prefaces to the listings is primarily for the purpose of deciding whether an impairment meets one of the listings. We have also added new paragraphs to 4.00A and 104.00A (already described above, in the summary of provisions) which remind adjudicators that individuals whose impairments do not meet any listing may still have impairments that are equivalent in severity to a listed impairment, or that may result in a finding of disability as a result of the adult or child sequential evaluation processes based on an individualized assessment of their functioning. We have also added language in several places in the prefaces to state that all evidence in the case must be considered, and that, if the evidence already obtained establishes a favorable decision, the adjudicator should not continue development just to determine whether an impairment meets a listing.

Finally, we do not agree that the rules provide less discretion to the State agencies than the former listings. As the former listings were based almost exclusively on objective, diagnostic medical findings, the revised listings which provide functional criteria as well as medical criteria arguably provide the decisionmaker with more discretion.

Comment: A number of commenters were concerned about the statement in the NPRM that "most individuals usually have received the benefit of a medically prescribed program of progressive physical activity accompanied by relevant educational and psychological support." They said that many of the people who file for title II and SSI disability benefits are outside the mainstream of health care and cannot afford such treatment or do not live where it is available. Three commenters thought that we would deny claims in which the claimants did not undergo treatment; one of these commenters said that the listings indicated that a claimant who did not undergo treatment should be denied. Two commenters were concerned that the rules were related to the rules on failure to follow prescribed treatment. One medical association suggested that the language in 4.00A should be, "Whenever there is such evidence, the clinical record must include a description of the therapy and positive evidence of rehabilitation measures prescribed by the treating source and response." One attorney commenter pointed out that listing 4.04 requires a regimen of prescribed treatment as a prerequisite under that listing, so that a claimant whose impairment satisfies the criteria in listing 4.04A but does not receive treatment still could not be found disabled under that listing.

Response: We adopted the comments by clarifying the final rules. In the second paragraph of final 4.00A, we now indicate that "many" individuals who have "listing-level" impairments will have received the benefit of such treatment. Although we agree that people with impairments of lesser severity than those in the listings may not necessarily receive this kind of therapy, we believe that the listing-level impairments are so severe that many, if not most, individuals with such serious—and in the case of cardiovascular impairments, potentially life-threatening—impairments will be placed on some sort of rehabilitative and supportive program.

However, we also state that, "[w]henever there is such evidence . . ., [i]t is important to document any prescribed therapy and response because this medical management may have improved the individual's functional status," recognizing that there will be cases in which no such treatment was prescribed or instituted. This does not mean that we will find individuals who have not received treatment to be not disabled, only that we cannot find them to have impairments that "meet" the requirements of those listings that require a regimen of therapy, as noted by the last commenter. For instance, an individual whose impairment meets the requirements of final listing 4.04A but does not have any treatment available to him or her, and who can establish duration of 12 months (prospectively or retrospectively) could be found to have an impairment that is equivalent in severity to the listing.

To underscore our policy that we never deny claims because of failure to meet the listings, we added new paragraphs in final 4.00A and 104.00A which stress the necessity for making an equivalence determination when a claimant's impairment(s) does not meet a listing, and for assessing residual functional capacity (or performing an individualized functional assessment) when a claimant's severe impairment(s) neither meets nor equals in severity any listing. We have also added reminders in final 4.00A and 104.00A to consider all relevant evidence whenever there is no evidence of therapy.

Finally, the rules on failure to follow prescribed treatment are quite complex, but have one simple underpinning: They do not come into play unless an individual's treating source has prescribed treatment for the individual which the individual is not following. If the individual's treating source has not prescribed treatment, or the individual does not have a treating source, the principle does not apply.

Comment: Two legal services organizations commented on the statement in 4.00A of the NPRM that individuals who are "deconditioned" may do poorly on exercise testing. The commenters thought that this meant our adjudicators would disregard deconditioning in the determination of disability. One commenter said that the section should instruct adjudicators to consider the effects of deconditioning as a basis for a finding of disability because it is a consequence of a cardiovascular impairment or a side effect of treatment (e.g., bedrest). The same commenter also said that we should include deconditioning as a listing criterion in listing 4.04.

Response: We did not adopt the specific comments, but we have clarified the rules. By "deconditioned," we meant the reversible exercise intolerance that comes from a lack of activity, such as from prolonged hospitalization, but which resolves with therapy. Thus, one of our reasons for requiring a longitudinal clinical record of at least 3 months was to allow time for the individual to become reconditioned; in this way our evaluation would consider the individual in the most stable condition and provide a more accurate picture of expected long-term functioning. We have, therefore, revised the proposed language (now in the second paragraph of final 4.00A) to clarify our original intent that inactivity or bedrest may result in a reversible deconditioned state, and that individuals in such a deconditioned state may do poorly on exercise testing if the testing is performed before they have been reconditioned.

We do not agree that the factor of deconditioning should be added to listing 4.04. Listing 4.04 already includes functional criteria that result from the listed impairment.

Comment: A national medical association noted that we had included arterial desaturation as one of the consequences of heart disease in 4.00B4. They noted, however, that we had not provided a methodology for quantifying this condition, and suggested that purchase of exercise pulse oximetry in conjunction with exercise testing may be an objective measure of degree of impairment.

Response: We did not adopt the comment. The documentation for arterial desaturation is found in the criteria in 4.06A or B and the methodology can be easily inferred from the criteria. We do not provide for the purchase of pulse oximetry to determine oxygen saturation because it is not as sensitive a measure as arterial blood oxygen tension. Of course, if we obtain it from a treating source, we will consider the results along with all other evidence.

Comment: One commenter thought that the requirement in 4.00C1 for a 12-lead resting ECG rather than a 3-lead test might pose a problem because the 3-lead test is used in smaller facilities and rural areas. Another commenter thought that the statement in 4.00C1 that the tracings of a resting ECG "must be submitted," might be misinterpreted by adjudicators to mean that a resting ECG must be purchased even if the documentation is sufficient to meet a listing that does not require an ECG.

Response: We did not adopt the comments. The provision is substantively the same as the provision in 4.00F of the prior listings, and has not been a problem in the past. We will, of course, accept a 3-lead ECG which is submitted to us by a treating or examining source, and use it to decide the case if the remaining information in the record is sufficient. If this information is not sufficient for a favorable decision, we may purchase the needed evidence. When we purchase an ECG, it will be a 12-lead test. With regard to the second commenter's point, inasmuch as the language in the final rule is nearly identical to the language in former 4.00F and there was no misunderstanding of that language, we do not believe that these rules will be misinterpreted either.

Comment: Two attorney commenters said that detailed descriptions of ECG findings without the original or legible copies of the tracing should be sufficient documentation, at least in the Second Circuit. The commenters, in nearly identical language, said that this regulatory requirement "frequently results in purchase of a CE with [its] attendant problems."

Response: We require ECG tracings to minimize error in adjudication and to ensure that standardized criteria for interpretation are applied to all claimants. Having the actual tracings, or a copy, also permits us to compare the findings with other evidence in the record—such as ECGs from emergency rooms and other medical sources—which the treating source might not have seen, and permits us to create a longitudinal picture through serial tests. Furthermore, our requirement for the actual tracings or a copy is only consistent with standard medical practice; it is routine to send the tracings along with other medical records whenever referring a patient to another physician, and for this reason it is rare for treating sources to provide the kind of detailed description of the tracing and the methodology in their narrative reports needed to understand the interpretation and methodology. Although it is possible that the physician could provide us with a narrative showing all of the information we would need to resolve any questions we might have, we have found it much simpler and more straightforward to review the tracings.

Under our recently published standards for obtaining consultative examinations, we do not anticipate the purchase of many consultative examinations merely because a treating source failed to submit actual tracings. The rules which are applicable nationwide, including in the Second Circuit, require us to make every reasonable effort to recontact the doctor to try to obtain the missing evidence. If the tracing is for some reason not available, the rules also provide that the treating source is the preferred source for a consultative examination, and that, furthermore, we may restrict the examination only to the evidence we need. Therefore, if it is necessary to complete the record, we may purchase a new ECG from the treating source, without the need for a full-blown, independent consultative examination. We will only purchase examinations from other physicians when we have good reason to do so, primarily because the treating source has been unable or unwilling to provide us with adequate information and is unable or unwilling to perform the consultation for us.

The commenters seemed to have assumed that our rule for obtaining tracings is one that only works to the detriment of the claimant. In fact, because we have specialists in cardiovascular disorders working in our State agencies, and because treating sources are not always specialists, we often find abnormalities on ECGs that were not noted in the treating sources' readings. Without review of the actual tracings by a knowledgeable program physician applying nationally acceptable standards, erroneous decisions could result which adversely affect claimants. The rule can, and does, work both ways.

Comment: A national medical association suggested that we should include a requirement for routine documentation of the effects of hyperventilation under 4.00C1b(4) because it is a necessary part of the examination.

Response: We partially adopted the comment. We could not add a rule requiring documentation of the effects of hyperventilation under 4.00C1b(4) because it is a general section that includes existing medical evidence. Inasmuch as this existing evidence often will not include documentation of the effects of hyperventilation, such a requirement would have little practical effect. However, we can require such testing when we purchase exercise testing on a consultative basis. Therefore, we have added this requirement under final 4.00C2b(1), the methodology section under "Purchasing Exercise Tests." The new language is an almost exact copy of the third and fourth sentences of the first paragraph of 4.00F2 of the former rules.

In addition, the comment made us realize that the statement in proposed 4.00C1b(4) was an inadvertent error. It required posthyperventilation tracings to be deferred for at least 10 minutes after hyperventilation, which is manifestly so long a recovery time that it would defeat the purpose of the hyperventilation study. We have, therefore, corrected the sentence to state what we originally intended—that the exercise test should be deferred for at least 10 minutes after hyperventilation.

Comment: A medical association provided an updated reference for exercise standards to be used in 4.00C1b(5).

Response: Instead of updating the literature citations, we decided to delete the references entirely in favor of a more general requirement to use generally accepted protocols for post-exercise electrocardiograms. In this way, we ensure that the criteria will not again become out-of-date in the future.

Comment: Several commenters asserted that the use of handrails and the Borg scale intensity rating of perceived exertion are not usually reported in the existing medical evidence. The commenters pointed out that, if this information is required for a program acceptable exercise test, it will require frequent recontact with the attending physician. They also said that this information will usually not be available from hospitals.

Response: In response to the comment, we deleted the language on the Borg scale and the use of handrails. We agree with the commenters that these are not usually reported or available.

Comment: One commenter referred to our statement in the third sentence of proposed 4.00C2a that "[e]xercise test reports, when available, must be included in the file." The commenter said that there are occasions when certain exercise tests would not be germane to a decision, e.g., an exercise test performed only a week after an acute event or when chest discomfort is clearly not of coronary ischemic origin.

Response: Although we do not agree with the commenter, we have deleted the sentence in our clarification of final 4.00C2a. Inasmuch as we have revised 4.00C2 to focus only on the purchase of exercise tests, it is no longer relevant to discuss available (i.e., existing) evidence.

Even though we have not retained the sentence, it is still our policy that all relevant, existing evidence must be requested. In the case of a person alleging chest pain, this will include reports of exercise testing. We do not believe that it would be fair, or accurate, to automatically conclude that a person's chest pain is "clearly" not of cardiac origin when a physician has seen fit to order a treadmill test for the individual. Moreover, as we have stated several times in this preamble, the treadmill test provides valuable information about an individual's aerobic capacity which will always be relevant to determining functional capacity.

Comment: An attorney said that we had acknowledged in 4.00C2b(1) that "there is more than a little uncertainty about what constitutes a MET or 5 METS" on an exercise test, and said that there was no universally accepted guideline in the medical community for how many METS a person has completed at a given level of a treadmill test. The commenter suggested, therefore, that we provide a definition in the listings that would serve as the standard for these listings.

Response: We did not adopt the comment. It is universally understood by physicians that the metabolic equivalent, or MET, is a unit of sitting, resting oxygen uptake equal to approximately 3.5 ml of oxygen per kilogram of body weight per minute. Exercise protocols, such as the Bruce protocol, are standardized and have been developed so that each level approximates a workload in multiples of METS; hence, stage I of the Bruce protocol approximates 5 METS. (See, e.g., The American Heart Association Special Report, "Exercise Standards: A Statement for Health Professionals From the American Heart Association," "Circulation," Vol. 82, No. 6, December, 1990, p. 2286; and Schlant, et al., "Guidelines for Exercise Testing: A Report of the Joint American College of Cardiology/American Heart Association Task Force on Assessment of Cardiovascular Procedures (Subcommittee on Exercise Testing)," printed in the "Journal of the American College of Cardiology," Vol. 8, No. 3, September, 1986, p. 725, and "Circulation," Vol. 74, No. 3, September, 1986, p. 653A.) An exact MET level can be determined only by measuring the volume of oxygen uptake directly, but this is rarely done in conjunction with the clinical use of the ECG treadmill exercise test.

We do not believe that there is any reason for us to define the criteria of these listings beyond what is universally understood by physicians. Inasmuch as we use the listings only to find people disabled, the fact that there is some variation from person to person in the actual workload imposed at a given level of an exercise protocol will not disadvantage any claimant. Those claimants who demonstrate the requisite findings of the listings on exercise will be found to have impairments that meet the listings; those who do not will still have the opportunity to demonstrate disability by showing equivalence to the listings or by an individualized assessment of their residual functional capacities.

Comment: A commenter from a national medical association suggested the use of measured oxygen consumption to quantify functional capacity objectively. The commenter said that use of measured oxygen consumption may enhance objectivity of data obtained during exercise testing and strengthen interpretation of exercise test results in cases where effort on the part of the claimant is questionable; while measurement of oxygen consumption is not frequently used in clinical practice, the additional data gathered from this relatively simple, noninvasive technique may allow for more objective interpretation of respiratory or cardiovascular limitation.

Response: Although we agree that exercise testing with measurement of maximal oxygen uptake provides an accurate determination of aerobic capacity, this procedure is not widely utilized and would also increase the cost of impairment evaluation prohibitively. Therefore, we are unable to adopt the comment. However, in response to the comment, we have added a discussion of the efficacy of such testing to final 4.00C3, "Other studies," and the importance of obtaining test results when the test has been performed by a treating source or other medical source.

Comment: A national medical association suggested the 5 METS or less criterion should apply to leg exercise only and that we should provide a threshold for arm exercises. The commenter recommended that the arm threshold should be 60 to 70 percent of the leg criterion. The commenter also said that some work does not require much leg work and, therefore, thought the treadmill test could be inappropriate. The commenter said that some people without the use of legs may still be capable of performing many jobs.

Response: The primary focus of the preface to the listings is on providing guidance for the evaluation of disability under the listings, although the guidance may also be applicable for disability evaluation at other steps of the sequence. The listings are only examples of impairments that are so severe that they preclude the performance of gainful activity. We believe that the exercise threshold levels we have provided in these listings are so limited as to render the issue of arm exercise at the listing level moot; even sedentary work requires some ability to walk, stand, lift, and carry, as well as the ability to travel to and from a job. We also want to make clear that, under our longstanding rules, people who do not have the use of their legs are always deemed to be disabled.

We do agree with the commenter that arm exercise testing is less sensitive than treadmill testing in the detection of myocardial ischemia because the large muscle groups are not exercised. We believe that the maximal aerobic capacity is approximately 80 percent of a treadmill exercise test. A MET level on arm exercise that would be comparable to the 5-MET level of the treadmill test would be 4 METs, and in such a circumstance, an equivalence determination would be appropriate.

Comment: Another commenter wondered whether the increase in systolic blood pressure before and early into exercise discussed in 4.00C2b(4) and the diastolic rise in blood pressure discussed in 4.00C2b(5) occur with sufficient frequency that they warrant inclusion in the listings. The commenter also suggested that we add a statement requiring review by a program physician to ensure the validity of the test interpretation. Another commenter recommended that we add the phrase, "at peak level," to the third sentence of proposed 4.00C2b(5), so that the sentence would begin: "Conversely, an increase in diastolic blood pressure with low workloads, together with a decrease in systolic blood pressure at peak level and decrease in pulse pressure * * *." Another commenter also asked if the significance of severe systolic hypertension during exercise needs to be mentioned.

Response: We believe that the guidance we provide in final 4.00C2b(4) is important and that the phenomenon does occur frequently enough to warrant inclusion in the methodology section of the preface. It is well known that there is a degree of apprehension in the patient-physician encounter which can result in a transient elevation of blood pressure, and this occurs commonly. As the test progresses, the transient elevation in blood pressure settles down. This degree of apprehension must be taken into consideration when blood pressure is measured early in an exercise test in order to distinguish between tests that are truly sign-or symptom-limited and those that register transient elevations in blood pressure early in the test as a result of apprehension.

We agree with the first commenter that the changes in diastolic blood pressure discussed in proposed 4.00C2b(5) were not as important; it is the decrease in systolic blood pressure which is associated with significant coronary disease and which increases the risk of further treadmill testing. Moreover, the paragraph could have been misleading; it is, in fact, not invariably true that an increase in diastolic blood pressure with low workloads, together with a decrease in systolic blood pressure and decrease in pulse pressure, indicates a response to inadequate cardiac output. There are other reasons why this may occur. Therefore, we have deleted the paragraph and redesignated proposed 4.00C2b(6) (with modifications to delete redundancies and cross-references not in the final regulation) as final 4.00C2b(5). We did not adopt the comment to add "at peak level."

We did not adopt the comment that asked us to add a sentence ensuring review by a program physician because the results of an exercise test must be considered in the context of all of the other evidence in the individual's case record, and we did not wish to mandate a physician review of only one specific type of evidence. We did not single out the issue of blood pressure because we believe that it is well understood by experienced program physicians that interpretation of exercise tests takes into account all relevant factors, including the individual's exercise time, stage of exercise, electrocardiographic manifestations, and blood pressure changes. Blood pressure changes as an isolated finding are not the basis of an exercise test interpretation.

Finally, we adopted the comment that asked us to add information about the significance of systolic hypertension on exercise. We added a sentence to the end of final 4.00C2b(4) that states: "In addition, isolated systolic hypertension may be a manifestation of arteriosclerosis."

Comment: A physician, noting that 4.00C2b(4) discusses the significance of a decrease in systolic blood pressure during exercise to below the usual resting level, wondered whether it would be helpful if we added a discussion of the significance of a decrease in systolic blood pressure during exercise that did not fall to below the resting level. The commenter noted similar passages in listings 4.02B2 and 4.04A4.

Response: As the commenter is undoubtedly aware, a decrease in systolic blood pressure that does not fall below the resting level may have no special significance. It is merely one of many factors the doctor must consider when interpreting the test results. Because we do not, and could not, list all of the factors that must be considered in exercise test interpretation, we do not believe that it would be helpful to address this particular one, either in the preface or the listings.

Comment: We received two general comments about the examples of risk factors in proposed 4.00C2c. One commenter noted that there were fewer risk factors named in the proposed rules than in 4.00G3 of the former rules, and that we had recategorized some of the risk factors militating against the purchase of exercise testing in the former rules to factors that would limit interpretation of exercise test results in the proposed rules. This commenter, who believed that the rules "contemplate[] large numbers of exercise tests purchased by State agencies notwithstanding the judgments of treating physicians," said that we would also be exposing claimants to additional risk "by abandoning many of the factors which have heretofore restrained the discretion of SSA decisionmakers." In a similar vein, the second commenter indicated a belief that, by changing some of the factors in the prior rules (such as musculoskeletal and neurological impairments) from risk factors to factors that would limit the usefulness of exercise testing, we were indicating that we would purchase exercise tests in which the results are likely to be of limited or no value because of the presence of other diseases.

Response: In response to the comments, we have restored to final 4.00C2c several of the examples from the former rules of conditions that would preclude purchase of exercise testing; specifically, Wolff-Parkinson-White syndrome (which we have deleted from final 4.00C2e(2)), marked aortic stenosis, marked pulmonary hypertension, and limiting musculoskeletal and neurological impairments. We had proposed to delete the examples of second-and third-degree heart blocks because they are subsumed under the category of arrhythmias; moreover, second-degree heart blocks are not always significant risks to exercise testing, as in the case of Wenckebach periods. However, in response to the comments, we have included third-degree heart blocks and Mobitz II, a kind of second degree heart block that would preclude the purchase of testing, as examples of arrhythmias. We deleted the example of ventricular aneurysms because we now know that they do not rupture; therefore, we no longer consider them to be a significant risk to exercise testing for that reason. Arrhythmias that result from ventricular aneurysms would already be covered. However, in response to the comments, we have revised the example of aneurysms that was in the former rules to include chronic or dissecting aortic aneurysms and have added it to the list of examples. We deleted the example from the former rules of recent onset angina because it is too imprecise. Unless the angina is unstable or progressive, there is no significant risk to the purchase of testing.

In some cases, the examples in the NPRM were the same as the examples in the former rules, but with more up-to-date or clearer language; for instance, "atrial" fibrillation instead of "auricular" fibrillation and the example of cardiac drug toxicity instead of the less precise "individuals on medication where performance of stress testing may constitute a significant risk." In the final rules, we have also clarified an example we proposed in the NPRM. Because "recent" aortic dissection was unclear, we have revised the example to "unrepaired" aortic dissection; thus, we will not require recency. The proposed rules contained the example of arterial dissection after coronary angioplasty; however, we eliminated this example because it was medically inappropriate. Coronary arterial dissection usually occurs during coronary angioplasty and is the major means by which arterial blood flow is reestablished. Therefore, it is not a risk factor for exercise testing. In addition, we have added language to 4.00C2c to indicate that the risk factors listed are only examples, not an all-inclusive list. Based upon the facts of each individual's case, the program physician can decide if other factors would pose a risk to the claimant.

With regard to the first commenter's remark that we intend to purchase many exercise tests despite the judgments of treating physicians, we believe that the commenter misread the NPRM. As we have already explained, the proposed rules clearly stated that the program physician would "give great weight to the treating physicians' opinions and will generally not override them," in this regard.

Comment: One commenter from a national medical association recommended that the unstable progressive angina in 4.00C2c should be well-documented as it could be attributable to anxiety, especially following coronary artery bypass graft surgery.

Response: We always try to document any impairment as well as possible. However, we believe that unstable angina due to any cause raises a sufficient question of risk that we will not purchase exercise testing.

Comment: One national cardiological association said that our proposal in 4.00C2d to wait 3 months following an acute myocardial infarction or angioplasty before purchasing an exercise test was too long. They indicated an exercise test can be safely done within 1 to 2 months. Conversely, a lay commenter suggested that we might extend the deferral period to 4 months.

Response: We did not adopt the comments. We chose a 3-month deferral period not for reasons of safety, but to provide a reasonable period of time for recovery so that we could realistically evaluate the benefits achieved from medical management. We also believe that 3 months is long enough to wait before purchasing the test, and that there is no need to extend the requirement to 4 months.

Comment: One commenter asked whether the factors in proposed 4.00C2e (final 4.00C2e(2)) could be used to rule out exercise testing, inasmuch as they limit test interpretation.

Response: The factors in this section may, on an individual case basis, be used for this purpose, although they generally will not preclude exercise testing. We have included this section to alert program physicians that these things sometimes make interpretation of an exercise test more difficult. They do not necessarily pose a risk to the individual, nor do they make interpretation impossible.

Comment: The same commenter asked us to define "prolonged periods of physical inactivity."

Response: We have added a parenthetical statement indicating that an example of a prolonged period of physical inactivity could be two weeks of bedrest.

Comment: Two attorneys, using identical language, said that they were concerned about the "requirement" for an exercise test for individuals with chronic heart failure, who generally are treated with digitalis. The commenters said that the medication alters the ECGs and can result in misinterpretation of the exercise test results, thereby denying benefits to disabled individuals. Two other attorneys, also using identical language, said that the rules authorized reliance on exercise tests where medication conceals the underlying impairment. One of these commenters said that we should instruct adjudicators to disregard results of exercise tests that are unfavorable to individuals when the individuals were on medication at the time of the test. Another commenter thought that we should not use treadmills or Holter monitors to evaluate ischemia when the individual takes medication that controls symptoms because a negative test result does not give a reliable picture of the person's ability to function on a sustained basis. This commenter also said that we routinely invalidate positive test findings when they are supportive of listing criteria because the claimant is on digitalis.

Response: The final rules do not require exercise testing for people with chronic heart failure. They provide a listing under which people who have chronic heart failure can show listing-level severity through exercise testing, in addition to two other listings for the impairment that do not include exercise testing among their criteria. Moreover, the exercise test criteria in listing 4.02 do not include ischemic findings; they require only an inability to exercise at 5 METS or less or a need to stop exercising because of certain abnormalities that require the cessation of the test. Therefore, any effect of medication on the ECG will have no relevance to whether the person meets this particular listing. However, because these commenters misunderstood the language we had proposed, we have revised listing 4.02B1 to make it clearer. We also repeat that the results of exercise tests alone will not result in denials of benefits.

With regard to the issue of the value of exercise testing when medication controls symptoms, a so-called medication-affected test in an asymptomatic individual indicates the effectiveness of the treatment. The absence of ST changes on an exercise test does not mean that the person is free of coronary disease; however, the absence of ischemic ECG changes or symptoms at a given level of exercise indicates what can safely be done in terms of aerobic capacity. As we have already stated, we do not use the exercise test to determine the existence of an impairment, but the aerobic functional limitations associated with the impairment. If medication sufficiently controls an impairment, the impairment is not disabling.

With regard to digitalis, we agree with the commenters that the presence of baseline segment abnormalities makes interpretation of an abnormal exercise test difficult and reduces the usefulness of such a piece of medical evidence in the evaluation of ischemic heart disease. This is because digitalis may induce or accentuate ST segment displacements on exercise, resulting in a false-positive test; for this reason, an abnormal exercise test of an individual taking digitalis could not in itself establish that the individual has met the criteria of the ischemia listing. However, this is not the case for normal tests, and a normal exercise test is valid. In any event, inasmuch as we have deleted the requirement to purchase exercise tests we believe that we have responded to the commenters' concerns.

Comment: A national medical association suggested that we should place less reliance on imaging techniques such as echocardiography and radionuclide tests in determining function. The commenter said that too much emphasis is placed on ejection fraction when tests must be related to aerobic capacity, and that the effective cardiac output and effective stroke volume are more important.

Response: We did not adopt the comment. It was not clear to us what the commenter meant by "effective" cardiac output and stroke volume. However, both cardiac output and stroke volume are measures of ventricular function that do not in themselves provide enough information from which one could infer reduced functional capacity. The implications of both depend on variables that we could not easily define in these listings. For instance, cardiac output depends on stroke volume and heart rate—which can vary through a wide range—as well as according to the individual's size. Therefore, neither test is amenable to a simple cutoff for listings purposes as the ejection fraction is. Moreover, stroke volume is one of the components used to calculate the ejection fraction. Even if we provide a threshold level for stroke volume in these listings, it would likely be redundant because it would also appear as a reduced ejection fraction.

Comment: One commenter felt that stricter guidelines need to be given as to when studies described in 4.00C3 may be purchased to ensure that the treating physician's opinion is carefully considered in order to protect the safety of claimants.

Response: We have clarified section 4.00C3 to indicate that these tests may only be purchased if the claim cannot be favorably decided on another basis, and the program physician, preferably one experienced in the care of patients with cardiovascular disease, has concluded that the test does not pose a significant risk to the individual.

Comment: One commenter asked us to provide criteria for purchase of two-dimensional echocardiograms and radionuclide studies in 4.00C3.

Response: We did not adopt the comment. We believe that this is a procedure that is more appropriate to our operational manual and other subinstructions than to the regulations.

Comment: One commenter suggested that the requirement in proposed 4.00D4 to obtain results of biopsies during the first few months after transplant was unnecessary because the listing directs an allowance once we have documented that a transplant has been performed.

Response: We adopted the comment by deleting the sentence. As the commenter correctly noted, under listing 4.09 there is a presumption of disability for 1 year following cardiac transplantation so that there is no need for requiring biopsy results within that period.

Comment: A physician noted the statement in proposed 4.00E1a that chronic heart failure may manifest itself as pulmonary congestion "or" systemic congestion could be clarified to indicate that it might also result in both pulmonary and systemic congestion.

Response: We adopted the comment.

Comment: One national medical association suggested that we add "chest wall syndrome" to the list of noncardiac conditions that may also produce symptoms mimicking myocardial ischemia.

Response: We adopted the comment in final 4.00E3f.

Comment: A national medical organization suggested three clarifying language additions for the second and third paragraphs of 4.00E4.

Response: We adopted two of the suggestions but not the third. In the first sentence of the second paragraph of 4.00E4, we added the phrase, "in the supine position," following the word, "determined," as suggested because this is standard practice for determining the ankle/brachial blood pressure ratio. Similarly, in the fifth sentence of the third paragraph of the section, we now indicate that the treadmill for Doppler exercise may be on a "10 or 12 percent" grade, instead of 12 percent alone, in recognition of variations in standard practice; for the purposes of this listing, the difference in the grading is so small as to be immaterial. We did not adopt the third suggestion, to indicate in the same sentence that the exercise should be for "up to" 5 minutes, because we already include the possibility that the individual will not complete the test in the fourth sentence of the paragraph.

Comment: Under 4.00E4 there is a requirement for a resting ankle/brachial systolic blood pressure ratio of less than 0.50. A national medical association commented that this is an arbitrary ratio and that pain and pallor must also be taken into account, even though the individual may register 0.45.

Response: We disagree. The requirement for a resting ankle/brachial systolic blood pressure ratio of less than 0.50 is not arbitrary, but is confirmed by the literature and clinical experience. (See, e.g., Carter, S.A., "Rate of Pressure Measurements in Vascular Disease," in Bernstein, E.F., "Non-invasive Diagnostic Techniques in Valvular Disease," The C.B. Mosby Company, Saint Louis, 1987.)

Furthermore, the existence of pain and pallor is implicit in the requirements. Listing 4.12 requires "intermittent claudication," which is a description of a kind of pain. People with the severely reduced ankle/brachial systolic pressure of 0.50 will have both pallor and rubor, among other clinical findings, and the symptom of pain associated with the degree of marked peripheral arterial disease described in the listings. We do not believe that it would be helpful to mention only one such common finding.

Comment: One commenter suggested that we add text to 4.00E4 explaining that varicose veins by themselves do not generally cause disability at the listing level, and that the listing concerns chronic venous insufficiency, not isolated varicosities.

Response: Although the commenter was correct, we do not believe that it is necessary to add this much detail to the listings, since it is evident from the listed criteria that a mere finding of varicose veins could not meet or equal the criteria of the listing.

Comment: Several commenters believed a definition of cardiac enlargement should be provided.

Response: We agree and have added two common examples of cardiac enlargement to listing 4.02A.

Comment: Two attorney commenters noted that proposed listing 4.02 introduced an exercise test criterion. One of these commenters said that we had presented no justification to support our conclusion that exercise testing is an appropriate measure of disability arising from chronic heart disease. The other commenter noted that in our former rules (in 4.00G3), we had listed congestive heart failure as a reason for not purchasing exercise tests. A third commenter asked us to define "markedly symptomatic exercise intolerance" as used in proposed listing 4.02B1.

Response: Experts in the field of cardiology indicate that exercise testing of individuals with chronic cardiac failure that is compensated is both safe and useful. Among the objectives for subjecting these individuals to exercise testing are the sensitive and specific detection of impaired cardiac performance, a gradation of the severity of chronic cardiac failure, the assessment of aerobic capacity of the patient, and the monitoring of the therapeutic response (see, e.g., "Cardiopulmonary Exercise Testing," Karl T. Weber, Joseph S. Janicki, 1986, W. B. Saunders Company, Philadelphia). Nevertheless, we still exclude individuals who have chronic heart failure resulting in NYHA functional class IV limitations in 4.00C2c.

Our proposal to use the term "markedly symptomatic exercise intolerance" was unclear. We have, therefore, clarified final listing 4.02B1 by deleting the phrase and revising the final clause to state clearly what we intended; i.e., that the person is unable to exercise in an exercise test at a workload equivalent to 5 METS or less because of symptoms of chronic heart failure. In this way, the criteria for symptoms will be evaluated only under B2 of final listing 4.02.

Comment: One commenter asked if in listing 4.02B the "recurrent fatigue and dyspnea" must occur at rest or with exercise, or either.

Response: NYHA functional class III connotes that the individual is comfortable at rest but symptomatic on ordinary physical activity. Similarly, NYHA functional class IV connotes that the individual is symptomatic at rest. We believe that we have clarified these principles by deleting the references to the functional classes and instead incorporating into the listings narrative descriptions of the functional criteria.

Comment: One commenter suggested that we include multiform ventricular premature contractions in listing 4.02B2a.

Response: We adopted the comment. We have rephrased this listing section to indicate, "Three or more consecutive ventricular premature beats or three or more multiform beats * * *."

Comment: Three commenters addressed the requirement in the opening sentence of proposed listing 4.04 for chest discomfort "occurring repeatedly." One commenter asked us to define the phrase. Two commenters pointed out that individuals may restrict their activities so as to avoid discomfort or lessen its frequency. One of the commenters said that we should delete the phrase, arguing that the issue is not how often chest discomfort of ischemic origin occurs, but whether it occurs consistently with exertion.

Response: We adopted the comment that asked us to delete the phrase. We agree with the last commenter that the issue is not how often the chest pain occurs, because the individual may have structured his or her activities so as to avoid chest pain, but whether the individual would have symptoms with a given level of exertion. Moreover, the specific listings criteria in the subparagraphs establish the level of functional severity for the listings.

Comment: One commenter thought that the requirement in listing 4.04A1 for findings in at least 3 consecutive complexes would make it too difficult to meet this listing. Another commenter asked if the requirement in listing 4.04A1 for depression in 3 complexes rather than 2 is correct.

Response: The requirement for findings in at least 3 consecutive complexes is correct. We changed the requirement from the prior listing's 2 complexes to bring these listings into line with standard medical practice. We do not intend to raise the severity level of the listing, but to make it more accurate and up-to-date by minimizing the effect of wandering baseline or artifacts.

Comment: One commenter requested that we consider expanding proposed listing 4.04A5 to include abnormal findings on stress echocardiogram and IV Persantine tests. The commenter said that these tests could improve sensitivity in assessing claimants with coronary artery disease who have negative routine exercise tests.

Response: We did not accept the comment because these are new technologies that are still being developed. Moreover, they are not widely available and there is currently no correlation indicating the sensitivity or specificity of these tests relative to more accepted technologies, such as thallium stress testing. For these reasons, it would not be appropriate at this time to include such modalities in our listings.

Comment: Two attorney commenters thought that proposed listings 4.04B and C tightened the criteria in the nonexercise test listings for ischemic heart disease and could not be met. One of these commenters said that this constriction of the listings was unfair and unwarranted, and that the former listing did not have these requirements. Another commenter asked why listing 4.04C requires both angiographic evidence of severe vessel occlusion and an ejection fraction of 30 percent or less.

Response: In response to the comments, we revised the rules so that final listing 4.04C requires only angiographic evidence instead of both angiographic evidence and a decreased ejection fraction. We also revised final listing 4.04B so that it has the same functional severity level as listing 4.04C, i.e., the condition must result in symptoms on ordinary physical activity even though the individual is comfortable at rest; this equates with NYHA functional class III. Proposed listing 4.04B had required NYHA functional class IV, which requires symptoms at rest.

We revised the listings employing ejection fraction criteria from the prior rules, which permitted a finding of "meets" based on the ejection fraction alone, because it is well supported in the medical literature that individuals with even markedly decreased ejection fractions can functionally live perfectly normal lives; the ejection fraction alone does not indicate functional capacity. This is also the reason we require functional limitations associated with angiographic evidence of coronary atherosclerosis; as we have already explained, this test does not provide information about whether any discovered disease is directly related to or predictive of functional status.

Comment: Two commenters noted that we deleted the criteria of former listings 4.04B1 through B5, which provided criteria for evaluating chest pain following myocardial infarctions based on resting ECG findings. One commenter said that resting ECGs should not be entirely omitted; another said that we should retain the prior listings.

Response: We did not adopt the comments. We deleted the criteria because the presence of a previous myocardial infarction does not, of itself, guarantee that subsequent chest pain is of cardiac origin. Therefore, we need other information to establish that the individual is experiencing ischemia resulting in chest discomfort. An exercise test may be useful for this purpose. Additionally, a resting and exercise two-dimensional echocardiogram or radionuclide ventriculogram may also demonstrate cardiac dysfunction, including ischemia. Again, the fact that we have deleted the criteria does not mean that individuals who demonstrate these findings will be found not disabled. It only means that they do not have impairments that "meet" the criteria of the listing.

Comment: Two commenters asked whether we deliberately deleted the word "proximal" in proposed listing 4.04C2b (final listing 4.04C1b), "70 percent or more narrowing of another nonbypassed coronary artery"; former listing 4.04B7b, the corresponding listing that was in effect at the time of the NPRM, had specified that the artery must be a proximal artery.

Response: We did delete the word intentionally to recognize that lesions in other than a proximal portion of the major arteries may result in significant impairment.

Comment: One commenter noted that we had deleted the criterion in former listing 4.04A5 for development of second-or third-degree heart block. The commenter thought that heart blocks and conductive defects with exercise are an important part of the ischemic heart disease listings and asked us to consider reinstating the criterion.

Response: We did not adopt the comment. We deleted the criterion because it is unnecessary. Individuals with significant heart block or conductive defects will have either blood pressure changes or exercise intolerance and would be considered under the standards for those consequences of the impairment.

Comment: Four commenters believed that the requirement in listing 4.05 that cardiac syncope be coincident with findings on a Holter monitor is an unduly restrictive requirement. They said that medical care is unavailable to many claimants, particularly in the SSI claimant population, who may only have the benefit of a Holter evaluation because of a consultative examination. The commenters also said that it was unreasonable to require that the claimant sustain an episode during an isolated twenty-four hour period.

Response: We did not adopt the comments. There are many causes for syncope, including noncardiac causes and, except for the use of a Holter monitor with the individual lying on a tilt table or by ambulatory monitoring there is no other way to determine whether there is true cardiac syncope. There has to be an association between the symptom and medically determinable arrhythmia to satisfy the requirements of this listing.

Even though we agree with the commenters that there will be some individuals who have true cardiac syncope but who have not been appropriately tested by treating sources, we do not believe that this will occur as often as the commenters do. Cardiac syncope is such a severe symptom that it is unlikely that many people who experience it will not be appropriately tested. In any case, and as we have stated throughout this preamble, those individuals who cannot present evidence showing that their impairments meet this listing may still establish disability in other ways.

Comment: A national medical association suggested that the addition of the phrase "near syncope" would make the listing description of symptoms associated with uncontrolled arrhythmias more accurate.

Response: We agree, and have added the phrase in final 4.00B3 and listings 4.05 and 104.05.

Comment: One commenter noted that the vast majority of claimants suffering from symptomatic congenital heart disease who suffer cyanosis at rest will meet either the requirements of listing 4.06A1 or A2. However, the commenter was concerned that we might deny claimants whose impairments do not quite meet the requirements of those listings. The commenter believed that cyanosis at rest should in itself be sufficient to meet listing-level severity.

Response: We did not adopt the comment. The commenter was correct that the vast majority of people who are cyanotic at rest will have impairments that meet this listing. However, we did not list cyanosis by itself because it can be too subjective a finding on which to base a listing. The listings are not intended to be all-inclusive. Those individuals who have cyanosis at rest but who do not demonstrate the criteria of the listing may still demonstrate disability through the equivalence concept and, if necessary, through an individualized assessment of their residual functional capacities. However, as we explained earlier in this preamble, and in response to this comment, we deleted the proposed functional criteria from final listing 4.06 partly in recognition of the fact that individuals who are cyanotic at rest will necessarily have functional limitations of at least NYHA functional class III.

Comment: One commenter suggested that we provide specific guidelines in the listings regarding when to purchase blood gas studies because these types of tests constitute an increased risk for claimants.

Response: We did not adopt the comment. It would be very unusual for a person with congenital cyanotic heart disease not to have at least one blood gas test available; therefore, we do not expect to purchase many of these tests. We also believe that the guidelines for the purchase of such tests more appropriately belong in our operational manual instructions. These tests do not pose a risk to the claimant.

Comment: One commenter was apparently concerned about the references to listings 4.02 and 4.04 in listing 4.08. The commenter noted that individuals with hypertrophic cardiomyopathy are high risk individuals and should not be subjected to exercise tests.

Response: Listing 4.08 refers to criteria in other listings for the evaluation of cardiomyopathy because the consequences of this

impairment are so varied (e.g., heart failure, ischemia, arrhythmia) that it is more efficient to refer to other listings to evaluate the consequences rather than repeating all the criteria under listing 4.08. We agree that individuals with hypertrophic cardiomyopathy are at risk for sudden death due to arrhythmias and physical exertion and they should not be required to do an exercise test. We have added this condition to the risk factors in final 4.00C2c to make it clear that we will not purchase exercise tests in these cases. We did not revise the cross-references in final listing 4.08 because some individuals may have undergone exercise testing and may be able to show that their impairments meet final listings 4.02B or 4.04A. Those who have not may still be evaluated under final listings 4.02A and C, and 4.04B and C, the listings that do not include exercise tests among their criteria.

Comment: One national medical association indicated that some individuals undergoing cardiac transplantation have recurring disability beyond 1 year. The commenter pointed out that serial observations and retesting must be continued.

Response: Final listing 4.09 does not conflict with these facts. The listing provides that, after the year of disability, the individual must be reevaluated under whatever listing is appropriate to his or her residual cardiac impairment, where a finding of "meets" or "equals" may be possible.

Moreover, under the medical improvement disability review standard that we are required to follow under §§ 404.1594 and 416.994 (and § 416.994a for children) we may not find that an individual's disability has ended "automatically" after a certain period of time. We are generally required to show that there has been medical improvement in the individual's disabling impairment, that the improvement is related to the ability to work, and that the individual is not currently under a disability. This means that, even if the individual's impairment has improved, we may still find that the individual continues to be disabled based on an assessment of residual functional capacity and consideration of vocational factors.

Comment: One commenter thought that there were too many criteria required to meet listing 4.11. The commenter said that it should be unnecessary to require all five of the listed criteria in order to find a claimant disabled and that brawny edema alone should be sufficient for a finding of disability. Conversely, another commenter questioned whether this listing is too liberal, but did not say why.

Response: We adopted the first comment. Extensive brawny edema is the most severe, end-stage manifestation of chronic venous insufficiency. However, we believe that because there can be variability in the severity of each of the other findings we proposed in listing 4.11, they must all be present together in order to establish listing-level severity based on chronic venous insufficiency that has not resulted in extensive brawny edema. We have, therefore, divided final listing 4.11 into two paragraphs. Final listing 4.11A will be met when there is extensive brawny edema resulting from deep venous insufficiency, while final listing 4.11B will be met with deep venous insufficiency with associated superficial varicosities, stasis dermatitis, and recurrent or persistent ulceration which has not healed following at least 3 months of prescribed medical or surgical therapy. We did not include a 3-month treatment requirement in final listing 4.11A because extensive brawny edema is a sufficiently severe finding in itself that treatment is not an issue in establishing duration.

We also reiterate for the first commenter that an individual does not have to demonstrate an impairment that "meets" a listing to be found disabled.

Comment: One commenter suggested that we consider adding a statement about the evaluation of lower extremity pain without evidence of chronic venous insufficiency.

Response: We did not adopt the comment. It was not clear to us from the comment what information the commenter was suggesting we provide. However, we have regulations for the evaluation of pain that we believe adequately cover the issue for all impairments.

Comment: One commenter suggested that we also require failure to visualize superficial arteries in listing 4.12A.

Response: We did not adopt the comment. When the common femoral or deep femoral artery cannot be visualized, superficial arteries will not be visualized either. Therefore, such a requirement would be superfluous.

Comment: Another commenter suggested that we add a third criterion, "ischemic pain at rest," to listing 4.12B.

Response: We did not adopt the comment. Individuals with peripheral vascular disease who have ischemic pain at rest will have more severe impairments than are already described in final listing 4.12B1. Such individuals have resting ankle/brachial systolic blood pressures in the range of 0.30 to 0.40, which is even more severe than the listing requirement of 0.50.

Comment: One commenter pointed out that the statement in the fifth paragraph of proposed 104.00A about age-appropriate activities for individuals age 14 1/2 to 18 (i.e., that they "are not unlike those of young adults") was inconsistent with the interim final rules for evaluating childhood disability we had issued on February 11, 1991 ("Determining Disability for a Child Under Age 18," 56 FR 5534).

Response: We have deleted the entire paragraph. The rules for evaluating childhood function in §§ 416.924 through 416.924e provide information on the evaluation of age-appropriate activities for children of all ages irrespective of the nature of their impairments. They are, therefore, also applicable to children with cardiovascular impairments.

Comment: One commenter thought that the requirement in the second sentence of proposed 104.00B, which provided in part that a copy of relevant echocardiographic views "must" be included in the record, was unrealistic. The commenter pointed out that the State agency is not always able to obtain these copies. Therefore, the commenter recommended that we change the word "must" to "should."

Response: We adopted the comment.

Comment: A physician recommended that we eliminate the reference to patent ductus arteriosus in the fourth paragraph of proposed 104.00D because this condition is almost always amenable to surgical or medical therapy with very low morbidity.

Response: We adopted the comment. Because of advances in the medical and surgical treatment of patent ductus arteriosus, infants born with this condition would not ordinarily be expected to experience chronic cardiopulmonary impairment. We have, therefore, deleted the paragraph. We have also clarified final 104.00C and D so that they clearly indicate the kinds of conditions and the level of severity intended by the discussion. In addition, we have included guidance in final 104.00D for those conditions which, even though not named, nevertheless result in listing-level impairment in individual cases.

Comment: A national medical association, noting our reference to "imaging techniques such as echocardiography and radionuclide studies" in the last sentence of proposed 104.00E, pointed out that pediatric cardiologists almost never do radionuclide studies. The commenter suggested that we revise the last sentence to refer to other tests, such as two-dimensional and Doppler echocardiography.

Response: We adopted the comment.

Comment: One attorney was concerned that our proposal to remove the tables for tachycardia and tachypnea from listing 104.02 would increase the severity level of the listing. The commenter thought that the exercise intolerance criterion was more restrictive in proposed listing 104.02 than in former listing 104.02G. The commenter also thought that there was no justification for removal of chest x-ray evidence demonstrating cardiomegaly. The commenter was also concerned that hepatomegaly and edema were not included as criteria for this listing, even though they were in the prior listing. This commenter and another commenter noted that there were no criteria for fatigue and dyspnea in proposed listing 104.02.

Response: Although the commenter was not correct in the belief that we had proposed to make the listing more stringent, we adopted most of the comments. Final listing 104.02A is now a criterion for persistent tachycardia at rest; final listing 104.02B is now a criterion for persistent tachypnea at rest or markedly decreased exercise tolerance; and we have included the corresponding tables from our former listings as Tables I and II at the end of the section. We moved the other criteria we had proposed in listing 104.02A into final 104.00E of the preface, and also added chronic dyspnea, tachypnea, and orthopnea to the list of symptoms (including fatigue and weakness) we had already included; we did not specifically mention all of them in the proposed listings because they are symptoms of chronic heart failure with which all physicians are familiar. Our proposed addition of the phrase "markedly reduced" (now "markedly decreased" in the final rules) to modify the requirement for exercise tolerance was only a clarification; infants with labored respiration on mild exertion have markedly decreased exercise tolerance.

The opinion of the commenter that there was no justification for our deletion of the rule for cardiomegaly demonstrated by x-ray was not correct. Our prior listing could be met only with the demonstration of two criteria associated with chronic heart failure. The proposed (and final) listing is the same, in that it requires as an initial matter the demonstration of findings—such as cardiomegaly—that confirm the existence of the impairment, and one other of the criteria in the subparagraphs. Our proposal to delete the requirement that cardiomegaly be demonstrated by x-ray was consistent with contemporary standards of care for infants and children with chronic heart failure, in which other, more accurate, imaging techniques, including echocardiography and other radionuclide studies, are generally used. Heart failure in children must be properly assessed; like cancer, it is not a diagnosis to be lightly entertained, and should not be based on an x-ray in a physician's office.

However, we concede that not all claimants receive the proper care and that there will still occasionally be x-rays in the evidence without appropriate imaging techniques. Therefore, in response to the comment, we have restored our prior rules on x-ray evidence of heart enlargement demonstrated by cardiothoracic ratio in final 104.00E, with the provision that the findings must also be accompanied by other signs of chronic heart failure or ventricular dysfunction. Because we restored these provisions, we were also required to clarify how enlargement is evaluated using appropriate imaging techniques, such as echocardiography. We have not included the former requirements for enlargement on serial x-rays or for measurement of chamber size based on x-rays because these are outdated methods of assessment.

The foregoing also explains why we removed the listing criteria for hepatomegaly and dependent edema. Like cardiomegaly, these are examples of medical findings that are used to confirm the existence of chronic heart failure, as described in 4.00E.

With regard to the comments on dyspnea, we also added two new paragraphs to final 104.00B providing for the possibility of exercise testing in children. Exercise testing has been increasingly used in children age 6 and older for the assessment of some arrhythmias, for the assessment of the severity of chronic heart failure, and for the assessment of recovery of function following cardiac surgery or other therapy; however, it is still less frequently indicated for children than for adults and can rarely be successfully performed on children under age 6. We do not expect to purchase many such tests for children, but they may be useful in those few cases in which we are unable to obtain sufficient evidence about a child's functioning (for instance, from treating sources and schools) to make a decision. The first paragraph, therefore, provides that we may purchase the test only if we cannot decide the case based on the available evidence; therefore, it cannot be used to disadvantage any child claimant. The paragraph also requires that the testing must be performed in a specialty center for pediatric cardiology or other facility qualified to perform exercise testing for children.

The second paragraph parallels the adult rules by requiring that such testing be performed using a generally accepted protocol consistent with the prevailing state of medical knowledge and clinical practice. In addition, it prohibits the purchase of the test for a child for whom the performance of the test constitutes a significant risk and adopts by reference the risk provisions in the adult rules, which are also applicable to children.

Comment: The same attorney commenter thought that the table for elevated blood pressure in listing 104.03 established a higher level of severity in the listings for children. The commenter also noted that we had not provided any explanation in the preamble to the NPRM of our reasons for making the change.

Response: As we explained in the preamble to the NPRM, Table III in final listing 104.03 comprises up-to-date values that are recommended by the American Academy of Pediatrics (56 FR 31269). The table is taken from the definition of "Significant Hypertension" in Table 5 of the "Guidelines of the American Academy of Pediatrics; Report of the 2d Task Force on Blood Pressure Control in Children—1987," "Pediatrics," 1987; 79:1-25. These guidelines should be in every pediatrician's office.

The levels in Table III are equal to or greater than the 95th percentile for age. In response to the comment, we have added this explanation to the opening of the final listing.

In most instances, the readings in final Table III are not more severe than those on the former table. We also provide more refined age category breaks. In any event, no child will be disadvantaged by our updating of the rule inasmuch as, to meet this listing, a child must also show another, consequent listing-level impairment, as in the former listing. Children whose blood pressures do not rise to the levels on the table but who have listing-level kidney, cerebrovascular, or heart impairment will still be found disabled at the listing level by virtue of their secondary impairments.

Comment: The same commenter thought that our proposal in 104.05A to require that any episode of cardiac syncope be coincident with the documented arrhythmia, was a new requirement.

Response: As we explained under the comments about the corresponding adult rule, listing 4.05, the proposed language merely clarified what is standard medical practice. In order for cardiac syncope to be related to the arrhythmia, it must be coincident with the arrhythmia. In response to the comment, we have also added a new second sentence to the second paragraph of final 4.00B, "Documentation," providing procedural guidance on when the purchase of an ambulatory ECG may be appropriate.

Comment: The same commenter thought that by cross-referencing the chronic heart failure listing in proposed listing 104.05B while omitting the exercise intolerance criteria in former listing 104.05C, we were suggesting that even a claimant meeting the more exacting stricture of the listing 104.02A exercise intolerance test would still be found not to meet the arrhythmia listing unless he or she also suffered from chronic heart failure. The commenter felt that an explicit exercise criterion should be included in the arrhythmia listing.

Response: We do not believe it is necessary to add the explicit exercise criterion to 104.05. Cardiac arrhythmias are common in children, and most childhood arrhythmias are essentially benign. When a child with a cardiac arrhythmia develops exercise intolerance with labored respirations on mild exertion, this is usually indicative of chronic heart failure. This is why we provided the cross-reference to listing 104.02 instead of incorporating the criteria from listing 104.02 into the listing on arrhythmia, which is what the prior listing essentially did. In addition, there are other criteria associated with arrhythmias in childhood, such as syncope or near syncope, which are more frequent manifestations of severe cardiac arrhythmias and which may not be necessarily associated with congestive heart failure.

However, in response to the comment, and because we have tried in these final rules to maintain consistency between part A and part B wherever possible and appropriate, we have revised listings 104.02 and 104.05. We have removed the reference to listing 104.02 from the arrhythmia listing and have instead added a criterion for arrhythmia under final listing 104.02C, by which that listing can now be met; this better describes our original intent than did proposed listing 104.05B. In addition, we revised the remaining language in final listing 104.05 into a single paragraph that is more consistent with the adult rules, as explained in the summary of provisions.

Comment: A medical association recommended that we add nonsyncopal arrhythmia to listing 104.05. Another commenter questioned why we did not mention A-V dissociation in the proposed listing; it was a criterion in the heading of the former listing.

Response: In response to the comment, we have revised final listing 104.05 to include near syncopal episodes clearly documented to be associated with arrhythmia. This is also consistent with our addition of the term to final listing 4.05.

We did not actually delete A-V dissociation from the listing. A-V dissociation is a kind of heart block and is, therefore, implicit in the term "heart block." We, therefore, proposed to delete the term because the prior listing language was ambiguous: The use of the conjunctive, "or," in the former rules did not mean that A-V dissociation was a separate medical entity from heart block; it meant that A-V dissociation was another way of describing a heart block. However, so that there is no misunderstanding, we have restored the term parenthetically after "heart block" in the final listing.

Comment: A physician questioned our approach in proposed listing 104.06A (final listing 104.06H). The commenter pointed out that the proposed rule to consider all children who meet the requirements of the listing to be disabled until age 24 months would result in inequitable outcomes. For instance, a child who had life-saving surgery at age 3 months would in effect be found disabled for at least 21 months, whereas a child who was found disabled under this listing at age 11 months would be found disabled for only 13 months. The commenter suggested that a more equitable approach would be to provide a recovery time period instead of an age cutoff.

Response: We adopted the comment, which actually conveyed our original intent. We have revised the final rule to state that we will consider an infant who meets the requirements of the listing to be disabled until the attainment of age 1, or for 12 months after surgery, whichever is later, and then reevaluate. We believe that 12 months is a sufficient period for stabilization and developmental recovery and, in this manner, all infants will be treated in the same way under the listing. We tied the recovery time period to surgery because the listing is for children who are expected to die within the first year of life without surgical intervention. If by age 1 an infant has not had surgery and has survived, the child should be reevaluated because the impairment may no longer be as severe as it originally appeared. Even if it is of listing-level severity at age 1, we should be able to evaluate it under one of the other listings; if it is less severe than listing-level, we will perform an individualized functional assessment and determine whether the child is still disabled.

In addition, we believe that the comment also required us to revise the period described in proposed listing 104.09, "Cardiac transplantation," because it used the same language. Indeed, proposed listing 104.09 could have been viewed as being even more inequitable. This is because, under the proposed rule, all children over age 1 would have received a period of 12 months (because a 12-month period for a child over 12 months old would always end later than age 24 months), whereas all children under age 1 would always receive a period of more than 12 months, some for nearly 2 years, others for closer to 1 year. In sum, we believe that the revisions to final listings 104.06H and 104.09 are simply fairer.

Comment: A commenter indicated that squatting, which was a criterion in former listing 104.04B, was not included in listing 104.06, the corresponding listing in these rules, and questioned its absence.

Response: We proposed to delete squatting because it is now a rare finding. In the past, squatting was usually seen in older surviving children with uncorrected cardiac defects associated with dyspnea and cyanosis at rest. Because the majority of children with cyanotic forms of congenital heart disease are now undergoing corrective surgery within the first year of life to restore oxygenation at more physiological levels, squatting is rarely seen. Nonetheless, to encompass the rare situation in which squatting may be present, we have restored it as a criterion under final listing 104.06A3.

Comment: A national medical association thought that the four manifestations of persistent, chronic hypoxemia we proposed in listing 104.06B might be too rigid. As an example, it said that a hematocrit of 55 percent or greater could create a problem for children who are anemic, and suggested that the qualifier "in the absence of documented anemia" should be added to the criterion. An attorney noted that former listing 104.04C required a "chronic" hematocrit of 55 percent or greater, whereas proposed listing 104.06B2 specified a hematocrit of 55 percent or greater "on two or more evaluations within a 3-month period." The commenter thought that this was a more stringent requirement than the one in the former listings, called it a "hidden" change in the listings, and strongly objected to it.

Response: We did not adopt the comments. Other than providing the example of anemia, the first commenter did not provide any other explanation for the belief that the four criteria in final listing 104.06A are too rigid. We believe that anemic and hypoxic children will be covered by final listing 104.06A1 because the arterial oxygen saturation should be significantly reduced as a result of anemia. The requirement that hematocrit results at 55 percent or greater be obtained twice or more within a 3-month period is consistent with medical practice in children, is manifestly a more precise programmatic definition of what is meant by the term "chronic" hematocrit of 55 percent or greater, and is not intended as a change in severity level. Moreover, we proposed the change through proper notice-and-comment rulemaking procedures and received no negative comments about the proposal aside from this lay comment.

Comment: The same attorney commenter from the preceding comment noted our statement in the preamble of the NPRM that we proposed to delete former listing 104.08, "Recurrent hemoptysis," because the condition could be evaluated under proposed listing 104.06E (56 FR 31269). The commenter said that, in fact, we had not mentioned hemoptysis (or bronchial collaterals, another criterion in the former listing) in proposed listing 104.06E. Another commenter asked whether nonvalvular stenosis should have been included in proposed listing 104.06F.

Response: We did not adopt the first comment. Hemoptysis is an uncommon manifestation of cardiovascular disease which occurs only in the presence of cyanotic heart disease or pulmonary hypertension of such seriousness that the primary condition would already have met or equaled the listings. Under final listing 104.06D, a child does not have to have an impairment so severe that it results in hemoptysis. Because of this fact, we have also deleted the criterion for recurrent hemoptysis in final listing 104.06A3; this, too, describes an impairment that exceeds listing-level severity.

We deleted the reference to bronchial collaterals for the same reason. Major collaterals (which generally are shown by cardiac catheterization) are only present in infants and children with severe cyanosis, occasionally combined with congestive failure, and such children will already have evidence that their impairments meet or are equivalent to one of the listings.

We adopted the second comment because we did intend to include nonvalvular stenotic lesions under this listing. We, therefore, revised final listing 104.06E (which was proposed listing 104.06F) to include congenital valvular or other stenotic defects, or valvular regurgitation. For consistency, we made similar revisions in final listings 4.07 and 104.07.

Comment: Another commenter noted that proposed listing 104.13 was essentially the same as prior listing 104.09A but that we had not included ECG findings as an example of an abnormal laboratory finding, as in the former listing.

Response: We have restored the example of abnormal ECG findings to final listing 104.13A. We had proposed to delete the example only because it is a minor criterion; it is not a specific requirement that abnormal ECG findings be provided in order to meet or equal the listing. The elevated sedimentation rate and abnormal ECG findings are only two of several minor criteria that can be used in establishing the diagnosis of continuing rheumatic activity. The use of elevated sedimentation rate to the exclusion of the ECG and other minor criteria was not intended to preclude the use or substitution of these other criteria for meeting the requirement that rheumatic activity be manifested by abnormal laboratory findings.

Comment: A national medical association suggested that we include a separate listing for Kawasaki syndrome. The commenter said that we should mention specifically the anatomical and arterial findings associated with the disorder and the chronic sequelae, including coronary lesions or valvular involvement.

Response: We have added a new listing 104.15 for Kawasaki syndrome. The listing will be met with Kawasaki syndrome associated with a major coronary artery aneurysm or with chronic heart failure of the level described in listing 104.02. The current treatment of Kawasaki syndrome with immunoglobulin therapy is thought to reduce the risk of the two major cardiac sequelae, coronary artery aneurysms and ventricular dysfunction. Listing-level impairment from either of these causes is rare, probably in the order of less than 5 percent of all infants affected, and it is unusual for the syndrome to be a cause of chronic cardiovascular impairment. However, in those instances in which such sequelae occur, other listings (such as listing 104.07, for valvular and other stenotic lesions) may be used for evaluation.