Response to Comments: Ambulatory Electrocardiograph (AECG) Monitoring

The commenter is concerned that the coverage criteria are too restrictive because they do not include all of the medically necessary indications. Additionally, the commenter feels that the relationship between the covered indications and each AECG service is vague leaving it open to interpretation. Comments were broken down into the following sections. Technology-Based Nomenclature: The commenter recommends: 1) Eliminating the nomenclature focused on AECG technology and adopting nomenclature that aligns with the AECG services as defined by Current Procedural Terminology (CPT) code. The recommendation was to place the CPT codes in the table with a list of devices associated with the CPT code; 2) Using the term "lead" for Holter monitors only and use "channel" for all modalities other than the 24-48 hour Holtor monitors; 3) Removing the language suggesting that single leads are not able to identify the location of an arrhythmia; 4) Providing clarification on the relevance of the limitations column in the table; 5) Clarifying that services using non-physician prescribed, hand/wrist held - smart phone based technology are not covered; 6) Revising the device table to include all labeled indications for each device and include daily symptoms for event monitors; 7) Clarifying that the choice of the AECG modality is at the discretion of the physician based on labeled indications of the technology and the patient's condition. Definitions: The commenter recommends revising the definition of a 24-hour monitoring station to remove the language “can direct the patient for the management of all emergencies” and replace it with “the technicians at the monitoring station must be able to contact the patient and the physician if an arrhythmia that meets prespecified criteria for being an [emergent] arrhythmia is identified.” Covered Indications: The commenter requests that the requirement of a 12-lead electrocardiograph (ECG) be clarified to indicate that there must be an ECG in the medical record, but that there is no requirement as to when that ECG is performed. Also, it was suggested that the the following covered indications be added: 1) All FDA cleared indications for each device; 2) Monitoring patients after radiofrequency ablation to determine whether the treatment was effective; 3) Monitoring patients to determine whether they need anticoagulation; and 4) Monitoring to determine the atrial fibrillation burden in order to determine whether the patient needs to have radiofrequency ablation. The commenter also suggested that Novitas clarify that performing a 24-48 hour Holter monitor is not a prerequisite or requirement for coverage of other types of AECG (e.g., event recorder, MCOT) and that they acknowledge that many clinically significant arrhythmias are asymptomatic and that extended AECG modalities are often necessary as initial tests to determine if asymptomatic arrhythmias are occurring. Provider Qualifications: The commenter requested clarification of the provider qualification statement to explain who "provider" is referring to, what is meant by "all aspects of care" and what is meant by "when all procedures are performed by appropriately trained providers in the appropriate setting." Documentation Requirements: The commenter would like the language "performed prior to initiation of cardiac event monitoring" when referring to the ECG be deleted. The commenter believes if an ECG was previously performed it is not medically necessary to perform another one prior to beginning monitoring. The commenter is seeking clarification on what is meant by the statement that the medical record must indicate the medical necessity for use related to the type of cardiac event monitoring device used. Their concern is that Novitas will require that providers use the lowest reimbursed test even if it is clinically more appropriate to use a higher reimbursed test. Clarification is requested regarding the intent of documentation #5 in the draft article. Specifically, the commenter is asking what type of documentation will be required. Several recommendations were made regarding documentation requirement #6: 1) The word "interpretation" should be removed because the technicians at the monitoring station do not interpret arrhythmias; 2) The word "lead" should be changed to "channel"; 3) The requirement for a signature should be removed since there are no signature requirements and requiring a signature would mean there would be over 30 signatures for each test; and 4) The seventh bullet in the draft article should indicate that in the absence of symptoms the technician should note that the patient did not record any symptoms. ICD-10-CM Codes that Support Medical Necessity: The commenter requested that the current ICD-10-CM codes that are included in A5660 and A52995 be maintained in this billing and coding article. In particular the commenter is requesting ICD-10-CM codes for unspecified arrhythmias. The following codes were requested: I44.30, I45.6, I45.9, I46.8, I46.9, I47.9, I48.20, I48.91, I48.92, I49.40, I67.841, I67.848, R00.0, R00.1, R06.00, R06.3, Z79.01

Thank you for your comprehensive comments. In response to your concerns with the table describing the different types of cardiac monitors, the Contractor has decided that it would be best to remove the table. Therefore, the table describing the types of monitors has been removed from the LCD and placed into the related billing and coding article. The intention of the LCD was not to suggest that the technicians in a 24-hour monitoring station could interpret rhythm strips. In order to provide clarification, the definition of 24-hour monitoring station has been revised. Regarding your comments on the covered indications; since the LCD does not state when the EKG should be performed, no change has been made to the standard EKG indication. When an arrhythmia is suspected, an EKG should first be performed and if there is no identifiable arrhythmia on the EKG then it would be reasonable to move to AECG monitoring. Additional covered indications have been added under the Covered Indications section and a note has been added in the Covered Indication section indicating that the type of AECG ordered should be based on the evaluation of the patient, the patient's symptoms and the device's indications. The LCD does not require that a Holter monitor be performed prior to other AECG monitoring. The expectation is that the provider will use the type of monitoring that is medically reasonable and necessary consistent with the patients symptoms, history etc. A limitation has been added to indicate that devices that do not have FDA clearance, including non-physician prescribed; hand/wrist held - smart phone based devices are not medically reasonable and necessary. Regarding the provider qualification section, "provider" refers to the physician or non-physician practitioner that is responsible for the patient's care and ordering the monitoring. Of note, per NCD 20.15 EKG services are covered diagnostic tests when there are documented signs and symptoms or other clinical indications for providing the service. Coverage includes the review and interpretation of EKGs only by a physician. The provider qualification section does not provide specific qualifications, it is suggesting that the services that are ordered/performed must be within the scope of practice of the provider ordering and/or performing them and that the provider must have the appropriate training. In response to your comments related to the documentation requirements: 1) Documentation requirement #3 has been reworded to advise that the diagnosis or clinical suspicion that led to the monitoring must be documented in the medical record. 2) There is no requirement in the LCD or the documentation requirements to use the lowest reimbursed monitoring device. While providers have the discretion to use the most appropriate monitor and they are not required to document why they did not choose another type of monitoring, the medical record should support the medical necessity of the device chosen. 3) For documentation #5 in the draft article (documentation requirement #4 in the final Article), whatever led the provider to order more than 24-hour monitoring should be reflected in the medical record. 4) The intention of documentation requirement #6 was not to infer that the technician at the monitoring station would be interpreting the strip, the requirements are for both the rhythm strip and the interpretation. Of note NCD 20.15 states; " A technician may be available at these centers to review transmitted data 24 hours per day. In some instances, when the EKG is determined to be outside certain pre-set criteria by a technician or other non-physician, a physician is available 24 hours per day to review the transmitted data and to make clinical decisions regarding the patient. " The language has been changed to try and provide clarification. The signature requirement has been removed and a new requirement was added to state that the medical record should include the name of the interpreting physician. Regarding the additional ICD-10-CM codes that you have requested, several of the requested ICD-10 codes represent unspecified diagnoses for specific conditions. Therefore, we would anticipate that the ordering provider would have a more specific diagnosis in these situations and/or that the patient would most likely have presented with symptoms leading to the monitoring in which case it would be more appropriate to report the symptoms since a diagnosis is not available. After further review, the Contractor has determined that the following requested ICD-10-CM codes are appropriate and they have been added to the final article: I46.8, I48.20, I48.91, I48.92, R00.0, R00.1 and R06.00. Once the LCD and related billing and coding article becomes effective providers may request additional ICD-10-CM codes if they feel there are codes that are supported by the indications of the LCD.

The commenter has concerns with the table in the LCD that explains the various types of monitors. They feel that the table is conflated and confusing and that it applies medical necessity criteria, requires excessive pretesting requirements and contains duplicative documentation submissions. The commenter mentions that the descriptors don't align with the CPT codes. It is recommended that the device types be aligned with the CPT codes. Additionally, the commenter is concerned that we are requiring a 12-lead ECG and a Holter monitor prior to another type of monitoring. The commenter recommends including two tables, one for cardiac monitoring devices and capabilities and one for cardiac monitoring services. Further, the commenter is concerned with the definition of 24-hour monitoring station language that states "and can direct the patient for the management of all emergencies". They recommend adding language directly out of the NCD in the LCD. There is also concern with the limitation regarding 24-hour attended monitoring stations as well as concerns listed in the covered indications for the requirement of a 12-lead ECG. When the clinical presentation is infrequent symptoms a 12-lead ECG and/or Holter would not likely be informative. The commenter is recommending the following additional covered indications: heart block, nocturnal arrhythmias such as those associated with sleep apnea or those associated with some channnelopathies, post-ablation, pre/post ICD reprogramming, stroke, syncope, dizziness, chest pain, palpitations, and shortness of breath. The commenter expresses concern that several pertinent publications are not included in the Summary of Evidence. One example given was the 2019 AHA/ACC/HRS Focused Update of the 2014 AHA/ACC/HRS Guideline for the Management of Patients With Atrial Fibrillation: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. It is recommended that Novitas consider the totality of peer reviewed publications as they are finalizing the LCD. Specifically mentioned were studies related to the Zio Service. Additionally, the commenter mentioned that there are pending articles describing the efficacy and health economics of extended or long-term Holter monitoring that they have been part of and that they would be happy to share the articles with the Contractor once they are published. There is also concern regarding the documentation for rhythm strip transmission and interpretation, specifically the lead of the AECG signal and the signature of the person interpreting the AECG strip as well as the documentation of symptoms at the time of transmission. The commenter requested clarification of these requirements and offered to further engage with the Contractor. Finally, the commenter is concerned regarding the elimination of several unspecified diagnostic codes and has requested that the codes be reinstated because monitoring services are diagnostic and the prescribing healthcare provider may not know the specific condition.

Thank you for your comments. Please refer to the response to comment #1 for your concerns related to the table with the devices and the 12-lead ECG requirement. Regarding the concern with the 24-hour monitoring stations, the Contractor agrees that the limitation needs clarification. The language has been revised to provide clarification. The Contractor has reviewed the additional covered indications that have been recommended and an indication for nocturnal arrhythmias, assessing the effectiveness of arrhythmia therapy, Pre/Post implantable cardiac defibrillator reprogramming, and syncope/dizziness/chest pain/palpitations/shortness of breath have been added. The indication for heart block was not added because the related billing and coding article contains diagnosis codes for heart block and the Contractor feels that heart block coverage would fall under covered indication #1 since heart block is generally diagnosed with standard ECG tracings. Regarding your concerns about the Summary of Evidence, the Contractor reviewed the 2019 AHA/ACC/HRS Focused Update and has added it to the bibliography for completeness. The other literature that is referenced in your comment is specific to the Zio Service. We performed a broad search for AECG monitoring services in general. If there is any peer reviewed literature that you feel would alter the indications/limitations of the LCD we encourage you to submit the full-text articles following the reconsideration process once the LCD is effective. The Contractor appreciates your concern regarding the removal of unspecified ICD-10-CM codes and as you mentioned, higher specificity is preferred. After further review, several unspecified codes have been added to the billing and coding article. For the other unspecified codes which are not included in the article, the Contractor believes they are for specific diagnoses for which the ordering provider would have the more specific information to code to the higher specificity and/or that the patient would most likely have presented with symptoms leading to the monitoring in which case it would be more appropriate to report the symptoms in the absence of a confirmed diagnosis. Once the LCD and related billing and coding article becomes effective, providers may request additional ICD-10-CM codes if they feel there are codes that are supported by the indications of the LCD.

The commenter supports the additional diagnoses and added indications for medically reasonable and necessary coverage in the Proposed LCD. However they have recommended that the A-Fib indication be broadened to include inherited cardiac arrhythmias and ischemic and non-ischemic cardiomyopathies. Further, the commenter recommends that indications listed elsewhere in the proposed LCD be listed in the Covered Indications section. The indications that are being recommended include syncope, dizziness, chest pain, palpitations, shortness of breath, post-ablation, stroke, nocturnal arrhythmias such as those associated with some channelopathies, heart block and pre/post ICD reprogramming. The commenter expressed concern about the pre-requisite requirements such as 12-lead ECG and 24-48 hour Holter monitor prior to using another type of AECG and has requested that those requirements be removed and that it be clear that it is up to the provider's medical judgement and training to select the most appropriate monitor. Additionally, the commenter offered concerns regarding the table describing the various types of monitors indicating that at times the information is vague and inconsistent with the FDA product description and indication. The commenter recommends that the table be removed and that focus be placed on the CPT codes and descriptions of the devices. The commenter also noted that 18 diagnosis codes are not included in the billing and coding article that are included in the current articles. The commenter explained that at the time of monitoring there may not be a specific diagnosis in which case an unspecified code would be appropriate. One example provided is for atrial fibrillation. The commenter is requesting that the following ICD-10-CM codes be added to the article: I44.30, I45.9, I46.8, I46.9, I47.9, I48.20, I48.91, I48.92, I49.40, I67.841, I67.848, R00.0, R00.1, R06.00, R06.3, Z79.01 and Z86.74. Concern regarding the definition of the 24-hour monitoring station was offered. The commenter disagrees that the staff at the monitoring station can direct the patient for the management of all emergencies, rather it is the physician who directs the patient as to next steps. Therefore, the commenter requests that the Contractor remove the language about the staff directing management of emergencies. Lastly, the commenter expressed concern that the documentation requirements are not well defined and that it is unclear as to what information is necessary to fulfill the requirements. For example, what type of information is required to support longer than 24 hour monitoring and documentation supporting the modality chosen. The commenter feels documenting the modality used should be sufficient and that the Contractor should not require physicians to document why they did not choose other modalities. In addition, the commenter would like to see the word "interpretation" removed from the rhythm strip transmission and interpretation requirement as well as the technician signature.

Thank you for your comments. Please refer to the responses to comments #1 and #2 regarding your comments on the covered indications, the pre-requisite requirements, the table of various devices, the definition of 24-hour monitoring station as well as the documentation requirements. After further review, the atrial fibrillation indication has not been expanded to include inherited cardiac arrhythmias and ischemic and non-ischemic cardiomyopathies. The Contractor welcomes the submission of full-text literature supporting the expansion using the reconsideration process once the LCD becomes effective. In response to your request to add the additional ICD-10-CM codes, please note that ICD-10-CM code Z86.74 was already included in the related draft billing and coding article. Please refer to the response to comment #1 for the other ICD-10-CM codes.

The commenter expresses concern with the matrix listing the monitor types in the LCD. The commenter stated that the categories do not map cleanly to the CPT codes, that it instructs that a 12-lead ECG is required, that cardiac event monitoring requires a 24-hour non-diagnostic Holter and that the categorization descriptors and coverage indications are not clear. The commenter requests that the matrix be revised to map to the CPT codes intended for each category. Under the Covered Indications Section the commenter requests that the following indications be added to ensure consistency with the currently covered indications: syncope, dizziness, chest pain, palpitations, shortness of breath, post-ablation, stroke, heart block, pre and post-ICD reprogramming. The commenter is also concerned about the unspecified ICD-10-CM codes being removed from the related billing and coding article because they are an IDTF and IDTFs rely on the diagnoses provided by the ordering physician. Further they state that it is reasonable that a diagnosis would be unspecified prior to monitoring. The commenter is requesting that the following ICD-10-CM codes be added to the related billing and coding article: I45.9, I46.8, I46.9, I47.29, I47.9, I48.20, I48.91, I48.92, I49.40, I67.841, I67.848, R00.0, R00.1, R06.00, R06.3. Lastly, the commenter is concerned about the documentation requirement that the medical necessity for the type of monitor used must be in the medical record. They feel this implies that there must be justification for one vs. another type of device, but the indications for the different devices overlap. Therefore, they are seeking clarification as to whether symptoms alone will suffice or if greater specificity is being sought. Additionally, if the intent is that the service would be disqualified if the report does not include all elements as laid out they would like this clearly stated. The commenter indicated that it would only be necessary to report the PR and QRS intervals if they are abnormal. The commenter would also like clarification as to whether the actions taken on abnormalities should be present on the final report or just in the medical record documentation.

Thank you for your comments. Please refer to the responses to comments #1 and #2 for your concerns related to the matrix listing the monitor types, the 12-lead ECG requirement and the covered indications. ICD-10-CM requires that providers code to the highest level of specificity. While the Contractor recognizes that for diagnostic testing there may be instances where an unspecified code may be appropriate, it is the provider of services responsibility to ensure that they are billing to the highest level of specificity. In response to several comments, the Contractor has added some unspecified codes to the billing and coding article. The Contractor believes that the diagnoses where unspecified codes were not added are specific conditions that the ordering provider should be able to bill to a higher level of specificity and/or that the patient would most likely have presented with symptoms leading to the monitoring in which case it would be more appropriate to report the symptoms. Please note that ICD-10 code I47.29 is included in the related draft billing and coding article. Refer to the response to comment #1 for further information regarding the ICD-10-CM codes. In response to your concerns related to the documentation requirements, several of the documentation requirements have been revised to provider further clarity. The intent of the documentation requirements is not to increase paperwork burden, rather the requirements are intended to provide guidelines to providers so that they ensure that the medical record supports the medical necessity of the service as outlined under Section 1862(a)(1)(A) of the Social Security Act. For any service reported to Medicare, it is expected that the medical documentation clearly demonstrates that the service meets the reasonable and necessary requirements.

The commenter is asking why the ICD-10-CM code for cryptogenic stroke is not included in the billing and coding article since several of the monitors can be used to assess for A-fib as a source of cryptogenic stroke.

Thank you for your comment. ICD-10-CM code I63.9 has been added to the billing and coding article with an asterisk note indicating that it should be reported for cryptogenic stroke.

The commenter feels that the proposed LCD is too general and needs to provide greater clarity and specificity. It is recommended that the classification of the products be structured by CPT code and that the table describing the types of monitors be deleted. The commenter is concerned about the removal of certain ICD-10-CM codes that were historically covered. They mentioned that it is standard practice for providers to prescribe a monitor to determine a more specific diagnosis and that unspecified codes are to be used when the medical record is insufficient to assign a more specific code. Further the commenter has suggested that the following indications be added to the Covered Indication section of the LCD: syncope, dizziness, chest pain, palpitations, shortness of breath, post-ablation, stroke, nocturnal arrhythmias, heart block, and pre and post-ICD reprogramming. The commenter indicated that the LCD should include a broader review of current evidence because the list is limited and contains only older studies. There were no specific articles/pieces of evidence suggested. The commenter expressed concern over the requirement for a 12-lead ECG and 24-hour Holter prior to other monitoring. They indicated that it would not make sense to require these tests for symptoms that are unlikely to appear in a 24-hour window. They stated that providers should be allowed to choose the appropriate technology based on the patient's condition, symptoms and labeled indications for the device. It was also noted that the commenter disagrees with the definition of a 24-hour monitoring station because the staff cannot direct the patient for the management of all emergencies. It was recommended that this definition be revised. Additionally, the commenter suggested that the LCD establishes an inappropriate burden of documentation for prescribers not consistent with CMS' goal of Patients over Paperwork. Finally, the commenter indicated that they agree with the LCD limitation that commercially available hand/wrist-held, smart phone-based devices should not be covered. The commenter requested a meeting to discuss the comments in greater detail before finalization of the proposed LCD.

Thank you for your comments. The majority of your concerns have been addressed in the prior responses. Please refer to the responses for comments one through five. In response to your concerns regarding the evidence, sixteen of the references in the LCD are within the last five to six years. The literature search for this LCD included a broad search for AECG monitoring in general and did not include articles for specific devices since the intent of the LCD is not to promote one type of monitor over another. Your comment did not include any suggested additional or newer articles. The Contractor is happy to review any additional literature that you may want to submit via the reconsideration process once the LCD becomes effective. We appreciate your offer to meet with us to discuss the comments in greater detail. At this time we do not feel that a meeting was necessary.

The commenter agrees with Novitas that patient-initiated, non-physician-prescribed ECG or event recorders without robust clinical report are considered pre-diagnostic and not designed to be acted upon without clinical review. We therefore support Novitas’ position that these are not considered appropriate services and should be non-covered within the context of this LCD. Additionally, the commenter concurs with the proposed LCD’s determination that coverage for AECG is medically reasonable and that providers determine the most appropriate AECG device consistent with the patients’ signs and symptoms. However, the commenter recommends that the term "patch recorders" be removed and replaced with "Long-Term ECG (LT-ECG)" consistent with CPT language. The commenter also recommends revising the language for Holter monitors in the table of devices to state 24-48 hours rather than 24-72 hours, again to be consistent with CPT coding. Further, the commenter recommends adding the following language under "Reports": LT-ECG reports “provide a summary of arrhythmic events with duration and frequency with reported symptoms” and that “used to assess arrhythmia sources for palpitations, dizziness, and pre-syncope; and used for detection of nocturnal arrhythmias such as those associated with sleep apnea” be added under "Indications" in the table. Additionally, the commenter recommends that the language regarding the possibility of electrical or motion artifact be removed from the table since the potential for interference is present in all AECG modalities. The commenter requests that in order to provide clarity, the device categories be listed with the relevant CPT code in the billing and coding article. The commenter recommends revising the definition for the 24-hour monitoring station to address only cardiac emergencies and that communications are made to physicians and not directly to patients. Further the commenter points out that 24-hour attended monitoring station applies only to external mobile cardiovascular telemetry. They are concerned that limitation #2 in the proposed LCD may be erroneously interpreted as any AECG without 24-hour attended monitoring as not being medically necessary. The commenter states that the ICD-10-CM diagnoses listing in the billing and coding article does not include all of the ICD-10-CM diagnoses for current FDA-cleared indications and that a number of unspecified ICD-10-CM codes are not included in the article that were included in the current Article A57476. It is requested that the ICD-10-CM codes for the FDA-cleared indications be added to the article and that the unspecified codes be added to the article because without these codes physicians are not able to order medically necessary AECG tests to further investigate a patient's underlying condition. A list of requested ICD-10-CM codes was not included with the comment. The commenter has expressed concern about the new documentation requirements and feels that these requirements will create unnecessary administrative burden for providers. Specifically the commenter requests that “or other abnormal test results which would indicate need for extended monitoring” be added to the documentation requirement that addresses the specific symptoms that should be in the medical record. In the Summary of Evidence the commenter requests that the evidence related to Holter monitors and LT-ECG recorders be separated since these are two distinct AECG modalities with different diagnostic yields. The commenter also provided a list of articles along with some proposed language for the Summary of Evidence. However, the full-text articles were not provided with the comments.

Thank you for your comments. Several of your comments are related to the table with the various devices. As mentioned in the above responses, this table has been removed from the LCD and additional indications have been added to the Covered Indications section in an attempt to provide further clarity. A table with the devices and the associated CPT codes has been added to the related billing and coding article. The definition of 24-hour monitoring station has been revised and we have added "cardiac" to the emergencies language as well as we added a parenthetical note that the 24-hour monitoring station is only applicable to mobile cardiac telemetry. Regarding your concerns with the ICD-10-CM codes please refer to the responses to comments #1 and #4. In response to your comment about documentation requirements for asymptomatic patients, the Contractor has added the requested language “or other abnormal test results which would indicate need for extended monitoring” to what is now documentation requirement #5. Please refer to the previous responses for additional information related to the documentation requirements. Lastly, the Contractor obtained the Rosenberg and Solomon articles and we have added your suggested summary for the Rosenberg article to the Summary of Evidence as well as adding both of these articles to the bibliography. The Freeman, Rho, and Smith articles listed in your references were not readily available to the Contractor. If it is felt that these articles would impact any of the LCD indications/limitations we welcome you to submit the full-text article following the reconsideration process once the LCD becomes effective. The May and Romme articles are related to children and therefore would not be pertinent to the majority of the Medicare population.