ASNC imaging guidelines for nuclear cardiology procedures

Introduction

The American Society of Nuclear Cardiology (ASNC) published a guideline for the reporting of myocardial perfusion imaging (MPI) in 2009.1 Over the last eight years there has been significant change in the breadth and depth of nuclear cardiology practice along with significant changes in the landscape of structured reporting. In consideration of this degree of change, it is appropriate that the guideline be updated and expanded to include a broader perspective of nuclear cardiology practice. At the same time, many things have not changed. This includes the fact that the report should provide a basic “bottom line” result to the referring physician and that this result must be clear and concise.2,3,4 This premise was expanded on by the American College of Radiology (ACR) with its development of a reporting and communication guideline with continued recent updates.5 All these documents emphasized the need for a defined structure containing standardized data elements to facilitate utilization of the complex data contained in an imaging report into the integrated healthcare of the patient through the electronic health record. The structured report is also an integral part to define quality in nuclear cardiology practices . There continues to be interest in the implementation of structured reporting as a mechanism to improve quality and outcomes and to reduce cost in fulfillment of the triple aim.

Since the publication of the prior guideline there have been significant developments in the field of nuclear cardiology. Examples of this include the development of the ImageGuide^TM Registry by ASNC, the development of additional registries for imaging internationally, the expansion of nuclear cardiology into greater utilization of positron emission tomography (PET) imaging, and new protocols for imaging inflammation, viability, and innervation.6 These additional areas of interest will be addressed in this updated guideline for nuclear cardiology procedure reporting in contrast to the prior document that was limited to perfusion imaging only.1 There is also new emphasis on the concept of interpreting the interpretation. Research regarding this important aspect of result utilization has focused on how the referring physician incorporates the report data to affect care and the differences between the referring physicians approach and the imaging physicians anticipated response to the report.7 This will become an increasingly important area of information science in the future. To help meet the needs of the referring physician, the appearance of a standardized report can and should vary from user to user. There should not be a single standard appearance of a report but one that best conveys the content to the end user. This may be in paragraph form for some laboratories while others might use a table or even a list of structured data elements. All would meet the guidelines for structured reporting as they are derived from defined structured data elements as outlined in this guideline.1,8

An essential part of structured reporting is the ability to use and incorporate other standards to facilitate data sharing among many different sources. These standards include the Digital Imaging and Communications in Medicine (DICOM) and the Integrating the Healthcare Enterprise (IHE) standards. The DICOM standard for stress reporting includes the data elements for structured nuclear cardiology reporting.9,10 The use of the DICOM elements has been integral to the clinical implementation of reporting software by both developers and manufacturers. This is supported through the utilization of the IHE standards for communication of data among different vendor systems and single and multimodality imaging environments.11,12 The data from this new IHE standard have been incorporated into this document.

Two important documents were utilized in the development of the first nuclear cardiology myocardial perfusion imaging reporting standard and remain important and relevant today. The American College of Cardiology (ACC) “Health Policy Statement on Structured Reporting and Cardiovascular Imaging” and the “Key Data Elements and Definitions for Cardiac Imaging: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Data Standards” remain as sentinel documents and facilitate the reporting of imaging studies in multimodality environments.13,14 In addition to the ACC documents, the European Association of Nuclear Medicine and the European Association of Cardiovascular Imaging have published a guideline regarding reporting nuclear cardiology.15 This important guideline addresses an update to the standards and serves as a guidepost as we move forward to standardized structured reporting internationally. The development of the ImageGuide^TM Registry for myocardial perfusion imaging has also been the cause for some redefinition of the data elements that were present in the prior version of the myocardial perfusion imaging study reporting standard. This updated image reporting guideline incorporates and harmonizes the recommendations of all these guidelines and unifies ASNC documents that have been published since the prior reporting guideline.

As with the prior document, this guideline consists of tables composed of the variables, their description (i.e., text, numeric, date), priority (i.e., required, recommended, or optional), and the allowed response(s). With regards to the allowed responses to numerical values, the writing group acknowledges that different units of measurement can be used to express the same value, such as millicuries (mCi) and megabecquerels (MBq). As this guideline is intended for international use, both traditional English units of measure and their metric equivalents are acceptable responses. It is required, however, that the user be consistent throughout the report regarding the system of units utilized. Acceptable units of measure are outlined in Appendix 1. As the structured report may be used to populate data in registries, such as ImageGuide™, it is a requirement of the registry submission process to provide the appropriate conversion factors from the structured report data to assure compliance with the allowed format from the registry’s data dictionary. Finally, examples of sample structured reports from numerous laboratories around the United States are incorporated in the appendix as a resource for the reader.

As was noted in the prior document, ASNC continues to support the mandatory use of structured reporting as a mechanism to improve the communication and reporting of nuclear cardiology reports. This has begun to be incorporated into the laboratory accreditation process, and there has been significant improvement over the course of eight years. There remain significant areas for improvement, particularly with regards to defect size and severity, and consistent reporting of these important variables.16 This guideline is designed to provide imaging physicians and technologists the necessary information to report nuclear cardiology procedures in a structured format using standardized data elements. While the content of the document has been carefully reviewed by many experts, the document should not be considered as a source of medical advice or professional service.

Table of Contents for Structured Data Elements

Site administrative data	Table 1
Study demographics
Patient demographics and study referral data	Table 2
Clinical Information	Table 3
Appropriate use reporting
Study description
Stress testing data	Table 4
Resting ECG data	Table 5
Stress ECG data	Table 6
Imaging data
Imaging parameters	Table 7
Additional imaging parameters specific to viability studies	Table 8
Imaging parameters specific for inflammation/infection	Table 9
Imaging parameters for Tc-99m PYP	Table 10
Qualitative LV perfusion assessment (SPECT and PET)	Table 11
Quantitative LV perfusion assessment (SPECT and PET)	Table 12
LV gated function volume assessment at stress	Table 13
LV gated functional and volume assessment at rest	Table 14
Additional PET-specific LV perfusion and function parameters	Table 15
Right ventricular perfusion and function parameters	Table 16
Miscellaneous data	Table 17
FPRNA/ERNA (rest and exercise)	Table 18
Viability—qualitative analysis	Table 19
Viability—quantitative analysis	Table 20
Inflammation/Infection—qualitative parameters	Table 21
Inflammation/Infection—quantitative parameters	Table 22
mIBG analysis parameters	Table 23
Tc-99m PYP analysis parameters	Table 24
Coronary artery calcium score analysis parameters	Table 25
Overall impression	Table 26
Combined conclusion	Table 27
Comparison to prior studies	Table 28
ImageGuide Registry CMS reported performance measures	Table 29

1. ECG, electrocardiographic; LV, left ventricular; RV, right ventricular; FPRNA, first-pass radionuclide angiography; ERNA, equilibrium radionuclide angiocardiography

Structured Reporting

Components of the Report

According to the ‘‘Health Policy Statement on Structured Reporting in Cardiovascular Imaging,”13 the standard components of a report include the following major headings: Administrative Information, Patient Demographics, Study Referral Data, History and Risk Factors, Study Description, Study Findings, and other reporting parameters. These elements are outlined in detail in ‘‘Key Data Elements and Definitions for Cardiac Imaging: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Data Standards,”14 which addresses specific details for each of these major headings for multiple cardiac imaging modalities and these remain unchanged from the prior document.

A few of the general data elements, and many of the specific data elements, may be recorded at the time that the test is performed. Some elements may not be required in the final report. This may be the case for some fields that are required for quality reporting, but not necessarily for reporting the findings from an individual patient’s study for specific patient management.

Many different structured reports can be generated from a set of structured data. The potential reports include: a clinical patient-specific report, summary quality report, billing report, reporting the data to registries, and other reports as needed. The greatest strength to structured data utilization is the ability to generate multiple report formats with varying levels of detail depending on the clinical or administrative need.

This document will harmonize these generalized concepts and apply them specifically to nuclear cardiology. Due to the variability of the study types encompassed by this document, some of the data elements are specific to certain types of acquisitions, or are dependent upon the study indication (e.g., viability determination by PET imaging). Therefore, some data elements may be required for certain acquisitions and clinical indications, while some may be optional or perhaps irrelevant for other indications.

A number of the data elements contained in the tables have been derived from, and harmonized with, other guideline documents, some multisocietal and others ASNC-specific.3,4,17,18,19,20,21 This update also addresses additional modalities that were not included in the prior versions of the document, such as: broader treatment of PET and viability, and non-perfusion imaging including amyloid detection, inflammation/infection, MIBG in heart failure and coronary calcium scoring, and its incorporation into the nuclear cardiology report. The data elements required for reporting the additional modalities have been added to specific tables where appropriate or additional tables have been added to the document to cover those items that were specific to the modality and could not be generalized to one of the existing table headings. Finally, a perspective on the future direction of nuclear cardiology reporting has been included as a guidepost for the future.

Site Administrative Data

The Site Administrative Data section of the report is the descriptor of the site performing the study. It includes elements such as the physical address, accreditation status, type of facility (e.g., hospital or office), and insurance payer. These data may only need to be collected as part of the reporting process, and some elements may not be recorded in the final report. Some elements may be necessary to inform registry submission of the data and as part of the quality initiatives as we transition from volume-to-value-based practice (Table 1).

Table 1 Site administrative data

Patient Demographics and Study Referral

The Patient Demographics and Study Referral data section provides the clinical indications for the study, information regarding the referring and interpreting provider in addition to the necessary demographic information that could impact the clinical outcomes of the study. Indications to be considered include the following major areas: diagnosis of coronary artery disease (CAD), extent and severity of known CAD, risk stratification including peri-operative risk, determination of viability, assessment of acute chest pain syndromes, evaluation of structural heart disease, and heart failure. The table also allows for a secondary indication to be selected. With the inclusion of the History and Risk Factors section, this would complete the data elements contained in Tables 2 and 3.

Table 2 Patient demographics and study referral data

Table 3 Clinical information

The specific purpose for which the test is being performed must be clearly identified. This provides the required documentation for the medical necessity of the study and focuses the report on the question asked by the referring physician. The structured data elements that relate to the indication are in Table 3. The structured reports must contain sufficient information from these areas to ensure correct identification of the patient. The reports must also convey the specific indications for the study and the pertinent portions of the clinical history that allow the caregivers to appropriately place the imaging results in clinical context. This would include the patient’s current symptoms or other indication for which the study is being performed, current medications, cardiac history with pertinent risk factors including risk factoring scoring, and prior testing, and therapeutic procedures.

Appropriate Use Reporting

Greater emphasis including elevating to required status for reporting AUC has been a significant change in this document. In response to rapid and unsustainable growth in utilization of radionuclide MPI, professional medical organizations developed appropriate use criteria (AUC) to guide physicians and payers on the effective use of these procedures.26 Based on symptoms, coronary risk factors, and cardiac history, the AUC classifies testing across a range of clinical scenarios in three categories: appropriate (established value), may be appropriate (uncertain value), and rarely appropriate (no clear value).25 A significant body of literature demonstrated that appropriate MPI use enhances its acumen in risk stratification, reduces radiation risk, and improves its clinical value.27,28,29,30,31,32,33 Physicians are faced with multiple, occasionally discordant, AUC from different organizations. For example, there is substantial discordance between the multimodality AUC for the detection and risk assessment of stable ischemic heart disease developed by the American College of Cardiology, ASNC, and several other societies and the Appropriateness Criteria set forth by the American College of Radiology (ACR). ASNC recommends the AUC promulgated by the ACC as they are best validated and have been shown to be more effective in guiding providers toward patients with greater potential for myocardial ischemia than the ACR Appropriateness Criteria.34

For the past decade, AUC has been promoted as a tool to optimize value of imaging studies. Many health organizations have implemented measures to reduce rarely appropriate studies as an academic or quality improvement exercise. Despite the importance of AUC in the clinical domain, documentation of adherence to AUC in the clinical reports has not been required or widely performed. This will change soon. The Centers for Medicare and Medicaid (CMS) is in the process of implementing §218 of Protecting Access to Medicare Act (PAMA) of 2014. As of 2018, this legislation will require the ordering physician to consult AUC using a CMS-approved, computer-based decision support tool (DST) when ordering MPI studies.35 Thus far, CMS has approved many qualified professional organizations that have developed or endorsed applicable AUC; among these, the ACC’s AUC.25 CMS finalized eight “priority clinical areas,” which will be used to benchmark providers according to their use of rarely appropriate imaging procedures. These clinical areas include suspected or diagnosed coronary artery disease, suspected pulmonary embolism, headache, hip pain, low back pain, shoulder pain, suspected or diagnosed lung cancer, and neck pain. Suspected or known CAD being a “priority clinical area,” the majority of MPI studies will be used to benchmark the ordering physician.35 Based on PAMA, the imaging specialists will not be paid for their services if they do not have documentation that the ordering physician consulted an AUC DST. After collecting two years of data in the aforementioned eight priority clinical areas, referring physicians who are considered “outliers” in terms of their utilization of rarely appropriate MPI will be subjected to prior authorization when ordering MPI studies. As a result, there will be a massive shift wherein the burden of reducing inappropriate use will move largely from payers to providers.36 Imaging specialists, practicing physicians, and health organizations need to adapt to meet this requirement. Nuclear cardiologist need to find practical ways to obtain and document AUC determination, as discerned by a CMS-approved DST used by the ordering physician.

Study Description

The Study Description should be the next section of the structured report. This section should include all the parameters used in acquiring the study. It must include a description of the stress test performed, including the type of stress test (i.e., exercise or pharmacologic). For stress tests, it is necessary to include the type of protocol, duration of exercise, and its adequacy as determined by exercise time, peak heart rate, percent maximal predicted heart rate (MPHR), pressure rate product (PRP), and estimated metabolic equivalents (METS). For pharmacologic stress tests, the pharmacologic agent used, the dose received, including the infusion rate and duration, hemodynamic response to the dose, and use of adjunctive exercise must be documented. If pharmacologic stress is performed after attempted exercise, exercise parameters should be reported in addition to pharmacologic parameters. The time of administration of radioactivity is also required for either modality. The specific data elements for this section as well as their responses are found in Table 4.

Table 4 Stress testing data

The electrocardiographic (ECG) data pertinent to the test should be reported next. This would include the presence of any baseline ECG abnormalities that might preclude a conclusive interpretation of the ECG stress portion of the test (Table 5).

Table 5 Resting ECG data

The stress ECG interpretation must evaluate the parameters defined in Table 6, commenting on any changes from baseline with regards to either the ST segments or onset of arrhythmias. Comparison to prior tests and inclusion of parameters that allow calculation of validated risk scores (e.g., the Duke treadmill score)37 are recommended. Ideally, Stress ECG data would be presented in a tabular format, with documentation of many of the following variables at each stage of stress and recovery.

Table 6 Stress ECG data

The structured report format continues with variables that define the imaging process including the protocol utilized, the patient position, and radiopharmaceutical doses administered to the patient. It also includes their time of administration and whether attenuation correction or other modalities were used. These data elements are presented in detail in Tables 7, 8, 9, and 10.

Table 7 Imaging parameters

Table 8 Additional imaging parameters specific to viability studies

Table 9 Imaging parameters specific for inflammation/infection

Table 10 Imaging parameters for Tc-99m PYP

Following the section on imaging parameters, the left ventricular (LV) perfusion results should be provided. The results will differ slightly for SPECT vs PET MPI. Every qualitative assessment of LV perfusion should include a summary that provides an overall statement of LV perfusion abnormality. This should be followed by the size, location, severity, and degree of reversibility of any perfusion defects as shown in Table 11. Perfusion defect location should be described according to the standardized 17-segment model (Appendix 6). This pattern can be repeated for multiple perfusion abnormalities. Inclusion of a bulls-eye polar plot showing the location and degree of perfusion defects can aid in visualization. The associated segmental function of myocardium with a perfusion defect can inform the clinical interpretation. A clinical interpretation of each perfusion defect provided in this portion of the report can help increase clarity (ischemia, infarction, peri-infarct ischemia). Any uncertainty can be reported here. For instance, probable ischemia (vs artifact) can be selected when perfusion is probably abnormal or probable artifact can be chosen if perfusion is categorized as probably normal. Classification of the perfusion defect as visual only, quantitative only, or visual and quantitative is optional but provides additional information on the degree of evidence to support the conclusions made. The presence or absence of transient ischemic dilation (TID) is a required element and can also be classified as visual, quantitative, or both. Reporting of the stress and rest perfusion cavity sizes and ratio of these two parameters (the TID ratio) are optional. The presence of normal LV tracer uptake and myocardial wall thickness vs increased values in the setting of LV hypertrophy should be documented. Finally, increased tracer uptake in the right ventricle and the lungs at stress and rest can be reported.

Table 11 Qualitative LV perfusion assessment (SPECT and PET)

Quantitative image processing for LV perfusion is recommended, with suggested data elements outlined in Table 12. Each segmental score should be adjusted for attenuation prior to calculation. No segment should have a negative score. The derived extents of perfusion and ischemia require division of the respective SSS, SRS, and SDS by 68, the maximal perfusion score of 4 across all 17 segments.

Table 12 Quantitative LV perfusion assessment (SPECT and PET)

Stress and/or rest-gated imaging should be performed when technically feasible. LV global and segmental function and volumes should be reported as detailed in Tables 13 and 14. The timing of stress function assessment (during stress [i.e., first-pass], post-stress, rest) is recommended. The following values can be repeated for each phase assessed (stress and rest). An overall assessment of global LV function is required, and the calculated left ventricular ejection fraction (LVEF) should be provided. Segmental functional abnormalities can be described both by regional thickening and wall motion. Severity should be described by location according to the 17-segment model.17 Numerical documentation of LV volumes and/or volume indices and subjective assessment of the LV cavity sizes at both end-diastole and end-systole are optional. The information in these tables may be repeated as required to describe multiple perfusion defects.

Table 13 LV gated functional and volume assessment at stress

Table 14 LV gated functional and volume assessment at rest

LV perfusion and function assessment by PET has additional parameters not typically assessed in SPECT studies that can be reported as shown in Table 15. Stress and rest myocardial blood flow (MBF) can be quantitated during PET MPI and can provide additional information on LV perfusion. Values are typically provided for stress and rest globally and by coronary perfusion territory (left anterior descending [LAD], left circumflex [LCX], right coronary artery [RCA]). The ratio of stress to rest flow is defined as the myocardial flow reserve. Stress MBF and MFR can be classified as preserved (>2 mL/min/g), mildly reduced (1.5^-2 mL/min/g), or severely reduced (<1.5 mL/min/g).20 Thresholds for MBF and MFR can vary by protocol and lab. The calculation of a true stress LVEF during vasodilator stress has led to calculation of LVEF reserve, the difference between stress and rest LVEFs that has diagnostic and prognostic significance. An LVEF reserve <0%, indicating a drop in LVEF with stress, has diagnostic and prognostic significance and can be optionally reported.38

Table 15 Additional PET-specific LV perfusion and function parameters

SPECT and PET MPI also allow interpretation of the perfusion, size, and global and segmental function of the right ventricle (RV). Data elements for this assessment are provided in Table 16. These parameters are not typically reported unless abnormal or in the presence of specific indications for their assessment.

Table 16 Right Ventricular Perfusion and Function Parameters

There are several miscellaneous factors that should be present in the report and will be detailed in Table 17. Comment on the overall study quality can assist in study interpretation and serve as a quality reporting mechanism for the nuclear laboratory. Appreciated artifacts seen on the primary MPI images and CT attenuation correction images should be documented. Increased lung uptake can be commented on, particularly in the setting of Thallium administration. Finally, any incidental findings should be documented including from any associated CT attenuation correction images.

Table 17 Miscellaneous data

FPRNA and ERNA

FPRNA and ERNA utilize a number of variables included in other tables, such as those describing LV and RV function at rest and with exercise. Some variables, however, are not covered adequately and are not assignable to other existing tables. Table 18 describes the variables that are recommended for FPRNA and ERNA at rest or with exercise. The majority of the variables in Table 18 are optional, with the required elements noted at the top.

Table 18 FPRNA/ERNA (rest and exercise)

Viability Imaging

Viability reporting should detail imaging parameters including patient dietary state; glucose loading or use of the euglycemic-hyperinsulinemic clamp; radiopharmaceutical dose; time of viability imaging; and time delay from injection of radiopharmaceutical to imaging (Tables 7 and 8). Resting left and right ventricular perfusion and function should be described according to parameters listed in Tables 11, 12, 14, and 16.

Assessment of myocardial viability should include visual and quantitative analysis. Metabolism defects, perfusion/metabolism matched defects, and perfusion/metabolism mismatched defects must be described with regards to location, size, and severity.20 The remaining elements in Table 19 are recommended for use in reporting myocardial viability.

Table 19 Viability—qualitative analysis

The use of quantitative image elements (i.e., number of viable segments and extent of matched and mismatched defects) is also recommended. Table 20 outlines the quantitative data for myocardial viability.

Table 20 Viability—quantitative analysis

Inflammation and Infection Imaging

Inflammation and infection imaging is based on increased glucose metabolism by activated immune cells.39 In inflammatory conditions (e.g., cardiac sarcoidosis, myocarditis) and infection (e.g., endocarditis, cardiac implantable electrical device [CIED] infections), immune cell activation and infiltration into the myocardium can be visualized by uptake of F-18 FDG, a glucose analog. An important aspect of imaging infection and inflammation is suppression of physiological cardiomyocyte uptake of glucose, so upon injection of F-18 FDG, uptake of the radiopharmaceutical is limited to inflammatory cells.20,40 Reporting should include patient preparation relevant to the suppression of physiological cardiomyocyte glucose uptake as well as abnormal uptake of F-18 FDG (Table 21).

Table 21 Inflammation/infection—qualitative parameters

Assessment of myocardial inflammation includes both visual and quantitative analysis. For sarcoidosis imaging, rest perfusion imaging is required for co-localization of F-18 FDG images with the myocardium and to evaluate for the presence of active inflammation.20,41 Current guidelines do not require myocardial perfusion images for the imaging of cardiovascular device or prosthetic infections.20 Reporting of left ventricular resting perfusion should follow the recommendations set forth in Table 12 of this document. Table 21 lists the qualitative parameters recommended for use in reporting myocardial inflammation and/or infection. The use of quantitative measurements for myocardial uptake of F-18 FDG and for measurement of blood pool (background) activity is summarized in Table 22.

Table 22 Inflammation/infection—quantitative parameters

Iodine-123 metaIodobenzylguanidine (I-123 mIBG) Imaging

Reporting metaiodobenzylguanidine (mIBG) imaging should include visual and quantitative analysis. Decreased mIBG uptake and heart-to-mediastinal ratio (HMR) are key components of I-123 mIBG imaging and should be clearly stated in the report.42 Calculation of washout and specific localization of sympathetic activity defects may also be included.43,44,45 The remaining elements in Table 23 are recommended for use in reporting mIBG imaging.

Table 23 mIBG analysis parameters

Tc-99m Pyrophosphate Imaging for Transthyretin Cardiac Amyloidosis

There is increasing use of Technetium 99m pyrophosphate (Tc-99m PYP) imaging to diagnose cardiac transthyretin amyloidosis (ATTR).46,47 The American Society of Nuclear Cardiology published a Practice Points statement detailing the critical components of Tc-99m PYP imaging and reporting.48 Reports should include semi-quantitative and quantitative analysis of cardiac uptake of Tc-99m PYP in addition to visual scan interpretation (Table 24). The report should include all applicable elements of a nuclear cardiology report as detailed in Tables 1, 2, 3, 7, and 10 of this guideline.

Table 24 Tc-99m PYP analysis parameters

Coronary Artery Calcium Scoring

Coronary artery calcium score, if performed with SPECT/CT or PET/CT imaging, should be reported quantitatively and by percentile ranking based on age and sex (Table 25).49,50

Table 25 Coronary artery calcium score analysis parameters

Section on Overall Impressions

The overall impression is the most important portion of the nuclear cardiology report, as it assimilates and summarizes the most important details presented in the preceding sections. Data elements specific to this section are outlined in Table 26. Summaries of LV perfusion, function, and viability (when indicated) should be provided with clear indication of normal vs abnormal findings. For perfusion defects, a statement of whether these findings indicate ischemia, infarction, or both should be provided. This information may have been provided in preceding sections but should be highlighted in the overall impression. The number of coronary territories involved and possibly even specific vessel territories can be indicated, though caution should be advised in correlating perfusion results to coronary anatomy in the absence of prior invasive or CT coronary angiography to precisely define the epicardial distributions. For positive studies, it is recommended that a statement be made regarding the significance of the LV perfusion results. The overall impression should also contain additional statements from the body of the report if additional emphasis is needed. For instance, if transient ischemic dilation or significant RV perfusion or functional defects are present, these should be mentioned. Furthermore, to ensure timely access to the data, the report needs to be compliant with the standard for timely reporting requiring completion of the interpretation within one business day and transmittal from the lab to the referring physician within two business days.51

Table 26 Overall impression

Conclusion and Communication of High-Risk Results

An important additional component of the overall impression section is a combined conclusion that incorporates results from both imaging and the stress test, including the electrocardiogram, hemodynamics, and stress-induced symptoms. It is also important to note discordant results between perfusion and non-perfusion imaging results, such as normal perfusion and increased lung uptake. As detailed in Table 27, combining the results is straightforward when the ECG and imaging are concordant. Likewise, when the studies are discordant with abnormal imaging, the combined test is typically treated as abnormal. However, the combined conclusion is more challenging when there are discordant results with a positive stress ECG and negative imaging. One solution is to categorize the cardiovascular risk as low, intermediate, or high. This is difficult if the reader is not the ordering physician. Detailing supporting clinical information used to classify the risk (such as young age or atypical presentation for low risk and stress angina or high-risk ECG findings such as multiple millimeters of persistent ST depression for intermediate or high risk) can inform the referring physician of the parameters considered even when the reader has not seen the patient. A clinical recommendation can then be offered based on the risk classification. A low-risk designation could suggest that further cardiac evaluation may not be necessary. Intermediate and high-risk designations could suggest that further cardiac evaluation “could” and “should” be considered, respectively.

Table 27 Combined conclusion

A complete report should include documentation of the communication of high-risk results, including what findings were communicated, the person to whom they were communicated, and the date and time of the communication.

A section comparing the current imaging to prior studies is recommended in all reports as shown in Table 28. The date of the study being compared should be provided, and a statement of whether there are new changes or if the imaging is unchanged. Changes in perfusion and function should be detailed, with comment on both changes in LVEF and segmental function. A statement on the clinical significance of the changes should be provided.

Table 28 Comparison to prior studies

Future Directions

Available and evolving technology solutions can ameliorate the burden of comprehensive nuclear cardiology reporting and further enhance the value of the report in providing diagnostic, prognostic, and decision-guiding information, while meeting all regulatory requirements. Taking full advantage of these technology tools will facilitate evidence-based and patient-centered reporting.

Structured Reporting Software

Providing high-quality medical care and satisfying all guidelines and regulatory requirements is ever more complex; this certainly applies to nuclear cardiology reporting. Building new habits to satisfy all reporting elements is rather difficult. Using structured reporting software with hard-wired, guideline-driven reporting standards as well as built-in reminders and hard-stops for high importance reporting elements would ensure a complete and informative report every single time. Structured reporting packages can be fitted with DSTs capable of exploiting the wealth of objective clinical, stress, ECG, perfusion, functional, and ancillary data (chamber volumes, mass, and TID) to produce diagnostic and prognostic assessment using a catalogue of widely accepted nuclear cardiology literature. These determinations can be translated into hard-wired, evidence-based, and patient-centered diagnostic, prognostic, and decision-guidance statements. Furthermore, structured reporting software can facilitate reporting to accreditation bodies, automate data entry in public registries, aid in conducting research and quality improvement initiatives, and track radiation dose and critical findings.

Structured reporting software packages vary in their quality, ease of use, and comprehensiveness. They also vary in terms of their ability to auto-populate readily available data in electronic health records, previous testing reports, and stress testing data. Commonly used nuclear cardiology analysis software packages are fitted with structure reporting capabilities. Other structured reporting software can import and auto-populate imaging data from nuclear cardiology analysis packages and stress testing data from the treadmill computer console. Finally, structured reporting software may facilitate the generation of all-encompassing nuclear cardiology reports by combining separately interpreted stress and imaging data while maintaining two provider signatures: a cardiologist (stress portion) and an imaging specialist (nuclear portion). Unfortunately, structured reporting software packages are not universally used across various practice settings. ASNC recommends the use of structure reporting packages to ensure comprehensive nuclear cardiology reporting to optimize decision-making and facilitate continuous quality improvement through accreditation and public reporting.

Decision Support Tools (DST)

Computer-based DSTs can complement nuclear cardiology reporting on two main levels.

• (1) Discerning Appropriate Use: Computer-based DST can mine data readily available in electronic health records in discerning appropriateness of MPI, and when testing is rarely appropriate it can provide guidance on appropriate alternative testing, for example, exercise tolerance test (without imaging) instead of stress MPI. Deep integration of DST in the electronic order entry in electronic health information systems can provide seamless, real-time guidance on study appropriateness with minimal provider burden. AUC adherence data can then seamlessly flow into interconnected electronic structured reporting software and hence to the clinical report. Such practical technologic applications can be easily developed to enhance adherence to AUC, improve value of imaging, and facilitate compliance with PAMA requirements.

• (2) Risk assessment and Guiding Decision-Making: Structured reporting software can be fitted with DST that can leverage the wealth of objective clinical, stress, ECG, perfusion, functional, and ancillary data in the nuclear cardiology study to provide individualized diagnostic and prognostic statements using a catalogue of widely accepted nuclear cardiology literature. Specific examples of such statements: (1) No history of CAD or diabetes mellitus, normal exercise stress MPI and ejection fraction, and no TID: Patient is at <1% annual risk for major adverse cardiac events; (2) Abnormal MPI and abnormally high TID ratio: Perfusion imaging is predictive of multi-vessel CAD and increased risk of adverse cardiac events; (3) Normal MPI but abnormal heart rate response to vasodilator stress agent: Patient is at increased risk of mortality and adverse cardiac events; (4) Ischemic myocardial perfusion deficit 15%: observational outcome data favor coronary revascularization over medical therapy (if clinically indicated and feasible); (5) Ischemic myocardial perfusion deficit 5%: observational outcome data favor medical therapy over coronary revascularization. In such fashion, structure reporting software can be leveraged to hard-wire evidence-based and patient-centered diagnostic, prognostic, and decision-guidance statements. Decision support in nuclear cardiology reporting can be further enhanced by applying machine learning algorithms.

Machine Learning

The interpretation of MPI is currently performed primarily by experienced readers who mentally combine clinical, ECG, stress, perfusion, and functional data to generate an overall diagnostic and prognostic impression. However, this interpretation is primarily subjective, semi-quantitative, and heavily dependent on reader’s wealth of knowledge, acumen, and experience.52 Furthermore, traditional prognostic risk assessment in patients undergoing nuclear cardiology imaging is based on a limited menu of clinical and imaging findings. Many of these findings are continuous variables (ejection fraction, chamber volumes, TID, SSS, etc.) that are difficult to incorporate in a simple diagnostic or prognostic determination.

Machine learning can consider a greater number (dozens) and complexity of variables and correlate them with specific outcomes in very large training datasets. These machine-learned algorithms are validated in testing datasets before they can be applied clinically.53,54 Unlike multivariate regression modeling, machine learning algorithms are not fitted models, and thus are not affected by collinearity between variables. Furthermore, they can be improved in an ongoing basis incorporating accumulative observations after clinical implementation. It has been shown that machine learning algorithms derived from integrating clinical, perfusion, and functional data elements for diagnosis of obstructive CAD yield results similar to or better than those obtained by experienced readers.55 Furthermore, machine learning applications, integrating clinical, ECG, exercise, hemodynamic, defect quantification, and ancillary imaging data provide a patient-specific estimate of likelihood of early revascularization and all-cause mortality, thus aiding in individualized decision-making in a way the human brain cannot do.53,56

Machine learning algorithms are a natural complement to nuclear cardiology analyses packages and structured reporting software, from which multi-faceted data can be derived to generate risk estimates factored in DSTs and patient-centered decision guidance.

Registries and Public Reporting

ASNC’s ImageGuide^TM Registry is the first registry of its kind focusing on SPECT and PET imaging. The primary purpose of the registry is quality improvement. It provides a fully integrated platform to seamlessly collect data from nuclear imaging laboratories to measure quality, safety, and efficiency. The registry contains hundreds of data elements such as referral information, demographics, clinical data, stress data, ECG data, imaging parameters, radiation dosing, perfusion, quantification, left ventricular function parameters, study quality, and signature date/time.54 Data elements in structured reporting applications within commercially available nuclear cardiology analysis packages are fully homogenized with the ImageGuide^TM. Thus, data from each study can be easily submitted from the laboratory to the ImageGuide^TM Registry, which in turn tracks and publicly reports, in real-time, indicators of excellence in radionuclide imaging, including crucial reporting measures.16,54,55 Such integration provides a constant quality improvement feedback loop for ever-improving report quality and patient care.57

The ImageGuide^TM Registry is a Qualified Clinical Data Registry (QCDR) through which participating physicians can receive CMS reimbursement credits for participating in a Physician Quality Reporting System (PQRS). Physicians satisfactorily reporting on a minimum of 9 CMS-approved quality measures can avoid reimbursement penalties based on the Merit-Based Incentive Payment System (MIPS). Table 29 lists 2017 CMS-approved nuclear cardiology quality measures. The ImageGuide^TM Registry and CMS yearly update the reported quality measures, such that old, highly achievable measures are retired and new measures are introduced in a sustained effort to continuously improve the quality of nuclear cardiology studies.

Table 29 ImageGuide ^TM CMS reported quality measures

The appendices to this guideline demonstrate model formats for structured reporting based on the principles and data elements contained in this document. Appendices 2 and 3 are model formats for exercise stress myocardial perfusion imaging, with Appendix 3 specifically demonstrating a combined conclusion. Appendices 4 and 5 are model formats for pharmacologic-based stress myocardial perfusion imaging. They are intended as examples only and ASNC fully acknowledges that there are many allowable structured formats for the reporting of nuclear myocardial perfusion images. Different structured report formats would be required for the other indications covered in this document (e.g., PET, exercise/rest FPRNA/ERNA, and viability imaging). Appendix 6 provides a diagram of the 17-segment model with corresponding vascular territories.17

Appendices

Appendix 1: Acceptable Units of Measure

Variable measured	Acceptable units of measure	Table location
Weight	Lbs; kg	2
Height	Inches; cm	2
Chest circumference	Inches; cm	2
HDL cholesterol	mg/dL; mmol/L	3
LDL cholesterol	mg/dL; mmol/L	3
Total cholesterol	mg/dL; mmol/L	3
Pharmaceutical stress dose	mg, mg/kg or µg·kg−1·min−1	4
Rest dose	mCi; MBq	7; 10
Stress dose	mCi; MBq	7
Reinjection dose	mCi; MBq	8
Blood Glucose level	mg/dL; mmol/L	8; 9

1. mg/dL, milligrams per deciliter; mmol, millimoles per liter; mCi, millicuries; MBq, megabecquerels

Note: Below are sample formats; please note, however, these do not include every variable.

Appendix 2: Sample Template for Exercise Myocardial Perfusion Imaging

(Single/2 day) Rest/Stress (or Stress/Rest) Exercise Stress Myocardial Perfusion Imaging with LV function analysis

Indication

(select one) (Diagnosis of coronary artery disease/known coronary artery disease/chest pain/shortness of breath/Preoperative assessment/Evaluation of myocardial viability/Risk Stratification/Other)

Clinical history:

X-year-old man/women with a history of:

Cardiac History:

Cardiac Risk Factors:

Prior cardiac imaging and procedures:

Prior nuclear stress test date:

Current symptoms:

Technique

At rest, the patient received x mCi of x tracer. X minutes later, resting tomographic images of the heart were obtained.

The patient then underwent exercise treadmill/bike stress testing according to the x protocol, exercising for x minutes, achieving a workload of x metabolic equivalents (METS). Resting HR was x with a peak heart rate of x bpm and x% maximum predicted heart rate and pressure rate product of x. Resting BP was x mm Hg and Peak BP was x mm Hg, which is a normal/hypertensive/hypotensive response. The heart rate response to recovery was normal/abnormal. The test was terminated due to chest pain/shortness of breath/fatigue/leg pain. Other symptoms included x.

The resting EKG showed x with no significant ST/T abnormalities that would preclude interpretation. The stress EKG showed (no) ST-segment changes consistent with myocardial ischemia, with x mm horizontal/upsloping/downsloping ST depression in the x leads. ST depressions began at x min of rest/stress and resolved at x min of rest/stress. The Duke Treadmill score was x, predicting a low/intermediate/high risk.

At peak stress, the patient received x mCi of x. Stress tomographic imaging was performed x minutes later. The rest and post-stress images were acquired with ECG gating, for assessment of left ventricular systolic function. All imaging was performed on a x camera and data were analyzed using x software.

Findings

The overall quality of the study is poor/fair/good/excellent. Review of the raw imaging demonstrates (no) significant motion during stress/rest image acquisition. Attenuation artifact was present/absent in the x walls.

Review of the perfusion images shows symmetric or improved uptake of tracer in all portions of the left ventricle from rest to stress imaging OR shows an x severity x sized perfusion defect in the anterior wall that is x reversible, a x sized x severity perfusion defect in the lateral wall that is x reversible and a x sized x severity perfusion defect in the inferior wall that is x reversible. Quantitative evaluation shows a summed stress score of x, a summed rest score of x, and a summed difference score of x. This represents a myocardial ischemic fraction of x%.

Gated SPECT images shows that the left ventricle is normal/enlarged in size and shows normal systolic performance. The LVEF at rest is x% and x% on post-stress images. No regional wall motion abnormalities are present during either stress or rest imaging.

Transient ischemic dilation, a high-risk marker, is/is not present. Left ventricular/right ventricular hypertrophy is/is not present. Left ventricular/right ventricular dilation is/is not present.

Impression

• 1. Myocardial perfusion imaging is normal with no evidence of ischemia or scar OR Myocardial perfusion imaging is abnormal with a small/moderate/large area of ischemia/infarction in the distribution of the x artery.

• 2. Left ventricular systolic function is normal/abnormal with (no)/x regional wall motion abnormalities. Left/Right ventricular hypertrophy/dilation is present.

• 3. In comparison with the previous study of x date, there has been (no)/a change in left ventricular perfusion, size, or function.

Appendix 3: Sample Template Exercise Myocardial Perfusion Imaging with Combined Conclusion

Reason for Study: Preoperative evaluation prior to non-cardiac surgery.

Clinical History: Mr. [XXXXXX] is a 56-year-old male with a history of hypertension and dyslipidemia with no prior known coronary artery disease who is currently asymptomatic. He has not had prior coronary angiography and has a SPECT myocardial perfusion imaging study from [xx/xx/xxxx] for comparison.

Stress ECG: (not provided in this appendix for brevity).

Isotope Administration

This was a gated SPECT myocardial perfusion imaging study. A one-day rest-stress imaging protocol was followed. The isotope used for imaging was ^99mTc-sestamibi. Rest imaging was performed after an injection of 7.1 mCi. Stress imaging was performed after an injection of 21.3 mCi.

Nuclear Stress Findings

Nuclear Study Quality

Overall imaging quality was good.

Perfusion Conclusion

LV perfusion is probably normal.

Perfusion Defect #1

There is a small region with moderate reduction in uptake in the apical to mid anterior segment(s) that is predominately reversible. There is normal wall motion in the defect area. The defect appears to be shifting breast artifact, but ischemia cannot be ruled out. The perfusion defect is visually present but not quantitatively significant.

Perfusion Comments

There is no evidence of transient ischemia dilation (TID). The rest study indicates well-preserved viability.

Function Comments

Left ventricular function post-stress was normal with an ejection fraction of 63%. The stress end-diastolic cavity size was normal (52 mL/m²). The stress end-systolic cavity size was normal (19 mL/m²).

Interpretation Summary

• The stress electrocardiogram was positive for electrocardiographic evidence of myocardial ischemia.

• The Duke Treadmill Score was intermediate risk at -5.

• The patient developed typical angina at peak stress.

• LV perfusion is probably normal.

• The small region with moderate reduction in uptake in the apical to mid anterior segment(s) appears to be shifting breast artifact but ischemia cannot be ruled out.

• Left ventricular function post-stress was normal with an ejection fraction of 63%.

Nuclear and Stress Combined Conclusion

The ECG and SPECT portions of the stress study are discordant, but the following factors support an intermediate risk of inducible myocardial ischemia:

• Poor exercise workload achieved during stress.

• Anginal symptoms during stress.

• Multiple cardiovascular risk factors.

Further cardiac evaluation for ischemic heart disease could be considered, especially in the setting of progressive or typical angina.

Nuclear Prior Study

Compared with the prior study dated [xx/xx/xxxx], the perfusion defect is new. There has been no significant change in left ventricular function.

Appendix 4: Sample Template for Pharmacologic-Based Stress Myocardial Perfusion Imaging

(Single/2 day) Rest/Stress (or Stress/Rest) Pharmacologic Stress Myocardial Perfusion Imaging with LV function analysis

Indication

(select one) (Diagnosis of coronary artery disease/known coronary artery disease/chest pain/shortness of breath/Preoperative assessment/Evaluation of myocardial viability/Risk Stratification/Other)

Clinical history

X-year-old man/women with a history of:

Cardiac History:

Cardiac Risk Factors:

Prior cardiac imaging and procedures:

Current symptoms:

Technique

At rest, the patient received x mCi of x tracer. X minutes later, resting tomographic images of the heart were obtained.

Pharmacologic stress testing was performed with adenosine/dipyridamole/dobutamine/regadenoson at a rate of ____ for ___minutes. Additionally, low-level exercise was performed along with the vasodilator infusion (specify: ____). Resting HR was x with a peak heart rate of x bpm and x% maximum predicted heart rate . The rest blood pressure was ___ mm/Hg and increased/decreased to ___ mm/Hg, which is a normal/hypotensive/hypertensive response. The patient developed significant symptoms, which included ____.

The resting EKG showed x with no significant ST/T abnormalities that would preclude interpretation. The stress EKG showed (no) ST-segment changes consistent with myocardial ischemia, with x mm horizontal/upsloping/downsloping ST depression in the x leads. ST depressions began at x min of rest/stress and resolved at x min of rest/stress.

At peak stress, the patient received x mCi of x. Stress tomographic imaging was performed x minutes later. The rest and post-stress images were acquired with ECG gating, for assessment of left ventricular systolic function. All imaging was performed on a x camera and data were analyzed using x software.

Findings

The overall quality of the study is poor/fair/good/excellent. Review of the raw imaging demonstrates (no) significant motion during stress/rest image acquisition. Attenuation artifact was present/absent in the x walls.

Review of the perfusion images shows symmetric or improved uptake of tracer in all portion of the left ventricle from rest to stress imaging OR show an x severity x sized perfusion defect in the anterior wall that is x reversible, a x sized x severity perfusion defect in the lateral wall that is x reversible, and a x sized x severity perfusion defect in the inferior wall that is x reversible. Quantitative evaluation shows a summed stress score of x, a summed rest score of x, and a summed difference score of x. This represents a myocardial ischemic fraction of x%.

Gated SPECT images shows that the left ventricle is normal/enlarged in size and shows normal systolic performance. The LVEF at rest is x% and x% on post-stress images. No regional wall motion abnormalities are present during either stress or rest imaging.

Transient ischemic dilation, a high-risk marker, is/is not present. Left ventricular/right ventricular hypertrophy is/is not present. Left ventricular/right ventricular dilation is/is not present.

Impression

• 1. Myocardial perfusion imaging is normal with no evidence of ischemia or scar OR Myocardial perfusion imaging is abnormal with a small/moderate/large area of ischemia/infarction in the distribution of the x artery.

• 2. Left ventricular systolic function is normal/abnormal with (no)/x regional wall motion abnormalities. Left/Right ventricular hypertrophy/dilation is present.

• 3. In comparison with the previous study of x date, there has been (no)/a change in left ventricular perfusion, size, or function.

Appendix 5: Sample Template for Pharmacologic-Based Stress Myocardial Perfusion Imaging

Appendix 6: Left Ventricular Segmentation17

Adapted and reprinted with permission from the American Society of Nuclear Cardiology; originally presented in Cerqueira MD, et al. J Nucl Cardiol 2002;9:240-5.