Physician Performance Measurement Set
American Academy of Neurology Board of Directors
Approved August 10, 2009
AMA Convened Physician Consortium for Performance Improvement
Approved March 9, 2010
Physician Performance Measures (measures) and related data specifications developed by the American Academy of Neurology (AAN) are intended to facilitate quality improvement activities by physicians.
These measures are intended to assist physicians in enhancing quality of care. Measures are designed for use by any physician who manages the care of a patient for a specific condition or for prevention. These measures are not clinical guidelines and do not establish a standard of medical care, and have not been tested for all potential applications.
Measures are subject to review and may be revised or rescinded at any time by the AAN. The measures may not be altered without prior written approval from the AAN. The measures, while copyrighted, can be reproduced and distributed, without modification, for noncommercial purposes (e.g. use by health care providers in connection with their practices). Commercial use is defined as the sale, license, or distribution of the measures for commercial gain, or incorporation of the measures into a product or service that is sold, licensed, or distributed for commercial gain. Commercial uses of the measures require a license agreement between the user and the AAN. Neither the AAN nor its members shall be responsible for any use of the measures.
THESE MEASURES AND SPECIFICATIONS ARE PROVIDED “AS IS” WITHOUT WARRANTY OF ANY KIND.
©2009 American Academy of Neurology. All rights reserved.
Limited proprietary coding is contained in the measure specifications for convenience. Users of the proprietary coding sets should obtain all necessary licenses from the owners of these code sets. The AAN and its members disclaim all liability for use or accuracy of any Current Procedural Terminology (CPT®) or other coding contained in the specifications.
CPT ® is a registered trademark of the American Medical Association.
American Academy of Neurology
Physician Performance Measurement Set
Nathan Fountain, MD
Paul Van Ness, MD
American Academy of Neurology Facilitator
Christopher Bever Jr., MD
American Academy of Neurology
Jeffrey Buchhalter, MD
Andres Kanner, MD
K. Babu Krishnamurthy, MD
Susan Naselli, MD
Piotr Olejniczak, MD
Rita Richardson, MD
Joseph Sirven, MD
Michael Sperling, MD
John Stern, MD
American Epilepsy Society
Allan Krumholz, MD
Paul Levisohn, MD
Epilepsy Foundation of America
Gregory L. Barkley, MD, FAAN
Michael C. Smith, MD
National Association of Epilepsy Centers
David Labiner, MD
Thaddeus Walczak, MD
American Academy of Family Physicians
Mark Potter, MD
American Academy of Pediatrics
Dennis Dlugos, MD
American Academy of Neurological Surgeons/Congress of Neurological Surgeons
Joshua Rosenow, MD
American Clinical Neurophysiology Society
William Tatum IV, DO
American College of Emergency Physicians
Andrew Jagoda, MD
American College of Radiology
Eric Russell, MD
American Psychological Association
Bruce Hermann, PhD
American Society of Neuroimaging
Ruben Kuzniecky, MD
Child Neurology Society
Kevin Chapman, MD
National Academy of Neuropsychology
Gregory Lee, PhD
National Organization of Rare Disorders
Suki Bagal, MD, MPH
Society of Nuclear Medicine
James M. Mountz, MD, PhD
American Academy of Neurology Staff
Rebecca Swain-Eng, MS
Sarah Tonn, MPH
American Medical Association Convened-Physician Consortium for Performance Improvement
Mark Antman, DDS, MBA
Timothy Kresowik, MD
Kay Schwebke, MD, UnitedHealth Care
Wesley Wong, MD, Wellpoint
Thomas James, MD, Humana
Robert Kropp, MD, Aetna
Purpose of Measures
These clinical performance measures which the American Academy of Neurology (AAN) developed using the model for performance measure development from the Physician Consortium for Performance Improvement (PCPI), are designed for use in individual quality improvement. The measures may also be used in data registries, continuing medical education (CME) programs, and board certification programs. Unless otherwise indicated, the measures are also appropriate for accountability if the necessary methodological, statistical, and implementation rules are met.
The measure titles listed below may be used for accountability:
Measure # 1:
Seizure Type(s) and Current Seizure Frequency(ies)
Measure # 2:
Documentation of Etiology of Epilepsy or Epilepsy Syndrome
Measure # 3:
Electroencephalogram (EEG) Results Reviewed, Requested, or Test Ordered
Measure # 4:
Magnetic Resonance Imaging/Computed Tomography Scan (MRI/CT Scan) Results Reviewed, Requested, or Scan Ordered
Measure # 5:
Querying and Counseling about Anti-Epileptic Drug (AED) Side Effects
Measure # 6:
Surgical Therapy Referral Consideration for Intractable Epilepsy
Measure # 7:
Counseling About Epilepsy Specific Safety Issues
Measure # 8:
Counseling for Women of Childbearing Potential with Epilepsy
Intended Audience, Care Setting, and Patient Population
These measures are designed for use by physicians and other eligible health professionals who provide care to individuals diagnosed with epilepsy. The measures may be used in the emergency department only if the physician or eligible provider uses the appropriate ICD-9 and CPT codes as described under each individual measure. The measures are intended to be used to calculate performance and/or to report measurement at the individual physician level.
The AAN seeks to specify measures for implementation using multiple data sources, including paper medical records, administrative (claims) data, and in particular, Electronic Health Record Systems (EHRS). Specifications to report on the measures for Epilepsy using administrative (claims) data are included in this document. The AAN has identified codes for these measures, including ICD-9 and CPT (Evaluation and Management Codes, Category I and, where applicable, Category II codes). Specifications for additional data sources, including EHRS, will be fully developed at a later date.
The AAN used the PCPI policy “Specification and categorization of measure exclusions: recommendations to PCPI work groups” as the basis for defining exclusions. (Available at: http://www.ama assn.org/ama1/pub/upload/mm/370/exclusions053008.pdf. Accessed September 2008)
This methodology is described below.
For process measures the PCPI provides three categories of reasons for which a patient may be excluded from the denominator of an individual measure:
Not indicated (absence of organ/limb, already received/performed, other)
Contraindicated (patient allergy history, potential adverse drug interaction, other)
Resources to perform the services not available
Insurance Coverage/Payer-related limitations
Other reasons attributable to health care delivery system
These measure exclusion categories are not available uniformly across all measures; for each measure, there must be a clear rationale to permit an exclusion for a medical, patient, or system reason. The exclusion of a patient may be reported by appending the appropriate modifier to the CPT Category II code designated for the measure:
Medical reasons: modifier 1P
Patient reasons: modifier 2P
System reasons: modifier 3P
Although this methodology does not require the external reporting of more detailed exclusion data, the PCPI recommends that physicians document the specific reasons for exclusion in patients’ medical records, for purposes of optimal patient management and audit-readiness. The PCPI also advocates for the systematic review and analysis of each physician’s exclusions data to identify practice patterns and opportunities for quality improvement. For example, it is possible for implementers to calculate the percentage of patients whom physicians have identified as meeting the criteria for exclusion.
Please refer to the documentation for each individual measure for information on acceptable exclusion categories and the codes and modifiers to be used for reporting.
Data Capture and Measure Calculation
The intent of this measurement set is to encourage physicians to collect data on each patient eligible for a measure. Physicians should receive feedback on measures both at the patient level to facilitate patient management and in the aggregate to identify opportunities for improvement across a physician’s patient population.
Measure calculations will differ depending on whether a rate is being calculated for performance or reporting purposes.
The method of calculation for performance follows three steps. First, identify the patients who meet the eligibility criteria for the denominator (PD); second, identify which of those patients meet the numerator criteria (A); and third, for those patients who do not meet the numerator criteria, determine whether an appropriate exclusion applies and then subtract those patients from the denominator (C) (see examples below).
The methodology also enables implementers to calculate the rates of exclusions and to analyze further both low rates and high rates, as appropriate (see examples below).
The method of calculation for reporting differs. One program that currently focuses on reporting rates is the Centers for Medicare and Medicaid Services (CMS) Physician Quality Reporting Initiative (PQRI). Under that program’s current design, there is a reporting denominator determined solely from claims data (CPT and ICD-9), which in some cases results in a reporting denominator that is much larger than the eligible population for the performance denominator. Additional components of the reporting denominator are explained below.
The components that make up the numerator for reporting include all patients from the eligible population for which the physician has reported, including the number of patients who meet the numerator criteria (A), the number of patients for whom valid exclusions apply (C), and the number of patients who do not meet the numerator criteria (D). These components, where applicable, are summed to make up the inclusive reporting numerator. The calculation for reporting will be the reporting numerator divided by the reporting denominator (see examples below).
Examples of calculations for reporting and performance are provided for each measure.
Calculation for Performance
For performance purposes, this measure is calculated by creating a fraction with the following components: Numerator, Denominator, and Denominator Exclusions.
Performance Denominator (PD) Includes: Number of patients meeting criteria for denominator inclusion
Denominator Exclusion (C) Includes: Number of patients with valid medical, patient, or system exclusions (where applicable; will differ by measure)
A (# of patients meeting numerator criteria
PD (# of patients in denominator) - C (# of patients with valid denominator exclusions)
It is also possible to calculate the percentage of patients either excluded overall or excluded by medical, patient, or system reason where applicable:
Overall Exclusion Calculation
C (# of patients with any valid exclusion)
PD (# of patients in denominator)
Exclusion Calculation by Type
C1 (# patients with
PD (# patients in denominator)
C2 (# patients with
PD (# patients in denominator)
C3 (# patients with
PD (# patients in denominator)
Calculation for Reporting
For reporting purposes, this measure is calculated by creating a fraction with two components: Reporting Numerator and Reporting Denominator.
Reporting Numerator includes each of the following components, where applicable. (There may be instances where there are no patients to include in A, C, D, or E.)
A. Number of patients meeting additional denominator criteria (for measures where true denominator cannot be determined through ICD-9 and CPT Category I coding alone) AND numerator criteria
C. Number of patients with valid medical, patient or system exclusions (where applicable; will differ by measure)
D. Number of patients not meeting numerator criteria and without a valid exclusion
E. All other patients not meeting additional denominator criteria (for measures where true denominator cannot be determined through ICD-9 and CPT Category I coding alone)
Reporting Denominator (RD) Includes:
RD. Denominator criteria (identifiable through ICD-9 and CPT Category I coding)
A (# of patients meeting additional denominator criteria AND numerator criteria) + C (# of patients with valid exclusions) + D (# of patients NOT meeting numerator criteria) + E (# of patients not meeting additional denominator criteria)
RD (# of patients in denominator)
Burden of Illness (Mortality and Morbidity)
Epilepsy is the third most common neurological disorder in the United States. (Alzheimer's disease and stroke are the two most common.1) It is equal in prevalence to cerebral palsy, multiple sclerosis, and Parkinson's disease combined1. Approximately 2.5 million Americans are affected by epilepsy2.
Epilepsy is not a single entity but a family of more than 40 syndromes3. Although people of all ages can be affected by epilepsy, the very young (under 2) and the elderly are the most affected2,4. In the United States, epilepsy currently affects more than 326,000 children under age 15 and more than 90,000 of them have severe seizures that cannot be adequately treated1.
There is an extremely high burden of illness associated with epilepsy. The mortality rate among people with epilepsy is two to three times higher than that of the general population, and the risk of sudden death is 24 times greater1. Every year 200,000 people in the United States will be diagnosed with epilepsy4 and an estimated 25,000 to 50,000 will die of seizures and related cause3. Thirty to 40 percent of people with epilepsy are severely affected and continue to have seizures even if given treatment1. Optimal seizure control can reduce the risk of epilepsy-related mortality, decrease morbidity, and greatly improve the quality of life.
Prevalence of active epilepsy (history of the disorder plus a seizure or use of antiepileptic medicine within the past 5 years) is estimated at approximately 2.5 million in the United States2. The prevalence rate is higher among racial and ethnic minorities than among Caucasians4.
The cumulative incidence (risk of developing epilepsy) within the United States increases with age. One percent of the population can be expected to have developed epilepsy by 20 years of age4. By 75-80 years of age, 3 percent of the population will likely have been diagnosed with epilepsy, and 10 percent will likely have experienced some type of seizure4.
Burden of Illness (Cost)
The annual economic burden due to epilepsy is estimated to be about $15.5 billion1,2 in the United States from associated health care costs and losses or reduction in employment, wages, and productivity.
Potential of Epilepsy Performance Measurement Set to Improve Health Outcomes
Epilepsy is a widely recognized neurologic condition, but is it often poorly understood, diagnosed, and treated. The lack of specialty care may lead to a delayed diagnosis and inadequate treatment. The health-related quality of life can be measured in days of activity limitation, pain, depression, anxiety, cognitive deficits, reduced vitality, and insufficient sleep or rest. The deficits in quality of life due to epilepsy and its treatment are comparable to conditions such as arthritis, heart problems, diabetes, and cancer1. A performance measurement set for epilepsy has the potential to increase patient safety, improve diagnosis, improve treatment, reduce the number of deaths due to epilepsy, and increase the quality of life for those who have epilepsy.
Variability in Clinical Practice:
Diagnosing epilepsy is a multi-step process that can involve multiple different tests and many different specialties. Epilepsy treatment is provided by multiple different specialties. These specialties include neurology, internal medicine, pediatrics, obstetrics and gynecology, psychiatry, neurosurgery, and family practice.
Medical professionals vary in their skill and approach to diagnosing epilepsy, determining seizure type, identifying causation, and administering appropriate therapy. A uniform performance measurement set is needed to clarify these roles and to determine how best to establish evidence-based standards of care.
Numerous recommendations for performance measures exist that could easily be applied by neurologists, epileptologists, and other health professionals.
In addition, there are multiple sources of nationally and internationally accredited guidelines available. The major guideline-producing entities are the American Academy of Neurology (AAN), Scottish Intercollegiate Guideline Network (SIGN), Centers for Disease Control (CDC), National Institute for Health and Clinical Excellence (NICE), and the Singapore Ministry of Health. Other, smaller relevant data sources for epilepsy guidelines are available as well.
The performance measures found in this document have been developed using these guidelines, enabling the physician to track his or her performance in individual patient care across patient populations. Please note that the provision of epilepsy care must be based on individual patients’ needs and the clinician’s professional judgment. Performance measures are not to be used as a substitute for clinical guidelines or individual physician clinical judgment. There may be instances where the age of an individual patient lies beyond the age range identified for the performance measure(s); however, this does not preclude the patient from receiving the service. Whether or not a patient should receive specific care is a decision that needs to be made between the patient and the physician while weighing the risks and benefits of the service with individual patient preference.
Measure #1: Seizure Type(s) and Current Seizure Frequency(ies)
This measure may be used as an Accountability measure
Clinical Performance Measure
Numerator: Patient visits with seizure type(s) specified and current seizure frequency for each seizure type documented in the medical record.
Denominator: All visits for patients with a diagnosis of epilepsy.
Documentation of medical reason for not documenting seizure type(s) and current seizure frequency for each seizure type (e.g. patient is unable to communicate and an informant is not available).
Documentation of patient reason for not documenting seizure type(s) and current seizure frequency for each seizure type (e.g. patient and/or informant refuses to answer or comply )
Measure: All visits for patients with a diagnosis of epilepsy who had the type(s) of seizure(s) and current seizure frequency for each seizure type documented in the medical record.
The following clinical recommendation statements are quoted verbatim from the referenced clinical guidelines and represent the evidence base for the measure:
Detailed history of the attack should be obtained from the person who had the attack + symptoms and from eyewitness(es) to the attack. (Level B) NICE (Oct. 2004)22
The seizure type(s) and epilepsy syndrome should be identified. (Level C) SIGN (April 2003) 23
When a patient with epilepsy receives follow-up care, then an estimate of the number of seizures since the last visit and assessment of drug side-effects should be documented. (Level D 1+/ Primary) Pugh (2007)17
IF a patient is thought to have a diagnosis of epilepsy THEN the diagnosis should include a best estimation of seizure types. (Level C 2+/Secondary) Pugh (2007)17
Rationale for the Measure:
Seizures are divided into generalized and partial (or focal) types based on whether they begin throughout the brain simultaneously or in one focal region (Dreifuss et al 1981). The main objective in treating epilepsy is to reduce the frequency of seizures and eventually achieve seizure freedom without medication side effects. In order to know that a treatment is effective, the patient’s seizure frequency must be known before an intervention is begun so it can be compared to the seizure frequency determined during follow-up visits after an intervention is instituted. Antiepileptic drugs reduce the frequency of seizures in controlled clinical trials. Seizure freedom is associated with improvement in health-related quality of life, for example after epilepsy surgery. Therefore, accurate assessment of seizure frequency is necessary to provide most forms of care for epilepsy.
Dreifuss FE, Bancaud J, Henriksen O, et al. Proposal for the revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia 1981;22:489-501.
Data Capture and Calculations:
Calculation for Performance
For performance purposes, this measure is calculated by creating a fraction with the following components: Numerators, Denominator, and Denominator Exclusions.
Performance Numerator (A) Includes:
Patient visits with seizure type(s) specified and current seizure frequency for each seizure type documented in the medical record.
Performance Denominator (PD) Includes:
All visits for patients with a diagnosis of epilepsy.
Denominator Exclusions (C) Include:
Documentation of medical reason for not documenting seizure type(s) and current seizure frequency for each seizure type in the medical record.
A (# of patient visits meeting measure criteria)
PD (# of patients visits in denominator) – C (# of patient visits with valid denominator exclusions)
Components for this measure are defined as:
# of patient visits with seizure type(s) specified and current frequency for each seizure type documented in the medical record
# of patient visits for patients with a diagnosis of epilepsy
# of patient visits for patients with valid medical reason(s) for not documenting seizure type(s) and current seizure frequency for each seizure type