Employers, insurers, and other purchasers of health care services collect data to profile the practice habits of pediatricians and other physicians. This policy statement delineates a series of recommendations that should be adopted by health care purchasers to guide the development and implementation of physician profiling systems.
As a result of rapidly rising health care costs, employers, insurers, governmental entities, and other purchasers of health care services have continued to seek effective ways of measuring the quality, utilization, and cost of services provided to patients. One approach commonly used to analyze patterns of care is profiling. Profiling is defined as a measurement of the quality, utilization, and cost of medical resources provided by physicians that is made by employers, third-party payors, governmental entities, and other purchasers of health care.1 Clinical profiling measures a physician's style of practice using treatment modality, utilization of services, and outcomes of care. Economic profiling examines the financial dimensions of a physician's practice style, case mix, and demographic factors.2 Profiling has been identified as having three primary “appropriate” applications: quality improvement, utilization review, and assessment of provider performance.3 Use of profiling to restrict provider participation based solely on economic considerations is an unfortunate use of these data in many markets, especially when third-party payers control a large share.
The Health Plan Employer Data and Information Set (HEDIS 3.0)4 is an example of data that are considered valuable to employers, reasonable for managed care organizations (MCOs) and other health plans to gather, and useful for improving patient care. The HEDIS set of performance measures includes a mix of outcomes and provides measures for acute and preventive care. However, it still has limited application to children with chronic conditions.
In the development and implementation of profiling measures, the American Academy of Pediatrics (AAP) recommends the following guidelines:
Pediatric input into and review of profile plans should be required before their use. The active involvement of local and/or national pediatric physician organizations is critical in the development of a profiling system.
Pediatricians have the responsibility to know and understand the profiling methodology and data requirements applied by those collecting and/or using the information.
Pediatricians should be notified in writing, with adequate advance notice as previously agreed to by all contracting parties, if a standard profiling methodology or data collection requirement of a profiling entity are going to be changed.5
A health plan, MCO, or other health care purchaser's request for the use of pediatricians' offices, staff and physician time, and the audit requirements of medical records must be reasonable, compensated, and should not compromise the care of patients. They should reimburse the practice for the time and materials for these audits and, where appropriate, pediatricians should be provided with the support and resources to analyze and work with these data. However, currently many MCOs and health care purchasers do not reimburse clinicians for this work.
Pediatric physician profiles should be based on the collection of valid, accurate, and objective data using a nationally recognized standardized approach and should reflect recognition of the nature of a pediatrician's practice, the quality of services rendered, and acuity of care (specifically age/sex-adjusted case mix severity), required by his/her panel of patients.6Also, profiles should account for the number or percentage of children with special needs that are cared for by the pediatrician because these children require a more comprehensive array of acute and chronic care services.
Data used as part of a quality improvement program for educational purposes or feedback on review of medical record documentation should be presented to the pediatrician in a user-friendly manner and by a physician.2
The methods for collecting and analyzing the profile data must be fully disclosed to both the pediatrician and the consumer. The methodology for determining the profiles must be explained to both providers and consumers in easily understandable language, because complex statistical analysis is the methodology often used.7
Norms should be based on valid data collection and profiling methodologies, and must use a sample size that is of sufficient statistical power. Interpreting results that are based on insufficient sample size may lead to erroneous conclusions and inappropriate actions. Data on norms should be accompanied by presentation of the limits at one and two standard deviations from the norm to delineate the range and not just the midpoint of the dataset.
Standards, guidelines, or practice parameters used for pediatric profiling should be derived from either evidence-based, generally accepted, local pediatric practices or from the published standards of the AAP.
Data sources used to develop profiles of pediatricians have many limitations. This is especially true of surveys, medical records, and claims data because of their limited ability to assess patients' health status and wellness. These limitations must be clearly identified and acknowledged by the health care payer and other reviewers to itself, its patients, and its enrollees.3
Comparisons among physicians' profiles should be adjusted for the following:
conformity to national quality of care guidelines, specifically the practice guidelines developed by the AAP and those endorsed by the AAP;
medical specialty (whether pediatrician, family physician or general practitioner; whether primary care pediatrician, pediatric medical subspecialist, or pediatric surgical specialist9);
patient care mix (although risk adjustments have not been widely tested for their validity and reliability in the pediatric population. If a system is used it should be widely accepted among the group of pediatricians being compared along with the number or percentage of children with special health care needs who are cared for by the pediatrician10);
geographic considerations (rural vs city, suburban vs inner-city);
distinctions between the ordering and referring physician and the physician providing the service or procedure2; configuration of practice type (solo, small vs large, single specialty vs multispecialty).
This latter point is of significant importance because in some cases individual patients are assigned to a specific physician within a practice. In a group setting the pediatrician who is assigned by the MCO may not be the one in the group whom the patient arranges to see regularly.
12. The data sources and methodologies, as well as the quality and accuracy of physician profiling, should be evaluated on a regular basis.1
13. The MCOs, health plans, and other health care purchasers using profiling data should institute effective controls and procedures to protect against the unauthorized use of pediatrician profiles. For example, the application of profiling data used for quality improvement through peer review can not be included during litigation or admissible for purposes of discovery whereas that used for economic purposes may.6 Additionally, profiling data should not be used to exclude specific pediatricians from panels solely on an economic basis, when payment adjustments could be made to accommodate the practice style of that pediatrician.
14. The quality and accuracy of pediatric-specific medical information should only be evaluated by the use of periodic on-site medical record audits and should be conducted by pediatricians or other clinicians who have both experience in providing and knowledge of pediatric care.
The American Academy of Pediatrics acknowledges the importance of pediatric profiling. If applied properly, profiling has the potential to improve quality of care, physician performance, and patient outcomes. It is important that the information collected regarding physicians and/or physician groups is accurate, interpreted appropriately, and only released to the public under controlled situations. Pediatric profiling can be a successful venture when there is ongoing communication and cooperation between pediatricians and evaluators.
Jack T. Swanson, MD, Chairperson
Edward O. Cox, MD
F. Lane France, MD
Katherine C. Teets Grimm, MD
James W. Herbert, MD
E. Susan Hodgson, MD
Allan S. Lieberthal, MD
Kyle Yasuda, MD
Committee on Practice and Ambulatory Medicine, 1998–1999
Todd Davis, MD
Robert D. Chessin, MD
Robert Sayers, MD
Emmanuel E. Eugenio, MD
Jan Ellen Berger, MD, Chairperson
Jeffrey Berman, MD
Jerome M. Buckley, MD
William Hugh Craft, MD
Charles H. Deitschel, Jr, MD
Geoffrey Evans, MD
John J. Fraser, Jr, MD, JD
Committee on Medical Liability, 1998–1999
Holly Myers, JD
Landon Combs, MD
Ken Heland, JD
American College of Obstetricians and Gynecologists Liaison
The recommendations in this statement do not indicate an exclusive course of treatment or serve as a standard of medical care. Variations, taking into account individual circumstances, may be appropriate.
- HEDIS =
- Health Plan Employer Data and Information Set •
- MCO =
- managed care organization •
- AAP =
- American Academy of Pediatrics
- ↵Schlackman N. Physician profiling—the road to where? Strategies for managed care: an update from the Committee on Child Health Financing. AAP News. 1995;10. Insert
- ↵American Medical Association, Department of Medical Review. Physician Profiling and the Release of Physician-Specific Health Care Data. Chicago, IL: American Medical Association; 1995
- ↵National Committee for Quality Assurance. HEDIS 3.0. Washington, DC: National Committee on Quality Assurance; 1998
- Keller RB
- ↵Horn SD. Physician profiling: how it can be misleading and what to do. Consultant. 1987;27:86–88, 93–94
- American Academy of Family Physicians. Policy Statement: Requirements Related to Managed Care Plans, Utilization Review Programs and Point of Service Plans. Kansas City, MO: American Academy of Family Physicians; 1996
- Politser P
- Copyright © 1999 American Academy of Pediatrics