Expansion of Quality of Care Measures (Q-SPAN) is a project of the Agency for Health Care Policy and Research designed to strengthen the science base of quality measurement while expanding the scope and availability of validated, ready-to-use measures. Q-SPAN builds on past work in quality measurement by public and private organizations through eight cooperative agreements to develop and test additional clinical performance measures for specific conditions, patient populations, and health care settings.
Once the new measures are ready for use, they will be added to the more than 1,200 performance measures included in AHCPR's Computerized Needs-Oriented Quality Measurement Evaluation
System (CONQUEST), a unique software tool for helping health plans, providers, and others identify,
understand, evaluate, and select clinical measures to assess and improve the quality of the health care they provide.
Overview
An overview of the eight cooperative agreements is provided with more detailed project profiles following:
Performance measures are being developed and tested for two major dental diseases: dental caries and periodontal
disease. They will be evaluated in various dental care sites in terms of outcomes, feasibility of implementation, and
perceived utility among purchasers, administrators, and providers.
Asthma is a common condition for which good primary care can reduce expensive emergency department visits and
hospital admissions. Measures of quality of care are being developed and tested on patients requiring emergency or
hospital care for acute asthma exacerbations, and modeling techniques will be used to identify the key predictors of the asthma outcomes measured.
Process measures are being developed for care delivered to adult men and women over age 50. A complimentary
project with the Health Care Financing Administration will address children and adult women. These measures will be
combined into a global measure of the performance of health plans.
This project focuses on a group of interrelated cardiovascular conditions (acute myocardial infarction, congestive heart
failure, and hypertension). Performance measures for episodes of care across the spectrum of care settings
are being developed and evaluated in several health plans, with evaluation including the cost of data collection.
Hip fracture in the elderly is frequent, serious, and expensive. The functional status of 500 patients is being followed
for 6 months from occurrence through hospital and rehabilitative care. Interventions to improve specific care processes
are being evaluated against outcomes.
A cross-section of Alabama providers and plans is being compared on various measures of quality of care for diabetes
mellitus and breast cancer screening using a specially designed approach to benchmarking. Interventions are being evaluated to measure the
effect of benchmarks on practice change by providers.
The National Committee on Quality Assurance's Health Plan Employer Data and Information Set (HEDIS) is widely
used to measure performance of health care plans. This project is developing operational specifications for certain
HEDIS measures and validating them against appropriate outcomes.
Subacute care in nursing homes and home care programs have grown in importance as hospital lengths of stay have decreased. This
project is measuring quality of post-acute care in these two transitional settings. Two different approaches to modeling
change in health status (such as activities of daily living, mobility, cognition, and pain) are being compared.
Project Profiles
Brief Description: The purpose of this study is to develop and evaluate a set of clinical performance measures for
dental care. The measures will be developed with guidance from Steering and Advisory Committees comprising
representatives of dental health plans, purchasers, and providers. The measures will be pilot tested for reliability and
validity. They will then be implemented in a large demonstration project involving two large managed care plans. A
supplement was awarded in September 1997 to add up to six additional dental care sites, such as state and local public
health delivery programs and national insurers. The final product will be a users' manual including information on the
development of the measures, their reliability and validity, how their numerators and denominators are constructed, and
how they may be used.
Conditions considered: Dental caries and periodontal disease.
Populations: Children and adult dental patients at two large HMO plans in Portland, OR and Minneapolis, MN; up to
six additional sites are to be selected.
Measurement approach: Primarily outcome measures.
Measures to be Developed or Evaluated: Measures for three dimensions of dental care will be developed:
effectiveness of care, access to care, and use of services.
Consortium Members: University of North Carolina at Chapel Hill, Group Health Foundation, and Kaiser Permanente Center for Health Research.
Principal Investigator and Institution:
James Bader, D.D.S.
University of North Carolina at Chapel Hill
Project Dates: September 30, 1996 - September 29, 1998
Progress to Date: The primary set of measures on effectiveness of care have been developed, with three initial sets of
specifications based on three possible sources of data for their calculation (administrative system, record audit and
claims data). Pilot testing will begin shortly. The measures for "access" and "use of services" are currently under
development.
Brief Description: The project contains two effectiveness studies (study I and II) and the pilot study. The study
consists of two phases. Phase 1 involves selecting the process of care criteria for an acute asthma exacerbation using
the National Heart, Lung, and Blood Institute's asthma evaluation and management guidelines. The investigators will
use the asthma process of care criteria to conduct study I. Administrative and clinical records will be used to assess the
medical care facility and physician characteristics, and the process of care for the acute asthma exacerbation. Two to
three weeks after patients have received treatment for an acute episode, they will complete four short questionnaires and
undergo allergy skin and objective lung function testing to assess patient characteristics, patient environmental factors,
and asthma outcomes. The asthma outcomes are: (1) inhaler use technique; (2) objective lung function testing (pre- and
post-bronchodilator FEV1; (3) asthma quality of life; (4) satisfaction with health care related to a specific acute asthma
exacerbation, (5) medical facility utilization for acute asthma exacerbation, and (6) asthma mortality. Statistical
modeling will be used to identify the asthma process of care criteria and the medical care facility, physician, and patient
characteristics (including patient environmental factors) that predict the asthma outcomes measured. The process of
care criteria that predict the optimal asthma outcomes will be used to develop the asthma quality of care candidate
measures. Before the pilot study, the investigators will develop a data retrieval system to obtain the data elements for
the candidate measures. The pilot study will assess the reliability of these measures and preliminarily assess their
validity, and feasibility and cost for their use with the data retrieval system. Phase II includes study II, which will test
the candidate measures for (1) their sensitivity, specificity, and predictive value as it relates to the asthma risk-adjusted
outcomes measured, and (2) the feasibility and cost for using them via a data retrieval system developed in Phase 1.
The key quality of care measures related to an acute asthma exacerbation will be selected based on the results of their
testing performance in study II.
Conditions considered: Asthma.
Populations: Adults age 18 and over (includes English and Spanish-speaking Hispanic Americans).
Measurement approach: Process measures.
Measures to be Developed or Evaluated: The process of care measures to be tested in Phase 1 are:
- Inhaled short-acting B2-agonists for all patients.
- Inhaled anti-cholinergics for severe asthma exacerbations and patients with impending or actual respiratory arrest.
- Oxygen for patients with an oxygen saturation <90% .
- Systemic corticosteroids for patients recently on systemic corticosteroids (within past month) or with severe or
moderate asthma exacerbations.
- Inhaled (at least double the current dose) corticosteroids for patients who are currently taking inhaled steroids with
mild asthma exacerbations.
- Inhaled B2-agonists at discharge for all patients.
- Inhaled corticosteroids (at least double the current dose for 7-10 days) at discharge for patients who are currently
taking inhaled steroids with mild asthma exacerbations.
- Systemic corticosteroids at discharge for patients with moderate and severe asthma or patients with mild asthma
exacerbations who were treated with systemic steroids recently (within past month).
- Medical facility followup visit within three days for asthma care for all patients.
- Patient education on asthma management at discharge (involving the correct use of inhalers).
- Patient education on asthma management at discharge (involving a review of a written plan for asthma discharge
medications) for all patients.
Consortium Members: University of Texas (UT) Southwestern Medical Center at Dallas, Parkland Health and Hospital System
Principal Investigator and Institution:
Yvonne Coyle, M.D.
UT Southwestern Medical Center at Dallas
Project Dates: September 30, 1996 - September 29, 1999
Progress to Date: Progress to date includes: (1) selection of 11 asthma process of care criteria for the treatment of an
acute asthma exacerbation, and (2) development of the data collection protocols and the testing procedures for the
study.
Brief Description: To provide methods for assessing the quality of care delivered in managed health care plans, this
project will examine care delivered to men ages 18-50 and women age 50 and over. Similar work is already in progress for
children and women ages 18 to 50, funded by the Health Care Financing Administration (HCFA). The aim is to
develop a single, global quality measure or at least measures for a small number of domains. A global measure of quality of care has two major advantages over more specific measures: first, it
is more accessible to consumers and other users; second, it is less vulnerable to "gaming." Conditions will be
selected to provide broad representation of care provided to the subject groups. Criteria and weights are derived from
the scientific literature using expert panels, organized into modules (such as "Preventive Services" or "Diabetes").
Branching logic will be developed to determine which modules apply to a given patient. Required data elements will
be determined, and a computer-based data collection tool will be designed for use on laptop computers to guide
collection of data from administrative and medical records. A sampling strategy will be designed to allow
representation of about 20 cases per condition per participating plan plus 100 randomly sampled cases (a total of 500
cases per plan). A pilot test of the data collection instrument will be conducted on a random subsample of patients prior
to beginning actual data collection toward the end of year 2. Data collection will be aimed to determine the reliability,
sensitivity, and specificity of the measures and to evaluate their utility for distinguishing among plans.
Conditions considered: Twenty-seven conditions for new indicator development (e.g., anxiety, colorectal cancer,
prostate cancer, menopause); 12 to be adapted from the HCFA study (e.g., alcoholism, breast mass, diabetes, low back
pain).
Populations: Men ages 18-50 and women over age 50 enrolled in participating health care plans.
Measurement approach: Mostly process-based measures.
Measures to be Developed or Evaluated: Global quality measures based on process of care, including screening,
diagnosis, treatment, and followup for conditions and populations as described above.
Consortium Members: Aetna-U.S. Healthcare, Harvard Pilgrim Health Care, PacifiCare, and United HealthCare
Principal Investigators and Institution:
Elizabeth A. McGlynn, Ph.D.
Steven M. Asch, M.D., M.P.H.
Eve A. Kerr, M.D., M.P.H.
RAND, Santa Monica, CA
Project Dates: September 30, 1996 - September 29, 1999
Progress to Date: Indicators will have been selected by January 1998, and operationalization will proceed, including precise statements of variables, development of software to guide data collection, and development of the method for weighting measures for aggregation.
Brief Description: Measures are being distilled from existing practice guidelines and clinical trial data and ranked for
importance by four panels of stakeholders (physicians, consumers, health-plan administrators, and purchasers)
representing physician organizations, health care purchasers, Medicare and Medicaid programs, and State quality
improvement organizations). Resulting measures will be tested to determine the most accurate and least expensive data
sources, and to determine the usefulness of the measures as indicators of the quality of health care delivered by various
provider types. Depending on the frequency of observations and the cost of collecting the data, determinations will be
made about the feasibility of measurement of quality performance of health plans, branch offices of health plans, or
individual providers, using a given measure.
Conditions considered: Congestive heart failure, acute myocardial infarction, and hypertension.
Populations: Enrollees of cooperating health plans who meet the disease-specific criteria.
Measurement approach: Mostly process measures.
Measures to be Developed or Evaluated: A total of 95 measures now being tested; specific information expected to
be available in late 1998.
Consortium Members: Participating health plans are Allina in Minneapolis, MN; PacifiCare in Portland, OR; United
HealthCare in Atlanta, GA and St. Louis, MO; and Prudential in Baltimore, MD and Houston, TX. Both United and
Prudential have expanded the number of sites they are including in the study. Also participating is the Health Education
and Research Foundation (HERF), in Minneapolis, MN, as a partner with Allina Health System.
Principal Investigators and Institution:
Barbara J. McNeil, MD, Ph.D., Principal Investigator
R. Heather Palmer, MB, MCH, SM, Co-Principal Investigator
Harvard University
Project Dates: September 30, 1996 - September 29, 2001
Progress to Date: The project is expected to yield measures of quality of cardiovascular care that will become national
standards and will be easily modified to reflect changing clinical practice.
Brief Description: The study will determine the feasibility and validity of functional status as a measure of health care
quality for patients with hip fracture. The first phase of the project will collect information on baseline function, patient
characteristics, and treatment on patients admitted with hip fracture at selected hospitals. All patients will be followed
longitudinally with assessments of functional status at 2 and 6 months after admission. The study will collect detailed
information on specific processes of care that are used during the treatment course of the patients by concurrently
following the course of the patients and by reviewing medical records. Investigators will evaluate the relationship
among risk-adjusted functional outcomes and specific medical care processes. The effort and feasibility of the data
collection and the reliability and validity of the measures will be examined. In the second phase, the project will plan
and implement a system to collect functional status outcomes across the health care system and test an intervention at
selected hospitals to determine whether these outcomes may be improved by focused interventions directed at
improving those processes of care that most affect outcomes. The final product of the project will be an operational set
of questions and procedures for collecting meaningful functional status data on patients with hip fracture.
Conditions considered: Hip fracture.
Populations: Approximately 500 hip fracture patients treated in four selected hospitals in the Mount Sinai Health
System, a network of 26 hospitals and 12 nursing homes in the New York metropolitan area.
Measurement approach: Functional and clinical outcomes.
Measures to be Developed or Evaluated: Functional independence measure (FIM)-MOTOR, social activity measure, and pain measure.
Consortium Members: Mount Sinai Medical Center; Hospital for Joint Diseases/New York University Medical
Center; Maimonides Medical Center, Brooklyn, NY; Phelps Medical Center, Sleepy Hollow, NY; University of
Maryland at Baltimore; and State University of New York at Albany.
Principal Investigators and Institution:
Albert Siu, M.D.
Mount Sinai Medical Center
Project Dates: September 30, 1996 - September 29, 2001
Progress to Date: The literature review has been finalized and is being submitted as a manuscript for publication. The
methods for identifying and collecting information from patients, including informed consent and followup interviews,
in Phase 1 have been tested. As a result, some modifications have been made to the data collection forms. Based on
tracking patients at several sites, it was determined that appropriate sample size for the study could be obtained. Patient
enrollment started in August 1997.
Brief Description: Benchmarking is a quality improvement technique involving identification of the "best in class," for
emulation by others. Frequently, the performance measured is not well selected in terms of scientific evidence, and the
actual benchmark has not always been data driven. This project relies heavily on AHCPR's CONQUEST database of
measures and will select and refine data-driven benchmark measures as "Achievable Benchmarks of Care" (ABCs).
The project will evaluate the effectiveness of the use of ABCs on a variety of medical conditions, demonstrate their use in
profiling physician performance across the continuum of care, test them at different levels of aggregation and in
different settings, and compare measures derived from different data sources. A major objective is to compare the
effect on provider performance of ABCs versus more traditional feedback methods, and seek associations between
ABCs and the process and outcomes of care.
Conditions considered: Diabetes mellitus (DM), breast cancer screening, urinary tract infection, and asthma.
Populations: DM—96 generalist physicians in Alabama; 1,955 ambulatory diabetic patients; patients with primary or
secondary admission diagnoses of DM at 15 Alabama hospitals. Breast cancer—1,096 eligible patients of 74 Alabama
physicians involved in the Health Care Financing
Administration's Ambulatory Care Quality Improvement demonstration project. A random sample of
Alabama Medicare beneficiaries (n=69,777 from October 1993 to January 1997), with a controlled trial of ABC feedback
to primary physicians is being considered. A similar trial with a managed care population (VIVA HMO) is under
development.
Measurement approach: Mostly process-of-care indicators.
Measures to be Developed or Evaluated: Diabetes mellitus—Inpatient measures, especially the association between the process of inpatient care and early readmission, and the method
of aggregating and weighting explicit process of care criteria for overall quality assessment. Outpatient measures, for
ABC testing: creatinine rates, triglyceride rates; cholesterol rates; flu vaccine; foot exam; eye exam; and long-term glucose
control. Breast cancer screening—screening mammography rates; clinician breast exam measures (under development).
Consortium Members: University of Alabama at Birmingham (School of Medicine, School of Health Related
Professions, School of Nursing, and School of Public Health), and Alabama Quality Assurance Foundation, Inc.
Co-Principal Investigators and Institution:
Catarina I. Kiefe, Ph.D., M.D.
Norman W. Weissman, Ph.D.
University of Alabama at Birmingham
Project Dates: September 30, 1996 - September 29, 2001
Progress to Date: Measures have been selected for DM; feedback of ABCs to half the sample of outpatient physicians
was done (February 1997), post-feedback measurement of performance is scheduled; physician opinions on ABCs have
been surveyed (April 1997); inpatient data collection tool is in beta test. For breast cancer, records of 1,096 patients of
74 physicians have been examined to determine whether chart versus claims data were best for determining
mammographies performed. Baseline performance rates have been obtained for recommendation and performance of
mammography. Feedback intervention is being planned. ABCs may enhance the ability to progress from quality measurement to actual changes in clinical practice.
In addition, the project will further advance the methodology of deriving quality measures from claims data. Claims
data were found superior to chart abstracts in the recording of screening mammography.
For more information on the Achievable Benchmarks of Care project, select the University of Alabama at Birmingham's Center
for Outcomes and Effectiveness Research and Education
(COERE) Web site at http://www.dopm.uab.edu/coere/grant.html
Brief Description: The Health Plan Employer Data and Information Set (HEDIS), under the auspices of the National
Committee on Quality Assurance (NCQA), is the most widely used system for evaluating health plan performance.
The system nevertheless can be criticized for not addressing many important aspects of health care delivery and for
some deficiencies in individual measures. HEDIS 3.0, the current version, identifies two types of measures: a
"reporting set," on which health plans will be expected to collect data, and a "testing set" of measures that are
promising but are less well developed or need further evaluation. This project will evaluate and strengthen reporting set
measures, develop complete operational specifications for testing set measures that are strong candidates for the next
version of HEDIS, and evaluate their performance as measures.
Conditions considered: See measures below.
Populations: Enrollees of health plans reporting to NCQA.
Measurement approach: Mostly process measures.
Measures to be Developed or Evaluated: Screening for chlamydia; continuity of treatment for depression; followup
after abnormal mammogram or Pap smear; appropriate medication for asthma; prevention of stroke in people with atrial
fibrillation; flu shots for older adults; treatment of otitis media; screening for chemical dependency; monitoring
diabetes patients; antibiotics for HIV-related pneumonia; family visits for children under 12; stage of breast cancer
detection; substance abuse counseling for adolescents; and low birth weight.
Consortium Members: Harvard Medical School, National Committee on Quality Assurance, RAND, and The Center for
Survey Research.
Principal Investigators and Institution:
Arnold Epstein, M.D., M.A.
Paul Cleary, Ph.D.
Harvard University
Project Dates: September 30, 1996 - September 29, 1999
Progress to Date: Health plans have been selected. Data collection on reporting set measures began in January 1997.
Data collection on testing set measures will begin in Fall of 1997. Widespread adoption of HEDIS measures means that
findings about their validity, reliability, appropriate data sources, variability, etc., will have rapid application in
practice.
Brief Description: The purpose of this study is to develop methods for evaluating quality of care within two critical
transitional settings following acute hospitalization: subacute nursing home programs and home care programs. The
specific aims are: to operationalize the outcome measures; to identify the causal covariance-risk factors that influence
these outcomes and test the applicability of these elements within disease categories; to establish covariance adjusted
decline and improvement rate profiles for these measures across different time windows for the entire universe of
patients and within disease subgroups; to evaluate changes in outcome patient subgroup clusters, identified using an ex-post facto iso-outcome classification schema; and to test the feasibility and validity of outcome measures and
benchmark standards in a prospective study in selected subacute and home-care-based rehabilitation programs.
Conditions considered: Disease groupings to be studied include: chronic heart disease, stroke, chronic lung disease, hip fracture, and advanced or metastatic cancer.
Populations: Medicare beneficiaries who are: in both subacute and home care programs run by the Hebrew Rehabilitation Center for the Aged (HRCA) or with whom HRCA has a research laboratory relationship; clients in post-acute care settings managed by Prism Health Group and represented in a large secondary database using either the Minimum Data Set (MDS) or MDS-HC (home care).
Measurement approach: Outcome measures.
Measures to be Developed or Evaluated: Eight outcomes measures for five broad domains will be developed: cognition and communication, activities of daily living (ADL) and mobility, bowel and bladder function, mood, and pain.
Consortium Members: Hebrew Rehabilitation Center for Aged, University of Michigan, Beth Israel Hospital, and Prism Health Group.
Principal Investigator and Institution:
John Morris, Ph.D., Principal Investigator
Hebrew Rehabilitation Center for Aged
Project Dates: September 30, 1996 - September 30, 1999
Progress to Date: During the first year, the investigators have defined the final set of measures against which the eight outcome measures derived from the MDS 2.0 and MDS-HC will be validated. Research protocols have been written for each assessment instrument and a research manual has been created. Testing of all of the cross validation instruments was begun in June 1997.
For additional information on Q-SPAN, contact:
Judy Sangl
Center for Quality Measurement and Improvement
Agency for Health Care Policy and Research
540 Gaither Road, Suite 3000
Rockville, MD 20850
Phone: (301) 427-1308
E-mail: JSangl@ahrq.gov
Internet Citation:
Expansion of Quality Measures (Q-SPAN). Agency for Health Care Policy and Research, Rockville, MD. January 1998. http://www.ahrq.gov/qual/qspanovr.htm