Using the AHRQ Quality Indicators for Quality Improvement
Day 1: Tuesday, September 27, 2005
View presentations at
1. Overview of the AHRQ Indicators- Speakers Irene Fraser, Center for Delivery and
Marybeth Farquhar, AHRQ
Measures and data can improve with use. “Good” measures and data can get better
(though not perfect) but even good measures with bad data can create mischief.
Currently, there is no gold standard and clinical, administrative, patient experience of
care data all have strengths and weakness.
AHRQ’s Measurement Initiatives assist with national tracking and benchmarks,
measuring local experience of care, measuring culture of safety, physician
measures, measuring hospital quality and safety, and measure potentially avoidable
The current AHRQ Quality Indicators are Prevention Quality Indicators, Inpatient
Quality Indicators, and Patient Safety Indicators. AHRQ is looking to expand the Quality
Indicators to include pediatric measures, women’s health measures, readmissions,
emergency department quality, and various other indicators.
The vision for AHRQ’s Quality Indicator Initiative is to develop, maintain, and evolve measures; strengthen administrative data at federal, state, local levels; create tools to
facilitate use; and bring change through strategies and partnerships.
The objective of the AHRQ Quality Indicators is to provide a tool to highlight potential
quality concerns; identify areas that need further study and investigation; track
changes over time; facilitate transparency through comparative information about the
quality of healthcare; facilitate decision making; and maximize existing resources.
The goal for the meeting was to help the participants learn from each other and provide
input as AHRQ refines their strategic vision in order to deliver what’s most useful.
2. Overview of the Pediatric Indicator Module - Kathryn and Sheryl Davies, Stanford
In 2000, there were 6.3 million children hospitalizations. The majority of the
hospitalizations were for premature newborns. The total cost of the hospitalizations was
$46 billion. The pediatric population is a unique population in that children are either
very well or sick. They also make up a large percentage of the poor population. They are
also unique clinically in that the coding is different than adults. The majority of the
pediatric population receives outpatient care than inpatient care.
Due to this population being unique, AHRQ is designing a pediatric indicator module. The indicators under consideration are intraventricular hemorrhage, respiratory distress syndrome, chronic respiratory disease, meconium aspiration syndrome rate, necrotizing enterocolitis, neonatal morality, noscocomial bacterima, proportion of VLBW infants born at Level III centers, and retinopathy of prematurity.
The Patient Safety and Mortality indicators under construction are aspiration pneumonia, postoperative pneumonia, catheter-associated venous thrombosis, other postoperative metabolic derangements, and trauma mortality.
The timeline for the PedQI software release with current AHRQ QIs adapted for pediatric cases is January 2006.
3. Session I- Using the AHRQ Area Level Indicators to Improve Population Health
Sandra Mahkorn, Wisconsin Department of Health and Family Services- Speaker
Wisconsin Medicaid Managed Care Program use Prevention QIs for quality improvement. The programs that use the QIs are the frail elderly, disabled, and chronic disease and disability programs.
Hospital admission rates and overall numbers of hospital days associated with seven chronic and acute Prevention QIs were used to measure the quality of care. Wisconsin used Prevention Quality Indicators to guide quality activities. The Prevention QIs provided Wisconsin with useful information on how to prevent hospitalizations.
The Prevention QIs helped the state and managed care programs to assess their effectiveness in reducing hospitalizations after members entered their programs; allowed managed care programs to compare their results with other programs; allowed programs and the state to track progress over time; and provided information that allowed programs to set quality improvement priorities.
Susan McBride, Dallas-Fort Worth Hospital Council
The Prevention Quality Indicators are used by the Dallas-Fort Worth Hospital Council to assess community health. The Council has designed an interactive website for Dallas Fort Worth Council Members to examine trends on AHRQ measures and public hospital discharge data. The site allows for users to drill into the numerators that are posted for the Prevention QIs.
The next step for the Council is to use the indicators at the regional level to examine overall performance and health trends; partner with the Department of Health, Public Health and Schools of Public Health to better utilize the measures to improve the health of the populations served; find funding to distribute data sharing capability; pursue ambulatory data projects; support Texas efforts for public reporting of hospital infection rates; and develop community interventions to address health concerns.
Sam Shalaby, General Motors
AHRQ partnered with General Motors to research quality and cost drivers of health care. AHRQ measured health care quality of the GM employees by using the PQI, IQI, and PSI measures.
AHRQ was able to provide GM cost data tables that detailed the average cost per discharge for each indicator in the Michigan area. The table displayed the number of discharges per year, total costs, and potential cost savings if the number of discharges were reduced by 10%, 20%, 30%, 40%, and 50%.
The proposed actions that came from this research was to integrate action plans with other Community Initiative projects; consider Pay for Performance for providers in specific counties; dovetail with Save Dollars/Save lives Project; and focus on the vital few projects (PTCA, CABG, CHF, Bacterial Pneumonia, COPD, & Diabetes).
4. Session II- Using the AHRQ Provider Level Indicators as a Catalyst for Quality Improvement
Ben Yandell, Norton Healthcare
Norton Healthcare provides consumers with all the National Quality Form indicators, JCAHO measures and patient safety goals; AHRQ PSIs and IQIs; and other measures such as pediatric ORYX and NICU mortality.
They publicly report every 12 months risk-adjusted rates using the AHRQ software. The source of the data is the Kentucky hospital discharge databases. They also create service line report cards.
The organization is unique in that they review charts to verify the accuracy of the data. They rarely find coding errors when they compare the administrative data to the medical charts.
Ben Yadell’s final thoughts were that data do not become valid until used; the number is what the number is; and even lousy indicators improve care.
5. Session II- Using the AHRQ Provider Level Indicators as a Catalyst for Quality Improvement
Carol Munsch, Covenant Healthcare
The Convenant Healthcare organization uses the Patient Safety Indicators. The Prevention Indicators have been used to manage diabetes in the patient and employee population. They also have resulted in the opening new comprehensive diabetes centers.
The organization has also learned from the use of the Prevention Indicators that some indicators still need more validation; coding problems are always a minor annoyance; avoid the tendency to react to measures in isolation; and watch for small samples, knee jerk reactions.
Joanne Cuny, University Health System Consortium
Ms. Cuny discussed the best practices to make a difference in preventing avoidable
deaths with rapid rescue teams. This relates to the Failure to Rescue quality measure. They were able to devise a best practice plan that involved a chain of command from noticing early warning signs, assessing the situation, rapid communication up chain of command; and the rapid response of the team to address the situation.
6. Session III- Implications of ICD-9-CM Coding Rules for Measuring QIs
Patrick Romano, UC Davis
Patrick Romano explained that in many cases the physician might misuse coding which makes it appear that something happened that did not. He stated that the coders in many cases are not the problem. He also discussed the many ways in which coding may present a problem. One such problem may be that in some events multiple codes should be used to specify the event. He also warned of overcoding that may present a problem. Mr. Romano discussion was very good and explained how if a record is coded wrong that it may skew that data in which we report using the AHRQ measures.