Performance Measurement Accelerating Improvement (2006) / Chapter Skim
Currently Skimming:
Appendix F Commissioned Paper: Improving the Quality of Quality Measurement--John D. Birkmeyer, Eve A. Kerr, and Justin B. Dimick
Appendix F Commissioned Paper: Improving the Quality of Quality Measurement--John D. Birkmeyer, Eve A. Kerr, and Justin B. Dimick
Pages 177-203
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 177... ...
The clinical context (e.g., cancer screening vs. high risk surgery)
|
From page 178... ...
. Although the measures could be sorted on other dimensions, we consider them below according to whether they focus on ambulatory care (preventive care and chronic disease management)
|
From page 179... ...
APPENDIX F 179 TABLE F-1 Primary Strengths and Limitations of Structure, Process, and Outcome Measures Structure Process Outcomes Examples Procedure volume, Majority of HEDIS Risk-adjusted mortality intensivists managed performance measures rates for CABG from ICUs for ambulatory care state or national registries Strengths Expedient, inexpensive Reflect care that Face validity patients actually Efficient -- one measure Measurement alone receive -- buy-in from may relate to several may improve outcomes providers outcomes (i.e., Hawthorne effect) Directly actionable for More predictive of quality improvement subsequent perfor activities mance than other measures for some Don't need risk adjust procedures ment for many measures Positive "spillover" effect to other processes Limitations Limited number of Sample size constraints Sample size constraints measures, none for for condition-specific Concerns about risk ambulatory care measures adjustment with Generally not May be confounded by administrative data actionable patient compliance and Expense of clinical data other factors Don't reflect perfor- collection mance of individual Variable extent to hospitals or providers which process measures link to important patient outcomes Levels of metabolic control (e.g., intermedi ate outcomes)
|
From page 180... ...
Heart attack 1. Percentage of adults (35 years or older)
|
From page 181... ...
Percentage of patients on antidepressant medication for the entire 12-week Acute Treatment Phase 3. Percentage of patients who remained on treatment for a full 6 month trial Low back pain Percentage of patients aged 18 to 50 years old who received imaging studies for low back pain (Plain X-ray, CT scan, or MRI)
|
From page 182... ...
Percentage of 3-, 4-, 5-, and 6-year old children with a well-child visit within the past year Adolescent well- Percentage of 12- to 21-year-olds with at least one primary care or care visit OB/GYN visit during the past year Frequency of Frequency of selected procedures that have wide regional variation procedures Inpatient utilization Number of admissions for general hospital acute care Ambulatory care Number of ambulatory medical care visits Inpatient utilization Number of admissions for nonacute care (skilled nursing facilities, rehabilitation, transitional care, and respite) Postpartum care Number of discharges and average length of stay postpartum Births and newborns Number of births and average length of stay for newborns Mental health 1.
|
From page 183... ...
Percutaneous coronary intervention X X X X (within 30 minutes) ACE inhibitor at discharge for X X X patients with low left ventricular function Risk-adjusted mortality rates X X X Congestive heart failure Smoking cessation counseling X X X Standardized discharge instructions X X X Assessment of left ventricular function X X X ACE inhibitor at discharge for X X X patients with low left ventricular function Risk-adjusted mortality rates X Coronary artery disease Hospital volume -- Percutaneous X X X coronary interventions Bilateral cardiac catheterization X Gastrointestinal hemorrhage Risk-adjusted mortality rates X Community acquired pneumonia Smoking cessation counseling X X X Assessment of oxygenation at X X X admission Blood cultures prior to antibiotics X X X Antibiotics started within 4 hours X X X Appropriate initial choice of antibiotics X X X Pneumococcal screen or vaccination X X X Influenza screen or vaccination X X X Risk-adjusted mortality rates X Hip fracture Risk-adjusted mortality rates X Asthma Use of relievers (<18 years)
|
From page 184... ...
184 APPENDIX F TABLE F-3 continued Endorser Current Users NQF AHRQ JCAHO CMS Leapfrog Obstetric Diagnoses Pregnancy and neonatal care Rates of 3rd and 4th degree perineal X X X lacerations Risk-adjusted neonatal mortality X X Cesarean delivery in low risk women X X Vaginal births after cesarean delivery X X X Birth trauma X Hospital volume -- Neonatal X intensive care Neonatal immunizations after 60 days X Surgical Procedures Abdominal aneurysm repair Hospital volume X X Risk-adjusted mortality rates X Carotid endarterectomy Hospital volume X Esophageal resection for cancer Hospital volume X X Risk-adjusted mortality rates X Coronary artery bypass grafting Hospital volume X X X Risk-adjusted mortality rates X X X Internal mammary artery use X X Pancreatic resection Hospital volume X X Risk-adjusted mortality rates X Pediatric heart surgery Hospital volume X Risk-adjusted mortality rates X Hip replacement Risk-adjusted mortality rates X Craniotomy Risk-adjusted mortality rates X Cholecystectomy Laparoscopic approach X Appendectomy Avoidance of incidental appendectomy X JCAHO = Joint Commission on Accreditation of Healthcare Organizations; CMS = Center for Medicare and Medicaid Services; NQF = National Quality Forum; AHRQ = Agency for Healthcare Research and Quality; Leapfrog = The Business Roundtable's Leapfrog Group.
|
From page 185... ...
Its original standards focused on three structural measures: hospital volume for high-risk surgery and neonatal intensive care; computerized physician order entry; and intensivist staffing for critical care units. Their recently updated standards have been expanded to include selected process and outcome measures.
|
From page 186... ...
Many can be assessed easily with readily available administrative data. Although some structural measures require surveying hospitals or providers, such data are much less expensive to collect than measures requiring patient-level information.
|
From page 187... ...
Both historical and subsequent mortality rates are adjusted for Medicare patient characteristics.
|
From page 188... ...
outcomes (e.g., blood pressure control, glycemic control reflected by HbA1C levels) could be considered with processes or with outcomes.
|
From page 189... ...
However, assessing the results of cholesterol tests requires expensive medical record review and profiling based on control alone is plagued by problems of risk adjustment and possible perverse incentives. Instead, administrative data are often used to simply count how many cholesterol tests were obtained.
|
From page 190... ...
A few proprietary health care rating firms and state agencies are assessing risk-adjusted mortality rates using Medicare or state-level administrative datasets. However, most of the momentum in outcomes measurement involves large clinical registries.
|
From page 191... ...
It may also be useful for discouraging gaming, e.g., hospitals or providers avoiding high-risk patients to optimize their performance measures. However, it is not clear how much the scientific validity of outcome measures is threatened by imperfect risk adjustment with administrative data.
|
From page 192... ...
Over time, analytic methods will be refined. Access to higher quality data may improve with the addition of clinical elements to administrative data sets or broader adoption of electronic medical records.
|
From page 193... ...
(b) New York State Hospitals FIGURE F-2 Mortality associated with coronary artery bypass surgery in New York State hospitals, based on data from the state's clinical outcomes registry.
|
From page 194... ...
For example, the ultimate objective of the CMS pay-for-performance initiative with acute myocardial infarction, pneumonia, and congestive heart failure is improving quality at all hospitals. Conversely, the Leapfrog Group's efforts with selected surgical procedures and neonatal intensive care are primarily aimed at getting patients to hospitals likely to have better outcomes (selective referral)
|
From page 195... ...
Many measures meet this latter interest while failing on the former. For example, Krumholz and colleagues used clinical data from the Cooperative Cardiovascular Project to assess the usefulness of Healthgrades' hospital ratings for acute myocardial infarction (based primarily on riskadjusted mortality rates from Medicare data)
|
From page 196... ...
Even if few measures became "obsolete," rotating conditions and measures might be a useful approach to renewing interest in quality improvement initiatives while minimizing data collection burdens. Broader Measures of Performance In addition to condition- and procedure-specific performance measures, there is also interest in broader measures of quality for profiling health plans and organizations.
|
From page 197... ...
Hospital-Based Care Intensivist-staffed ICUs X Acute myocardial infarction Time to thrombolysis or PCI X Appropriate use of aspirin, beta-blockers, and X ACE inhibitors Risk-adjusted mortality rates, CABG X X Procedure volume, pancreatic resection and esophagectomy X Perioperative beta-blockage during noncardiac surgery X (high-risk patients) These measures incorporate data from a wide range of clinical conditions and have the added advantage of familiarity.
|
From page 198... ...
With the latter, for example, Hannan and colleagues noted that risk adjustment models derived from administrative data for CABG would approximate the reliability of those from clinical data with the addition of only three variables (ejection fraction, reoperation, and left main stenosis)
|
From page 199... ...
Thus, rather than gathering process of care data on all patients, such information would be collected on the smallest subset of patients necessary to achieve adequate precision. For outcome measures, clinical data could be sampled for risk adjustment purposes or to monitor for gaming, while administrative data are used to assess the complete numerator and denominator.
|
From page 200... ...
Top: conventional 30-day mortality rates; Bottom: filtered 30-day mortality rates.
|
From page 201... ...
or process of care measures on the reliability of predicting future performance. Broader Domains of Quality For procedures, the large majority of performance measures currently reflects aspects of technical quality.
|
From page 202... ...
2002. Evaluation of a consumer oriented Internet health care report card: The risk of quality ratings based on mortality data.
|
From page 203... ...
1996. Consistency in performance among primary care practitioners.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.