Synthesizing Lessons Learned From Get With The Guidelines

Abstract

HomeCirculationVol. 128, No. 22Synthesizing Lessons Learned From Get With The Guidelines Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissionsDownload Articles + Supplements ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toSupplemental MaterialFree AccessResearch ArticlePDF/EPUBSynthesizing Lessons Learned From Get With The GuidelinesThe Value of Disease-Based Registries in Improving Quality and Outcomes A. Gray Ellrodt, MD, Gregg C. Fonarow, MD, FAHA, Lee H. Schwamm, MD, FAHA, Nancy Albert, PhD, CCNS, CHFN, CCRN, NE-BC, FAHA, Deepak L. Bhatt, MD, MPH, FAHA, Christopher P. Cannon, MD, Adrian F. Hernandez, MD, MHS, Mark A. Hlatky, MD, Russell V. Luepker, MD, MS, Pamela N. Peterson, MD, MSPH, Mathew Reeves, BVSc, PhD and Eric Edward Smith, MD, MPH, FRCPC, FAHA A. Gray EllrodtA. Gray Ellrodt Search for more papers by this author , Gregg C. FonarowGregg C. Fonarow Search for more papers by this author , Lee H. SchwammLee H. Schwamm Search for more papers by this author , Nancy AlbertNancy Albert Search for more papers by this author , Deepak L. BhattDeepak L. Bhatt Search for more papers by this author , Christopher P. CannonChristopher P. Cannon Search for more papers by this author , Adrian F. HernandezAdrian F. Hernandez Search for more papers by this author , Mark A. HlatkyMark A. Hlatky Search for more papers by this author , Russell V. LuepkerRussell V. Luepker Search for more papers by this author , Pamela N. PetersonPamela N. Peterson Search for more papers by this author , Mathew ReevesMathew Reeves Search for more papers by this author and Eric Edward SmithEric Edward Smith Search for more papers by this author Originally published28 Oct 2013https://doi.org/10.1161/01.cir.0000435779.48007.5cCirculation. 2013;128:2447–2460Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: January 1, 2013: Previous Version 1 IntroductionThe American Heart Association/American Stroke Association (AHA/ASA) is a trusted source of scientific information in cardiovascular medicine. The AHA/ASA has a longstanding commitment to support state-of-the-art scientific research in cardiovascular disease and stroke. The AHA/ASA has also developed a leadership role in translating cardiovascular science into internationally respected guidelines. In 2000, however, the AHA/ASA concluded that, to provide maximal benefit for patients with cardiovascular disease and those at risk, it needed to develop a rigorous approach to translating its guidelines into clinical practice. The result was a comprehensive suite of programs collectively called Get With The Guidelines (GWTG). Modeled in part on the University of California, Los Angeles Cardiovascular Hospitalization Atherosclerosis Management Program (CHAMP),1,2 GWTG was successfully piloted by the AHA in Massachusetts.3 Based on the success of the Massachusetts pilot, the AHA committed significant human and financial resources to extend the program across the United States. This commitment included the development of a national steering committee composed of AHA/ASA volunteers, and the addition of multiple modules including hospital-based management of coronary artery disease, heart failure, stroke, and resuscitation after in-hospital cardiac arrest. The scientific foundation of the program is the best evidence from the latest American College of Cardiology/AHA/ASA guidelines. GWTG staff work with participating hospitals to implement these guidelines by using AHA/ASA quality improvement professional consultation, workshops, and Webinars. In addition, the AHA developed sophisticated clinical databases (registries) through which hospitals and physicians collect information in real time for the assessment of quality, regional, and national benchmarking, national recognition, and the generation of new science. The AHA underwrites a portion of the costs associated with the technology platform and data collection tools to reduce the financial burden on participating sites.From the 4 GWTG disease-specific registries, >200 articles have been published in peer-reviewed journals. This statement will distill the information from those articles into a concise review of particular use to policy makers throughout the country and world. The scope and impact of GWTG program is profound. Of the 6280 hospitals registered with the American Hospital Association, 1956 US hospitals participate in at least 1 GWTG program (31%).4 Many institutions use ≥2 modules (Figure 1). These hospitals have submitted almost 5 million patient records since the inception of GWTG, with >2 million patients in the Stroke registry alone (Figure 2). In certain diseases, such as stroke, GWTG now enrolls >40% of all index hospitalizations in the United States annually (Figure 2).Download figureDownload PowerPointFigure 1. Growth in GWTG hospital participation. The American Heart Association GWTG-CAD registry was merged into AR-G in 2011. Only admission records counted as part of AR-G are listed in this figure. ARG (CAD) indicates ACTION Registry-GWTG Coronary Artery Disease; FY, fiscal year; GWTG, Get With the Guidelines; HF, GWTG-Heart Failure; Resus, GWTG-Resuscitation; Stroke, GWTG-Stroke; and Uniques, unique hospitals participating.Download figureDownload PowerPointFigure 2. Growth in total patient admission records in GWTG, overall and by module. The American Heart Association GWTG-Coronary Artery Disease registry was merged into AR-G in 2011. Only admission records counted as part of AR-G are listed in this figure. AR-G indicates ACTION Registry-GWTG; FY, fiscal year; GWTG, Get With the Guidelines; HF, GWTG-Heart Failure; Resus, GWTG-Resuscitation; and Stroke, GWTG-Stroke. Source: Outcome: A Quintiles Company.This statement will also highlight lessons learned from GWTG and elucidate their policy implications in 4 distinct sections. The first section presents an overview of the GWTG program and evidence that the GWTG performance improvement system can be implemented nationally and is associated with a substantial improvement in quality of care over a broad range of cardiovascular conditions in multiple settings. The second section addresses the power of assembling GWTG data to (a) generate new knowledge, (b) identify areas for future quality improvement efforts and measures, (c) drive measure prioritization and construct, (d) improve the development of measures, guidelines, and implementation strategies to reduce racial and sex disparities, and (e) determine the safety and effectiveness of therapies applied in routine practice. The third section presents evidence that GWTG has made a contribution to increasing the value of health care for our hospitals, payers, and our patients. We specifically explore reductions in length of stay and readmission rates and improvements in patient safety. In the fourth section, we address how the linkage of the GWTG clinical database with administrative data (eg, Medicare fee-for-service claims data) fosters real-world clinical effectiveness research (CER) and can be used to assess current practice, identify focused areas of health disparities, and develop successful implementation strategies to address opportunities to improve care. The final 2 sections address the limitations of GWTG and summarize policy recommendations.National Implementation and Impact on Quality of CareThere are substantial gaps in quality of care for cardiovascular disease, including disparities by sex, race/ethnicity, and geographic region.5 The GWTG program was launched to improve in-hospital quality of care and reduce disparities by providing a platform for hospitals to review their own data with hospital-selected regional and national benchmarks. The program provides clinical decision support tools, Webinars, conferences focused on quality improvement, and other educational materials. In addition, the program provides expert AHA field staff to assist providers in developing sophisticated quality improvement programs.3,6 For more detail on the scope of activities in the GWTG program readers are referred to references 3, 6, and 7. In addition to the significant commitment of resources by the AHA/ASA, hospitals must commit quality improvement professionals and data abstraction capabilities to be successful. By also facilitating real-time hospital and individual physician access to benchmarked performance data, providers can compare their performance against other hospitals based on a large selection of variables. These include benchmarking against all GWTG hospitals, hospital systems, geographic regions, hospital teaching status, bed size, and a variety of other hospital characteristics.GWTG uses chart abstraction of clinical data elements into a Web-based platform provided by Outcome: A Quintiles Company (Cambridge, MA). Case ascertainment is either concurrent with the use of clinical criteria or retrospective based on coding. Process measures are divided into achievement (performance) measures and quality measures. Achievement measures have the strongest supporting evidence and strongest link between the process and health outcomes. Most of these measures are endorsed by the National Quality Forum (NQF). These measures drive the GWTG Performance Achievement Award program and are shown in Tables 1, 2, and 3. Quality measures are supported by strong evidence, but not as robust as that supporting achievement measures. These quality measures are important future candidates for achievement measures, and extend the scope of interventions available to already high-performing hospitals. Finally, these quality measures can support recent changes to the NQF endorsement process requiring previous field testing of candidate performance measures.Table 1. Time Trends in GWTG-ACS Quality-of-Care Measures, 2006–201110Quality of Care Measure2006*2007*2008*20092010*2011*Aspirin within 24 h of admission94.792.891.290.99797.6Aspirin at discharge94.495.894.995.59898.3β-Blockers at discharge92.894.694.594.99696.7Lipid-lowering medication at discharge84.585.681.686.892†98.4†Lipid therapy at discharge if LDL cholesterol >100 mg/dL89.190.791.992.5NMNMARB/ACEI at discharge for patients with LVEF <40%87.391.191.991.98687.8Adult smoking cessation advice/counseling94.397.498.498.49898.4Cardiac rehabilitation referral for AMI patients71.163.652.049.17576.5Values are percentages.AMI indicates acute myocardial infarction; ARB/ACEI, angiotensin receptor blocker/angiotensin-converting enzyme inhibitor; GWTG-ACS, Get With The Guidelines–Acute Coronary Syndrome; LDL, low-density lipoprotein; LVEF, left ventricular ejection fraction; and NM, not measured.*The American Heart Association (AHA) GWTG-Coronary Artery Disease (CAD) registry was merged into the Acute Coronary Treatment and Intervention Outcomes Registry (ACTION) registry. Therefore, measures from 2006–2009 are from the AHA’s GWTG-CAD registry. 2010/2011 measures are from the ACTION registry.†Represents statin use.Reproduced with permission from Go et al.10 Copyright © 2012, American Heart Association, Inc.Table 2. Time Trends in GWTG-HF Quality-of-Care Measures, 2006–201110Quality of Care Measure200620072008200920102011LVEF assessment*93.896.296.898.29899.2ARB/ACEI at discharge for patients with LVSD*85.589.191.693.094.295.4Complete discharge instructions†78.884.888.590.993.393.5Postdischarge appointment (new for 2011)*–––––13.3Adult smoking cessation advice/counseling†90.894.797.197.699.399.2β-Blockers at discharge for patients with LVSD, no contraindications†89.990.292.592.794.896.2Evidence-based specific β-blockers*67.758.954.145.248.458.4Anticoagulation for atrial fibrillation or atrial flutter, no contraindications62.961.660.768.970.275.4Values are percentages.ARB/ACEI indicates angiotensin receptor blocker/angiotensin-converting enzyme inhibitor; GWTG-HF, Get With The Guidelines–Heart Failure; LVEF, left ventricular ejection fraction; and LVSD, left ventricular systolic dysfunction.*Indicates the 4 key achievement measures targeted in GWTG-HF.†Indicates historical key achievement measures in GWTG-HF.Reproduced with permission from Go et al.10 Copyright © 2012, American Heart Association, Inc.Table 3. Time Trends in GWTG-Stroke Quality-of-Care Measures, 2006–201110Quality of Care Measure (%)200620072008200920102011IV tPA in patients who arrived ≤2 h after symptom onset, treated ≤ 3 h*55.860.263.973.176.278.3IV tPA in patients who arrived <3.5 h after symptom onset, treated ≤4.5 h†––––42.557.9IV tPA Door-to-Needle Time ≤60 min22.524.925.928.029.533.8Thrombolytic complications: IV tPA and life-threatening, serious systemic hemorrhage20.817.316.115.113.115.7Antithrombotics <48 h after admission*94.895.896.096.296.396.7DVT prophylaxis by second hospital day*85.388.992.292.792.293.5Antithrombotics at discharge*94.195.197.097.897.798.1Anticoagulation for atrial fibrillation at discharge*88.289.593.193.593.593.1Therapy at discharge if LDL cholesterol >100 mg/dL or LDL cholesterol not measured or on therapy at admit*70.376.382.186.288.189.8Counseling for smoking cessation*86.192.294.396.296.797.0Lifestyle changes recommended for BMI >25 kg/m242.545.751.757.357.857.8Composite quality of care measure85.988.991.793.393.794.4Values are percentages.BMI indicates body mass index; DVT, deep venous thrombosis; GWTG-Stroke, Get With The Guidelines–Stroke; IV, intravenous; LDL, low-density lipoprotein; and tPA, tissue-type plasminogen activator.*Indicates the 7 key achievement measures targeted in GWTG-Stroke.†New quality measure subsequent to the European Cooperative Acute Stroke Study III published in 2008.Reproduced with permission from Go et al.10 © Copyright 2012, American Heart Association, Inc.Measures are developed by physician volunteers and AHA staff and are harmonized with measures developed by the American College of Cardiology/AHA, US Department of Health and Human Services, and The Joint Commission and those already endorsed by the NQF. For each measure, hospitals report the proportion of eligible patients receiving the intervention divided by the total number of patients eligible for the measure without explicit documented contraindications. A performance achievement award program with public recognition is designed to provide further incentives to participating hospitals. Tiered award levels are based on the number of quarters or years of consistent performance at a level of at least 85% adherence for all achievement measures, with additional opportunities for distinction in the quality measures domain.7Although GWTG is a voluntary program, participation has grown remarkably in each disease module since program inception (Figures 1 and 2). The high uptake of GWTG within the stroke community is likely driven by recommendations from the Brain Attack Coalition and ASA that primary stroke centers should engage in ongoing continuous quality improvement, and by the proliferation of state regulations or legislation requiring designation of centers that receive acute stroke patients.8,9 In addition, many hospitals use GWTG-Stroke as the platform for data collection for submission to The Joint Commission, to the Centers for Disease Control and Prevention Paul Coverdell Registry, or to public health authorities as part of state-based regulatory or quality improvement initiatives. In 2012, the AHA/ASA and The Joint Commission developed and launched the Comprehensive Stroke Center program based on consensus guidelines and performance measures supported by AHA scientific statements and the Brain Attack Coalition, parallel to the process used to design the primary stroke center process. The data collection specifications for the Comprehensive Stroke Center program are in addition to the basic primary stroke center requirements and are in pilot phase in 2013, after which they will be finalized.Outcome: A Quintiles Company serves as the data collection (through their Patient Management Tool) and coordination center for GWTG. The Duke Clinical Research Institute serves as the data analysis center and has an agreement to analyze the aggregate deidentified data for research purposes.Examples of the rapid and sustained year-over-year improvement in evidence-based care for stroke, heart failure, and coronary artery disease are shown in Tables 1, 2, and 3.10 In some instances, quality of care was poor at program inception with a substantial potential for improvement.An audit using central reabstraction of GWTG-Stroke data suggests that hospitals of all types and sizes are accurately reporting their data without a bias toward overestimating their own performance.11 Hospital-reported improvement in care is driven by real changes rather than by changed documentation or judgments of eligibility.12 The GWTG Steering Committee plans ongoing audits to ensure the validity and reliability of the GWTG databases. In addition, the databases contain optional data fields to drive hospital internal performance which have varying degrees of completeness. These fields do not, however, include the variables needed for quality measure calculation or recognition.GWTG includes large numbers of academic and community hospitals in urban and rural settings in all 50 states, the District of Columbia, and Puerto Rico. In the 1956 hospitals participating in GWTG, median bed size is 243, 13% have bed size <100, 34% are academic, 46% are nonteaching, and 78% are urban and 20% are rural. In the United States, 54% of hospitals have <100 beds, 24% are teaching, and 65% are urban.4 Thus, the GWTG hospital distribution is more heavily weighted toward larger, teaching, and urban hospitals. This diversity demonstrates the feasibility of implementing GWTG in all types of hospitals,13 but the different distribution raises questions about blanket applicability at the hospital level. The patient population, however, appears to be representative of the overall US population for stroke patients.14Hospital retention to date has been excellent (mean of 90% over the past 3 years), but important challenges in sustaining GWTG and promoting additional program growth exist. Manual chart abstraction yields high-quality, reliable clinical data, but is time consuming and must be self-funded by the participating hospitals. In response to this challenge, GWTG has developed tools to facilitate data abstraction and has partnered with electronic health record vendors to support batch uploading and data integration. Central program coordination and database maintenance and analysis, although very efficient, are still costly. Policies that provide incentives for hospital quality improvement (eg, through accreditation programs and pay-for-performance incentives with financial support for participating hospitals) could further enhance the growth and sustainability of GWTG.Hospital participation in GWTG is strongly associated with rapid and sustained improvement in evidence-based care in participating hospitals. All the GWTG modules have been associated with significant improvements in multiple processes of care strongly linked to improved outcomes.15,16 These improvements are summarized in Tables 1, 2, and 3. These changes have been rapid and sustained over many years. In addition, GWTG has produced multiple specific improvements in guideline adherence documented in the specific sections that follow.Facilitate Synthesis of Data to Better Understand Practice Patterns and Disease StatesGenerate New Knowledge and UnderstandingInformation about healthcare trends and outcomes are critical to our understanding of and improvement in the treatment and prevention of acute and chronic diseases.17 The GWTG program is designed to improve the care of individual patients and to facilitate the rapid evaluation of current practices at the provider, clinic, hospital, and regional level.Through its ongoing real-world surveillance function, GWTG can serve as a data source for agencies that track and report disease incidence and burden.18 Many recognize that research and its synthesis into guidelines do not have the same results or outcomes in real-world practice. This may be because research results are not applied in the same way as in a trial, diffuse slowly through the medical community, or are not applied at all.19 Surveillance of practice allows the characterization of these changes over time, the opportunity to accelerate improvement and demonstrate differences between practitioners and geographically diffuse healthcare institutions.Disparities in cardiovascular health and health care in the United States are well documented in the 2003 Institute of Medicine report, Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care.20 These inequalities continue and include differences by sex, education level, income, insurance, and geographic location.21 For example, a recent review found that blacks were less likely to be admitted to a coronary care unit, spent less time in the unit, and were likely to be discharged after a shorter hospital stay.22GWTG studies have shown both reductions in and elimination of disparities in the delivery of hospital-based cardiovascular and stroke care. In each case, new knowledge emerged from the GWTG analyses, and valuable lessons learned paved the way for new GWTG strategies to reduce health disparities elsewhere. For example, GWTG researchers reported substantial underuse of implantable cardiac defibrillators (ICDs) in eligible women and black patients in comparison with white male patients with systolic heart failure.23 The program then focused educational efforts in workshops and Webinars on the importance of ICD placement in eligible patients.24 From 2005 to 2009 in GWTG-Heart Failure–participating hospitals, the authors observed a significant increase in ICD therapy use in all sex and race groups, eliminating the observed disparities.24 In hospitals participating in GWTG-Stroke there were substantial improvements in acute ischemic stroke care demonstrated over time in all race/ethnic groups (Figure 3).25Download figureDownload PowerPointFigure 3. Equitable improvement in quality of care among race/ethnic groups with stroke: defect-free measure adherence for Get With the Guidelines Hospitals. Defect-free care means the percentage of patients in whom all interventions for which the patient is eligible are provided. This is also called perfect care %. Reproduced with permission from Schwamm et al.25 Copyright © 2010, American Heart Association, Inc.The GWTG platform is also a valuable tool in tracking the rate of diffusion of new knowledge and the adoption of new guideline recommendations. GWTG researchers demonstrated that the proportion of eligible patients with acute ischemic stroke treated with intravenous tissue-type plasminogen activator (tPA) within 4.5 hours of onset increased rapidly and substantially after publication of study results and an ASA advisory.26 GWTG was able to facilitate rapid adoption, track the rate of adoption, and identify areas for further improvement.Finally, the GWTG program has engaged young investigators in the examination of registry information, generation of new ideas, and publication of articles that have not only contributed to new knowledge and understanding, but also have helped launch young investigator careers. To date, the program has awarded 41 young investigator awards.Essential for Identifying Areas for Future Quality Improvement Efforts/MeasuresData generated by the GWTG program are essential for identifying areas for future quality improvement efforts and quality measurement. Analyses of GWTG data have identified gaps in evidence-based care. With awareness of such gaps, quality measures can be highlighted and adherence subsequently tracked over time. This iterative process makes the GWTG registries dynamic—they not only report data, but they facilitate quality improvement by providing feedback to sites about their individual performance in comparison with aggregate data.As noted, registry data provide real-world information about how patients are actually being treated in a way that randomized clinical trials cannot.27 For example, GWTG-Heart Failure demonstrated dramatic underuse of aldosterone antagonists despite randomized clinical trial data supporting their use.28 Only one third of eligible patients without any documented contraindications were prescribed aldosterone antagonists on discharge.29 By incorporating aldosterone antagonist prescription in appropriate patients into Target: Heart Failure, an extension of GWTG, this and other clinical opportunities can be addressed and performance improved.Despite compelling evidence and national guidelines, a pivotal GWTG report identified immense hospital variation in the percentage of ischemic stroke patients who received their thrombolytic dose within 60 minutes of hospital arrival.30 Based on these data, GWTG-Stroke has implemented Target: Stroke, a national quality improvement initiative to reduce door-to-needle time in intravenous tPA use through the use of 10 key evidence-based strategies.31 The Target: Stroke program is now assessing the ability of the 1200 participating hospitals to achieve the 60-minute door–to–tPA target in 50% of patients.Drive Prioritization of Measures, Identify/Test New Measures, and Determine Measure ConstructThe AHA/ASA and GWTG have helped convene key stakeholders to promote harmonization and definition of evidence-based measures with strong process outcome links. Data from GWTG have been invaluable in updating American College of Cardiology Foundation/AHA performance measure sets for acute myocardial infarction and for heart failure. In addition, the AHA/ASA helped convene stakeholders to identify and prioritize measures forming the basis of stroke quality improvement over the next decade. Over the past decade, various healthcare accreditation and quality improvement organizations in the United States have become actively involved in improving the quality of care for stroke patients.32 These early efforts centered on the development and standardization of achievement measures designed to measure specific aspects of the structure, process, and outcomes of acute stroke care. Most of these measures were developed based on the recommendations of evidence-based clinical guidelines,33,34 and the benchmarks proposed by the Brain Attack Coalition for the establishment of primary stroke centers, as well.8 In 2003, the Stroke Performance Measure Consensus Group—a collaboration between the AHA/ASA, The Joint Commission, and Centers for Disease Control and Prevention—began the essential process of harmonizing definitions and abstraction rules for stroke performance measures. The consensus process resulted in 10 harmonized acute hospital care stroke performance measures. Eight of the 10 measures were subsequently endorsed by the NQF, which acts as an independent authority and clearinghouse for the establishment of quality-related measures.35 Centers for Medicare & Medicaid Services (CMS) will likely include the 8 NQF-endorsed stroke measures as part of its core measure pay-for-reporting initiative beginning with inpatient discharges in fiscal year 2015.36,37 These publicly reported performance measures will be available to consumers on the CMS Hospital Compare Web site.38 Additionally, physician-level stroke performance measures have been endorsed by the NQF for use in the CMS Physician Quality Reporting Initiative program, an incentive pay-for-reporting program.39The GWTG-Stroke Program has played an essential role in the development and tracking of these stroke performance measures; data from the GWTG-Stroke registry, as well as other stroke quality improvement initiatives,40,41 have served to illustrate the importance of ongoing and systematic data collection. For example, analysis of >320 000 ischemic stroke patients treated at 800 GWTG-Stroke hospitals between 2003 and 2008 found clinically meaningful improvements in all 7 measures examined with absolute increases ranging from 3.2% to 30.7%.15 Based on the strength of the GWTG program, and its reputation for reliable data collection, several states now use GWTG as a required component for state stroke center designation, and the vast majority of the states participating in the Centers for Disease Control and Prevention’s Paul Coverdell National Acute Stroke Registry also use GWTG as their data entry and quality improvement vehicle. These public–private partnerships are another critical example of how collaboration helps to identify and refine appropriate measures and ensure effective and broad implementation.Improve Development of Process of Care Measures, Guidelines, and Their ImplementationThe GWTG program and database include processes of care measures that inform clinicians about practice patterns, the level of equitable care, and adherence to clinical practice guidelines, and facilitate clinical quality improvement. In 1 report of Medicare patients, hospitals enrolled in the GWTG-Heart Failure registry demonstrated better processes of care than other hospitals and had lower 30-day readmission rates.42 In specific patient populations, including women,43 Hispanics,44 and Asian Americans,45 a decrease or elimination of sex and race/ethnic disparities was observed. GWTG researchers found that lower-risk patients were referred more often to disease management programs and that quality and performance measures were higher in those referred.46 Such data could be used to optimize the selection of patients for referral. In patients with acute coronary syndromes, learning that intensive lipid-lowering therapy was not prescribed at hospital discharge for most eligible patients47 provides evidence that greater attention is needed in the implementation of guideline-re