- Email: [email protected]
Designing a Sustainable National Registry for Stroke Quality Improvement
Designing a Sustainable National Registry for Stroke Quality Improvement Lee Schwamm, MD, Mathew J. Reeves, PhD, Michael Frankel, MD Abstract:
Several studies have shown wide variations in the delivery of hospital-based care to patients with acute ischemic stroke. The findings of these studies suggest that recommendations drawn from published evidence-based or consensus-based guidelines are implemented inconsistently. Although rates of adherence to stroke quality indicators can be increased through the use of targeted quality improvement (QI) efforts, stroke QI programs are still in their infancy. Current stroke QI programs are often highly variable and poorly coordinated, rely on differing definitions of key data elements and/or indicators, and are challenging to implement and sustain in resource-constrained healthcare environments. Key barriers to long-term success of these programs include inadequate funding at the local and national level, lack of infrastructure to support electronic data capture for QI as part of the process of patient care, lack of a single clearinghouse for uniform data definitions and performance indicator descriptions, competing survey instruments to monitor hospitalized stroke care, and constraints on inpatient and post-discharge data collection imposed by the new Federal Health Insurance Portability and Accountability Act Privacy Rule. In addition, the competing needs of registry activities (e.g., complete case ascertainment) versus QI efforts (e.g., incremental tests of change) must be balanced. Potential solutions include: (1) financial incentives to healthcare providers and institutions for participation in QI initiatives; (2) financial incentives to healthcare providers and institutions for measurable improvements in care; (3) mandatory data reporting on key measures of stroke care; and (4) promotion of active and sustainable collaborations among key stakeholders including healthcare providers (e.g., physicians, nurses), healthcare organizations (e.g., hospitals, physicians’ groups), quality improvement organizations, health payers and insurers, public health departments, and state and federal health agencies to create a single national stroke registry for stroke QI. (Am J Prev Med 2006;31(6S2):S251–S257) © 2006 American Journal of Preventive Medicine
Rationale for Stroke Quality Improvement
T
here is abundant scientific evidence that supports consensus- and evidence-based guidelines for in-hospital and postdischarge care in cardiovascular and cerebrovascular disease management.1– 4 Despite ready access to published guidelines and widespread health-provider awareness of the recommended interventions, guidelines are poorly implemented in both in-hospital and postdischarge environments.5– 8 Some studies suggest that providers overestimate both likelihood and severity of complications of therapy in patients of advanced age or frail health when compared to the risk of recurrent stroke.9 Risk factors are adequately controlled in only 25% of patients, despite the From the Department of Neurology, Massachusetts General Hospital, Harvard Medical School (Schwamm), Boston, Massachusetts; Department of Epidemiology, Michigan State University (Reeves), East Lansing, Michigan; Department of Neurology, Grady Hospital, Emory University (Frankel), Atlanta, Georgia Address correspondence and reprint requests to: Lee H. Schwamm, MD, Department of Neurology VBK-915, Massachusetts General Hospital, 55 Fruit Street, Boston MA 02114. E-mail: [email protected].
fact that up to 75% of cardiac and cerebrovascular disease is preventable with successful risk-factor management.10 Table 1 lists the common modifiable and nonmodifiable risk factors for ischemic stroke and transient ischemic attack. The effort to close the gap between what is known and what is practiced has been the focus of several major studies and initiatives in recent years. The influential Institute of Medicine report Crossing the Quality Chasm12 offered a set of performance expectations for the 21st century healthcare system, a set of new rules to guide patient-to-clinician relationships, an organizing framework to better align incentives inherent in payment and accountability with improvements in quality, and key steps to promote evidence-based practice and strengthen clinical information systems. In another initiative, the National Institutes of Health Roadmap attempts to reorganize the federal funding of medical research to address some of the barriers inherent in translating not only the results of basic research but also the results of clinical trials
Am J Prev Med 2006;31(6S2) © 2006 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/06/$–see front matter S251 doi:10.1016/j.amepre.2006.08.013
Table 1. Common risk factors for ischemic stroke and TIA and recommendations for secondary prevention interventions in adults Risk factor Age Race/ethnicity Gender Heredity Hypertension Diabetes
Atrial fibrillation (AF) Noncardioembolic ischemic stroke or TIA Coronary heart disease Dyslipidemia
Cervical internal carotid artery stenosis
Cigarette smoking Alcohol use Obesity Physical inactivity
Recommendations for secondary stroke prevention in adult patients with ischemic stroke/TIA Nonmodifiable Nonmodifiable Nonmodifiable Nonmodifiable Antihypertensives after the hyperacute period (I-A) along with lifestyle modifications (IIb-C). Optimal drug regimen uncertain but diuretics and the combination of diuretics and an ACEI are effective (I-A) Rigorous control of dyslipidemia and HTN, with BP targets of 130/80 mm Hg (IIa-B). ACEIs and ARBs are first-choice medications (I-A). Glucose control is recommended to near normoglycemic levels to reduce microvascular complications (I-A) and possibly macrovascular complications (IIb-B). Hemoglobin A1c goal ⬍7% (IIa-B) For persistent or paroxysmal (intermittent) AF, use adjusted-dose warfarin (INR 2.5; range, 2–3) (I-A). For patients unable to take oral anticoagulants, aspirin 325 mg per day (I-A) Antiplatelet agents rather than oral anticoagulation to reduce risk of recurrent stroke and other cardiovascular events (I-A). Options for initial therapy (IIa-A) include aspirin (50–325 mg qd), aspirin plus extended-release dipyridamole (25/200 mg bid), clopidogrel (75 mg qd) Antiplatelet agents (I-A), lipid-altering agents per NCEP III (I-A) and management of comorbid risk factors for atherosclerosis as outlined Patients with elevated cholesterol, coronary heart disease, or evidence of an atherosclerotic etiology should be managed according to NCEP III (I-A). Statins are recommended with target LDL-C of ⬍100 mg/dL and ⬍70 mg/dL for the very high-risk (I-A). Reasonable to consider statins to reduce risk of vascular events in those without NCEP indication (IIa-B) For ipsilateral severe stenosis (70%–99%), CEA is recommended (I-A) and CAS may be considered (IIb-B) if (1) difficult to access surgically, (2) medical conditions exist that greatly increase risk for surgery, or (3) with radiation-induced stenosis or restenosis after CEA and (4) when performed by operators with morbidity and mortality rates of 4% to 6% (IIa-B). For ipsilateral moderate stenosis (50%–69%), CEA is recommended depending on age, gender, medical comorbidity, and severity of symptoms (I-A). For ipsilateral mild stenosis (⬍50%), there is no indication for CEA or CAS (III-A). Smoking cessation through use of counseling, nicotine replacement, or other medications (I-C) Heavy drinkers should eliminate or reduce (⬍2 drinks/day males, ⬍1 drink/day females) their consumption of alcohol (I-A) Consider weight reduction through increased activity and decreased caloric intake for all overweight patients to maintain goal BMI of 18.5 to 24.9 (calculated as kg/m2) or waist circumference ⬍40 in (102 cm) for men and ⬍35 in (88 cm) for women (IIb-C) In capable patients, 30 minutes of moderate-intensity exercise most days. For those with disability, supervised exercise regimen (IIB-C)
Note: Adapted from the 2006 Guidelines for Prevention of Stroke in Patients with Ischemic Stroke or Transient Ischemic Attack from the Stroke Council of the American Heart Association.11 Class and levels of evidence according to American Heart Association Guidelines represented in parentheses as (class-level). ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; bid, twice a day; BMI, body mass index; BP, blood pressure; CAS, carotid artery angioplasty and stenting; CEA, carotid endarterectomy; HTN, hypertension; INR, International Normalized Ratio; LDL-C, low-density lipoprotein cholesterol; NCEP, National Cholesterol Education Project; qd, once a day; TIA, transient ischemic attack.
into improved health outcomes for all citizens. On the topic of translating efficacy into effectiveness, the current director of the National Institutes of Health, Dr. Zerhouni, states; “It is the responsibility of those of us involved in today’s biomedical research enterprise to translate the remarkable scientific innovations we are witnessing into health gains for the nation.”13,14 Large-scale quality improvement (QI) programs in cardiovascular disease have demonstrated the power of this patient-focused approach to increasing rates of adherence to national guidelines.15–17 These successful QI programs have helped provide the rationale for a more organized and sustained approach to stroke QI efforts.
Preliminary Experience with Stroke QI The original criteria for stroke QI organizations into operational definitions of performance measures.6,18 Several small-scale stroke QI programs have demonstrated significant improvements in treatment rates and health outcomes in cerebrovascular disease when evidence-based guidelines are applied in-hospital.19 –22 Because much of the science around prevention addresses ischemic stroke and transient ischemic attack (approximately 85% of all stroke),23 QI programs largely target the ischemic stroke subtype; however, some programs have also included interventions that address hemorrhagic stroke. These programs have
S252 American Journal of Preventive Medicine, Volume 31, Number 6S2
www.ajpm-online.net
shown that a coordinated stroke QI program can lead to substantial clinical and statistically significant improvements in: (1) use of intravenous (IV) tissue-type plasminogen activator; (2) early dysphagia screening, antithrombotics, and deep vein thrombosis prevention; and (3) discharge interventions of smoking-cessation counseling, antithrombotic agents in general with anticoagulation for atrial fibrillation, and lipid-lowering agents for dyslipidemia. In addition, there have been several major efforts by national healthcare organizations to define performance measures for high-quality stroke care. These include the Centers for Disease Control and Prevention’s Paul Coverdell National Acute Stroke Registry (www.cdc.gov/cvh/stroke_registry.htm); the Joint Commission’s Primary Stroke Center certification program (www.jcaho.org/dscc/psc/); the Center for Medicare and Medicaid Services, the Medicare Quality Monitoring System reports (www.cms. hhs.gov/QualityInitiativesGenInfo/); the American Heart Association’s “Get with the Guidelines–Stroke” program (www.strokeassociation.org); and the American Academy of Neurology’s Stroke Performance Improvement Network (www.aan.com/professionals/ patient/spin.cfm). Of these, the current largest experience is within the Get with the Guidelines–Stroke program, with over 130,000 patient admissions captured as of October 2005 (Tammy Gregory, American Heart Association, personal communication, December 1, 2005). Experience from these various QI programs has identified some important challenges: To date, these programs are largely unfunded, requiring significant investment of resources at the local hospital level. They also rely on data that must be collected prospectively or manually extracted from the medical record. In addition, abstraction of extensive, overlapping data on stroke care and attempts to achieve complete case ascertainment can distract from the more complex task of improving care delivery, especially when the same human resources are expected to perform both functions. Most importantly, a universally accepted set of stroke performance measures and data-element definitions is still lacking. Finally, concerns about privacy and the definition of a class of “protected health information,” have placed new federal restrictions on the acquisition of patient information for the purposes of medical research. A conservative institutional interpretation of these regulations has had the unfortunate consequence of inhibiting both in-hospital and postdischarge QI activities. This has prevented collection of QI data, especially postdischarge health outcomes, even though QI activities are explicitly excluded from the purview of the new federal privacy rule.24 –26 December 2006
Tension Between Data-Collection Requirements for Disease Registries Versus QI-Based Systems Although the Centers for Disease Control and Prevention’s Paul Coverdell National Acute Stroke Registry is referred to as a registry, its main function is quality improvement, because its primary goals are to track the delivery of care to hospitalized stroke patients and to guide and monitor improvements in the quality of acute stroke care.27 The distinction between a pure disease registry versus a QI project that incorporates ongoing surveillance activity has important ramifications in terms of its data-collection requirements. For true disease registries, such as cancer or birth defects registries, much attention is appropriately given to collecting “research-grade” data with very high rates of case ascertainment, inter-rater reliability, and accuracy. A large proportion of resources (both time and money) are therefore allocated toward data quality assurance activities—such as training, development of detailed protocols, and ongoing data checks in order to ensure that the collected data are of the highest quality. In addition, these registries’ activities are often supported at the hospital level as an institutional function or by reimbursement on a per-case basis by federal or state agencies. By contrast, QI-based activities do not usually have a continuous source of funding, and so the issue of sustainability is paramount. Given the high costs of data collection, particularly for data that must be collected in real time or by the review of medial records (rather than extracted from administrative sources), there is frequently a need to create a more focused data-collection strategy compared to that used by pure disease registries. Data collection within a QI project must be aligned with the goal of creating a measurable and sustainable change in performance, and, therefore, only those data elements necessary to make such measurements should be collected. Without such a change in emphasis, the costs of data collection can consume an excessive amount of the available resources and negatively affect the primary goal of the QI project, which is to improve care through sustainable changes in care processes and systems. However, despite this change of emphasis, it is absolutely critical that QI-based registries collect data in a consistent and reliable fashion, for there to be comparability both across time within a particular system (so that trends may be documented reliably) and across different systems or sites (so that variations among systems or sites can be identified accurately). Even when a QI project is appropriately focused on measuring only those variables that relate to its mission, data-collection activities still often require a high level of resources. For example, the complexity associated with defining the appropriate eligible population for a given stroke performance indicator often requires the collection of several variables in order to adequately Am J Prev Med 2006;31(6S2)
S253
define the denominator and the numerator. The accurate and reliable construction of performance indicators, therefore, requires development of detailed definitions for data-element and standardized data-collection methods. In turn, these requirements dictate that QI projects provide for adequate staff training at each site and that they demonstrate adequate data reliability and validity early in their development, which is then maintained throughout the course of the project. This also underscores the need for competing agencies to work together to agree on common standards for data elements and performance indicators, so that data-collection requirements are kept to a minimum. National organizations such as the National Quality Forum are ideal bodies to vet these proposed definitions and generate consensus among the various sponsors. Once initial data reliability and validity have been demonstrated (ideally through a standardized methodologic approach) and progress is made toward measurable improvements in care, then the focus of the QI efforts should shift to system change and sustainability. Sustainability can be met by performing ongoing data collection with periodic data audits designed to ensure the maintenance of data accuracy and completeness. Alternatively, data-assurance systems can be sustained by collecting in-depth data periodically through random sampling of the pool of eligible cases.28 The challenge to this sampling approach is to make sure that there is no selection bias that could result in cases with better outcomes being more likely to be included; and to avoid the potential for a Hawthorne effect, which might occur when heightened awareness because of the presence of measurement activities produces a higher quality of care that is not sustained between measurement periods.
Building Infrastructure to Support Data-Collection Activities The adoption of electronic medical records (EMR) and computerized order entry has been recognized by many influential healthcare policy groups as a critical step in transforming the United States’ healthcare system and improving the safety and reliability of care,12 (e.g., Leapfrog Group, www.leapfroggroup.org). These electronic systems hold the promise of providing decision-support algorithms to increase adherence to evidence-based guidelines and avoid preventable medical errors, such as unfavorable medication interactions or inaccurate transcription of orders. These systems also move further toward the goal of integrating highquality data collection into the routine delivery of medical care so that the parallel activity of manual data reabstraction is no longer required.
The ability to collect data required for QI registries in an ongoing fashion through EMR systems will be an enormous step toward ensuring a registry’s long-term sustainability. Until recently, electronic capture of medical data has focused largely on the collection and analysis of administrative data sets that are of very limited utility in understanding the nature of inadequate care delivery, and of even less value in providing insights into how to fix the identified problems. To be useful, these electronic data-capture systems will have to meet the same data quality standards of the current manual data-collection methods in terms of providing clinically relevant data that have high validity and reliability. In addition, these systems will need to interface with larger external data systems for the purposes of uploading data to create aggregate national samples while at the same time providing local practitioners with readily accessible feedback on performance. Although the challenges of developing a fully integrated and functional electronic-based data-collection system are clearly massive, they are nevertheless achievable given sufficient financial and intellectual investment, as reflected in the high priority placed on EMR creation and adoption by the Institute of Medicine report and the new Presidential Consolidated Health Informatics initiative (www.whitehouse.gov/omb/egov/c-3-6-chi. html).
Calculating Costs and Benefits of Stroke QI Activities The costs associated with data collection and quality improvement are difficult to quantify. The huge burden that stroke places on the U.S. and world economy, however, suggests that there are significant opportunities for financial savings through improved prevention. There are no published data on the cost and benefits of stroke QI activities. Development of a model of costs associated with these activities needs to take into account both the direct and indirect costs associated with these activities. These include costs for data collection and analysis (data abstraction based on the average abstraction time per chart, data validation, data cleaning, data entry, data analysis, report generation and distribution), as well as costs associated with provider education and feedback, system and process reengineering, active executive-level sponsorship, and reassignment of staff away from other activities that likely generate more profit for the healthcare institution. There can be significant cost and data quality differences when comparing approaches to QI implementation, such as the more expensive but more reliable method of concurrent case ascertainment and data collection versus a retrospective International Classification of Diseases, ninth edition-based model. Higher quality care in the acute stroke setting might lead to measurable cost savings. For example, if QI
S254 American Journal of Preventive Medicine, Volume 31, Number 6S2
www.ajpm-online.net
efforts increased use of IV tissue-type plasminogen activator for ischemic stroke, which has been shown to generate a cost savings of $4000 per treated patient,29 these savings could offset the direct costs of the QI program. QI programs should also help increase the efficiency within the healthcare system by promoting the use of more cost-effective preventive strategies, such as antithrombotic and antihypertensive therapy. In addition, hospital and provider involvement in QI activities should result in some direct increases in reimbursement. For example, the increased reimbursement under the newly released DiagnosisRelated Group 559 can offset costs associated with QI that increases the rates of thrombolytic administration in eligible patients. Adoption by federal, state, and third-party payers of pay-for-performance or pay-for-participation initiatives is critical to sustained QI. Increased reimbursement for evidence-based interventions in acute stroke care and secondary prevention has perhaps the greatest potential to improve patient outcomes by aligning long-term financial incentives and good clinical practice through structured stroke QI activities.30
The Need for Long-Term Health Outcomes and the Impact of Privacy Regulations Ultimately, any successful QI stroke registry will need to show improvement not just in surrogate measures, such as rates of guideline adherence by providers at discharge, but in actual patient health outcomes. This is critical because long-term stroke recurrence rates range from 4% to 14% annually, and many evidencebased interventions are designed to reduce this risk.31,32 Because the posthospitalization period is an important time for control of risk factors to minimize these recurrent vascular events, strategies that maximize patient adherence with risk reduction need to be identified and implemented. Modifiable risk factors for stroke include hypertension (the highest attributable stroke risk), smoking, excessive alcohol consumption, obesity, cardiac disease, and hyperlipidemia. Unfortunately, the relationship between risk factors and outcomes remains incompletely understood, particularly in populations that are disproportionately affected by stroke. More information is needed on how to translate the successful initiation of interventions in the hospital into successful long-term adherence to these interventions in the community. This is particularly important if disparities in healthcare outcomes seen among ethnic, racial, and socioeconomic groups are to be eliminated, which is why actual health outcomes, rather than just surrogate outcomes, are still needed. A national stroke QI registry must have the ability to determine the impact of implementing primary and secondary in-hospital prevention strategies in December 2006
order to focus resources in areas that need them the most. Feedback on the effectiveness of public health, community, and hospital- or physician-practice– based stroke QI initiatives require representative samples of patient-level data that will by necessity contain some protected health information. But collecting postdischarge data about stroke patients in a large disease-specific registry is challenging, particularly when new privacy rules and stricter local institutional review board policies require patient written informed consent with increasing frequency. The Canadian Stroke Network, a prospective registry based at 20 major stroke centers in Canada, recently published their experience with the impact of new consent rules on patient participation in their registry.25 They reported that they received the required consent for only 50% of potentially eligible patients, and that this created a consent-election bias such that the patients in the registry were no longer representative of the stroke patients in their target population. Privacy rules had a similar impact on enrollment in an acute coronary syndrome registry in Michigan, where consent for follow-up declined from 96.4% in the pre-Health Insurance Portability and Accountability Act of 1996 (HIPAA) period to only 34.0% in the post-HIPAA period (p⬍0.01); those who were included in the post-HIPAA period were older, more likely to be married, and had higher survival rates. Incremental costs associated with complying with the HIPAA Privacy Rule were substantial; for the first year they were $8704.50, and $4558.50 annually thereafter.26 Thus, the HIPAA Privacy Rule has a significant negative impact on the feasibility and sustainability of all QI-based registries. This rule decreases the number of patients available for outcomes research, introduces selection bias, and increases costs, all of which clearly limit the utility of the data for improving the monitoring and planning of healthcare delivery. Clarification of prior legislation, or the creation of new legislation, or regulations on patient privacy and policies permitting waivers of informed consent for minimal-risk observational research is clearly needed, so that the original goals of a national stroke registry can be achieved.
Need for Collaboration To design a sustainable national registry for stroke quality improvement, there is a need for collaboration among key partners in medical research and policy, including patient advocates, healthcare providers, healthcare systems, QI organizations, health payers, and state and federal health agencies. First and foremost, the scientific basis of the evidence must be solidly established and agreed on so that there are no controversies regarding the appropriateness of recommended interventions. These recommendations must be in the form of established patient-care guidelines that are well Am J Prev Med 2006;31(6S2)
S255
Table 2. Characteristics of an ideal hybridization of quality improvement projects and disease registries Focus or % effort
QI project
Disease registry
Ideal hybrid “QI registry”
Time spent on data collection activities Predominant data sources
Moderate
High
Moderate
Paper or electronic medical record (EMR)
Administrative data
Type of case ascertainment
Concurrent clinical identification
Retrospective case identification
Type of abstraction
Concurrent case review
Retrospective case review
Data quality assurance Completeness of case ascertainment Quality improvement Available funding support Type of support
Minimal Minimal
High High
High Minimal Nonprofit or federal
None or minimal High State, federal, or regulatory
Maintaining data quality
Ongoing abstractor training
Type of healthcare change
Provider/system change to improve patient outcomes
Rigorous data cleaning and ongoing data audits Provider/system change to improve data quality
EMR that supports QI and clinical research Concurrent case ascertainment with retrospective backfill Concurrent case review with retrospective data capture as needed Moderate Moderate with spread over time High High Multipayer pay for performance Random sampling or periodic audits System change to simplify data collection and improve patient outcomes
QI, quality improvement.
publicized. The medical research community needs to provide up-to-date, evidence-based information and keep guidelines current, based on emerging data from controlled clinical trials. The medical research community also needs to become much more active in the process of translating guidelines into performance measures—that is, measurable patient-care objectives that are linked to specific interventions that can be used to improve the process of care and patient health outcomes. QI organizations must provide expertise in the crafting of performance measures and work to foster common definitions of key performance variable and indicators. This should occur through educating the healthcare community about the scientific methods of quality improvement and emphasizing collaboration among competing organizations in diseasespecific areas. Just as physician leaders have been trained and recruited to become physician managers and hospital executives, so too must physician leaders embrace QI methodology and become champions for system change at the hospital and national levels. In addition, these efforts cannot happen in isolation. Ideally, they should be integrated with state and federal health agencies to bring uniformity and comparability to the process. Initially, this could take the form of pilot grants to state departments of public health (such as
has occurred in the Paul Coverdell National Acute Stroke Registry program), but eventually this should be administered as a national-level, stroke-specific disease registry program under uniform governmental or nonprofit administration. This program must ultimately be recognized by the major health-payer organizations, so that stroke performance will move from voluntary measurement to incentive-based adherence, with the ultimate goal of the public reporting of verifiable data. Hand-in-hand with this, better performance should be rewarded with peer recognition, stroke center designation/certification with routing of selected patients to higher performing sites, and higher rates of reimbursement for better performance. In summary, the design and implementation of a sustainable national registry for stroke quality improvement requires: (1) the endorsement of a single set of data definitions and relevant performance measures, (2) the provision of adequate funding for data-collection and QI activities, (3) the improvement of electronic capture of clinical data for QI purposes, (4) the removal of barriers to accessing critical patient information both in the hospital and after discharge, and (5) the creation of a vibrant culture of quality in which innovative solutions can be fostered to improve the health outcomes of patients with cerebrovascular disease.
S256 American Journal of Preventive Medicine, Volume 31, Number 6S2
www.ajpm-online.net
Funding for Drs. Schwamm (U50/CCU120238 –1), Frankel (U50/CCU420275– 01), and Reeves (U50/CCU520272– 01) was provided under Paul Coverdell National Acute Stroke Registry Pilot Prototype Grant cooperative agreements from the Centers for Disease Control and Prevention (CDC). No financial conflict of interest was reported by the authors of this paper.
References 1. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Bethesda MD: U.S. National Heart, Lung, and Blood Institute, National Institutes of Health, 2003. 2. The Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Bethesda MD: U.S. National Heart, Lung, and Blood Institute, National Institutes of Health; May 2001; Publication No. 01-3670. 3. Smith SC Jr, Blair SN, Bonow RO, et al. AHA/ACC Scientific Statement: AHA/ACC guidelines for preventing heart attack and death in patients with atherosclerotic cardiovascular disease: 2001 update: a statement for healthcare professionals from the American Heart Association and the American College of Cardiology. Circulation 2001;104:1577–9. 4. Adams HP Jr, Adams RJ, Brott T, et al. Guidelines for the early management of patients with ischemic stroke: a scientific statement from the Stroke Council of the American Stroke Association. Stroke 2003;34:1056 – 83. 5. Jencks SF, Cuerdon T, Burwen DR, et al. Quality of medical care delivered to Medicare beneficiaries: a profile at state and national levels. JAMA 2000;284:1670 – 6. 6. Holloway RG, Benesch C, Rush SR. Stroke prevention: narrowing the evidence-practice gap. Neurology 2000;54:1899 –906. 7. Rogers WJ, Canto JG, Lambrew CT, et al. Temporal trends in the treatment of over 1.5 million patients with myocardial infarction in the U.S. from 1990 through 1999: the National Registry of Myocardial Infarction 1, 2, and 3. J Am Coll Cardiol 2000;36:2056 – 63. 8. Reeves MJ, Arora S, Broderick JP, and the Paul Coverdell Prototype Registries Writing Group. Acute stroke care in the U.S.: results from 4 pilot prototypes of the Paul Coverdell National Acute Stroke Registry. Stroke 2005;36:1232– 40. 9. Quilliam BJ, Lapane KL, Eaton CB, Mor V. Effect of antiplatelet and anticoagulant agents on risk of hospitalization for bleeding among a population of elderly nursing home stroke survivors. Stroke 2001;32:2299 –304. 10. Hachinski V. Stroke: the next 30 years. Stroke 2002;33:1– 4. 11. Sacco RL, Adams R, Albers G, et al. American Heart Association; American Stroke Association Council on Stroke; Council on Cardiovascular Radiology and Intervention; American Academy of Neurology. Guidelines for prevention of stroke in patients with ischemic stroke or transient ischemic attack: a statement for healthcare professionals from the American Heart Association/American Stroke Association Council on Stroke. Stroke 2006; 37:577– 617 and Circulation 2006 Mar 14;113:e409 – 49. 12. Committee on Quality of Health Care in America, Institute of Medicine. Crossing the quality chasm: a new health system for the 21st century. Washington DC: National Academies Press, 2001. 13. Zerhouni E. The NIH roadmap. Science 2003;302:63–72. 14. Zerhouni EA. Translational and clinical science—time for a new vision. N Engl J Med 2005;353:1621–3.
December 2006
15. Eagle KA, Montoye CK, Riba AL, et al. American College of Cardiology’s Guidelines Applied in Practice (GAP) Projects in Michigan; American College of Cardiology Foundation (Bethesda [MD]) Guidelines Applied in Practice Steering committee. Guideline-based standardized care is associated with substantially lower mortality in Medicare patients with acute myocardial infarction: the American College of Cardiology’s Guidelines Applied in Practice (GAP) Projects in Michigan. J Am Coll Cardiol 2005;46:1242– 8. 16. Bhatt DL, Roe MT, Peterson ED, et al. CRUSADE Investigators. Utilization of early invasive management strategies for high-risk patients with non–STsegment elevation acute coronary syndromes: results from the CRUSADE Quality Improvement Initiative. JAMA 2004;292:2096 –104. 17. LaBresh KA, Ellrodt AG, Gliklich R, et al. Get with the guidelines for cardiovascular secondary prevention: pilot results. Arch Intern Med 2004;164(2):203–9. 18. Holloway RG, Vickrey BG, Benesch C, Hinchey JA, Bieber J. National Expert Stroke Panel. Development of performance measures for acute ischemic stroke. Stroke 2001;32:2058 –74. 19. Hinchey JA, Shephard T, Furie K, Smith D, Wang D, Tonn S, Stroke Practice Improvement Network Investigators. Formal dysphagia screening protocols prevent pneumonia. Stroke 2005;36:1972– 6. 20. Schwamm LH, LaBresh KA, Albright DG, et al. “Get with the Guidelines— Stroke” improves acute intervention in patients hospitalized with ischemic stroke or TIA. [abstract]. Stroke 2005;36:436. 21. LaBresh KA, Schwamm LH, Albright DG, et al. Does Get with the Guidelines improve secondary prevention in patients hospitalized with ischemic stroke or TIA? [abstract]. Stroke 2005;36:454. 22. Stoeckle-Roberts S, Mullard A, Choate L, Wehner S, Reeves MJ, for The MASCOTS Writing Group. Closing the gap between evidence-based acute stroke care and hospital-based practice using a collaborative approach. 6th Scientific Forum on Quality of Care and Outcomes Research in Cardiovascular Disease and Stroke. Washington DC, May 2005 [abstract]. Circulation 2005;111(20):E331. 23. American Heart Association. Heart disease and stroke statistics—2005 Update. Dallas TX: American Heart Association, 2004. 24. The Health Insurance Portability and Accountability Act of 1996 (HIPAA), Public Law 104 –191, enacted on August 21, 1996. 25. Tu JV, Willison DJ, Silver FL, et al. Investigators in the Registry of the Canadian Stroke Network. Impracticability of informed consent in the Registry of the Canadian Stroke Network. N Engl J Med 2004;350:1414 –21. 26. Armstrong D, Kline-Rogers E, Jani SM, et al. Potential impact of the HIPAA privacy rule on data collection in a registry of patients with acute coronary syndrome. Arch Intern Med 2005;165:1125–9. 27. Wattigney WA, Croft JB, Mensah GA, et al. Establishing data elements for the Paul Coverdell National Acute Stroke Registry: part 1: proceedings of an expert panel. Stroke 2003;34:151– 6. 28. Reeves MJ, Page C, Zhang H. Design and sample size recommendations for state-based registries participating in the Paul Coverdell National Acute Stroke Registry. Poster presentation. 30th International Stroke Conference 2005, New Orleans LA. February 2005 [abstract]. Stroke 2005;36:479. 29. Fagan SC, Morgenstern LB, Petitta A, et al. Cost-effectiveness of tissue plasminogen activator for acute ischemic stroke. NINDS rt-PA Stroke Study Group. Neurology 1998;50:883–90. 30. Berthiaume JT, Tyler PA, Ng-Osorio J, LaBresh KA. Aligning financial incentives with “Get With The Guidelines” to improve cardiovascular care. Am J Manag Care 2004;10:501– 4. 31. Sacco RL, Wolf PA, Kannel WB, McNamara PM. Survival and recurrence following stroke: the Framingham Study. Stroke 1982;13:290 –5. 32. Petty GW, Brown RD Jr, Whisnant JP, Sicks JD, O’Fallon WM, Wiebers DO. Survival and recurrence after first cerebral infarction: a population-based study in Rochester, Minnesota, 1975 through 1989. Neurology 1998; 50:208 –16.
Am J Prev Med 2006;31(6S2)
S257
Several studies have shown wide variations in the delivery of hospital-based care to patients with acute ischemic stroke. The findings of these studies suggest that recommendations drawn from published evidence-based or consensus-based guidelines are implemented inconsistently. Although rates of adherence to stroke quality indicators can be increased through the use of targeted quality improvement (QI) efforts, stroke QI programs are still in their infancy. Current stroke QI programs are often highly variable and poorly coordinated, rely on differing definitions of key data elements and/or indicators, and are challenging to implement and sustain in resource-constrained healthcare environments. Key barriers to long-term success of these programs include inadequate funding at the local and national level, lack of infrastructure to support electronic data capture for QI as part of the process of patient care, lack of a single clearinghouse for uniform data definitions and performance indicator descriptions, competing survey instruments to monitor hospitalized stroke care, and constraints on inpatient and post-discharge data collection imposed by the new Federal Health Insurance Portability and Accountability Act Privacy Rule. In addition, the competing needs of registry activities (e.g., complete case ascertainment) versus QI efforts (e.g., incremental tests of change) must be balanced. Potential solutions include: (1) financial incentives to healthcare providers and institutions for participation in QI initiatives; (2) financial incentives to healthcare providers and institutions for measurable improvements in care; (3) mandatory data reporting on key measures of stroke care; and (4) promotion of active and sustainable collaborations among key stakeholders including healthcare providers (e.g., physicians, nurses), healthcare organizations (e.g., hospitals, physicians’ groups), quality improvement organizations, health payers and insurers, public health departments, and state and federal health agencies to create a single national stroke registry for stroke QI. (Am J Prev Med 2006;31(6S2):S251–S257) © 2006 American Journal of Preventive Medicine
Rationale for Stroke Quality Improvement
T
here is abundant scientific evidence that supports consensus- and evidence-based guidelines for in-hospital and postdischarge care in cardiovascular and cerebrovascular disease management.1– 4 Despite ready access to published guidelines and widespread health-provider awareness of the recommended interventions, guidelines are poorly implemented in both in-hospital and postdischarge environments.5– 8 Some studies suggest that providers overestimate both likelihood and severity of complications of therapy in patients of advanced age or frail health when compared to the risk of recurrent stroke.9 Risk factors are adequately controlled in only 25% of patients, despite the From the Department of Neurology, Massachusetts General Hospital, Harvard Medical School (Schwamm), Boston, Massachusetts; Department of Epidemiology, Michigan State University (Reeves), East Lansing, Michigan; Department of Neurology, Grady Hospital, Emory University (Frankel), Atlanta, Georgia Address correspondence and reprint requests to: Lee H. Schwamm, MD, Department of Neurology VBK-915, Massachusetts General Hospital, 55 Fruit Street, Boston MA 02114. E-mail: [email protected].
fact that up to 75% of cardiac and cerebrovascular disease is preventable with successful risk-factor management.10 Table 1 lists the common modifiable and nonmodifiable risk factors for ischemic stroke and transient ischemic attack. The effort to close the gap between what is known and what is practiced has been the focus of several major studies and initiatives in recent years. The influential Institute of Medicine report Crossing the Quality Chasm12 offered a set of performance expectations for the 21st century healthcare system, a set of new rules to guide patient-to-clinician relationships, an organizing framework to better align incentives inherent in payment and accountability with improvements in quality, and key steps to promote evidence-based practice and strengthen clinical information systems. In another initiative, the National Institutes of Health Roadmap attempts to reorganize the federal funding of medical research to address some of the barriers inherent in translating not only the results of basic research but also the results of clinical trials
Am J Prev Med 2006;31(6S2) © 2006 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/06/$–see front matter S251 doi:10.1016/j.amepre.2006.08.013
Table 1. Common risk factors for ischemic stroke and TIA and recommendations for secondary prevention interventions in adults Risk factor Age Race/ethnicity Gender Heredity Hypertension Diabetes
Atrial fibrillation (AF) Noncardioembolic ischemic stroke or TIA Coronary heart disease Dyslipidemia
Cervical internal carotid artery stenosis
Cigarette smoking Alcohol use Obesity Physical inactivity
Recommendations for secondary stroke prevention in adult patients with ischemic stroke/TIA Nonmodifiable Nonmodifiable Nonmodifiable Nonmodifiable Antihypertensives after the hyperacute period (I-A) along with lifestyle modifications (IIb-C). Optimal drug regimen uncertain but diuretics and the combination of diuretics and an ACEI are effective (I-A) Rigorous control of dyslipidemia and HTN, with BP targets of 130/80 mm Hg (IIa-B). ACEIs and ARBs are first-choice medications (I-A). Glucose control is recommended to near normoglycemic levels to reduce microvascular complications (I-A) and possibly macrovascular complications (IIb-B). Hemoglobin A1c goal ⬍7% (IIa-B) For persistent or paroxysmal (intermittent) AF, use adjusted-dose warfarin (INR 2.5; range, 2–3) (I-A). For patients unable to take oral anticoagulants, aspirin 325 mg per day (I-A) Antiplatelet agents rather than oral anticoagulation to reduce risk of recurrent stroke and other cardiovascular events (I-A). Options for initial therapy (IIa-A) include aspirin (50–325 mg qd), aspirin plus extended-release dipyridamole (25/200 mg bid), clopidogrel (75 mg qd) Antiplatelet agents (I-A), lipid-altering agents per NCEP III (I-A) and management of comorbid risk factors for atherosclerosis as outlined Patients with elevated cholesterol, coronary heart disease, or evidence of an atherosclerotic etiology should be managed according to NCEP III (I-A). Statins are recommended with target LDL-C of ⬍100 mg/dL and ⬍70 mg/dL for the very high-risk (I-A). Reasonable to consider statins to reduce risk of vascular events in those without NCEP indication (IIa-B) For ipsilateral severe stenosis (70%–99%), CEA is recommended (I-A) and CAS may be considered (IIb-B) if (1) difficult to access surgically, (2) medical conditions exist that greatly increase risk for surgery, or (3) with radiation-induced stenosis or restenosis after CEA and (4) when performed by operators with morbidity and mortality rates of 4% to 6% (IIa-B). For ipsilateral moderate stenosis (50%–69%), CEA is recommended depending on age, gender, medical comorbidity, and severity of symptoms (I-A). For ipsilateral mild stenosis (⬍50%), there is no indication for CEA or CAS (III-A). Smoking cessation through use of counseling, nicotine replacement, or other medications (I-C) Heavy drinkers should eliminate or reduce (⬍2 drinks/day males, ⬍1 drink/day females) their consumption of alcohol (I-A) Consider weight reduction through increased activity and decreased caloric intake for all overweight patients to maintain goal BMI of 18.5 to 24.9 (calculated as kg/m2) or waist circumference ⬍40 in (102 cm) for men and ⬍35 in (88 cm) for women (IIb-C) In capable patients, 30 minutes of moderate-intensity exercise most days. For those with disability, supervised exercise regimen (IIB-C)
Note: Adapted from the 2006 Guidelines for Prevention of Stroke in Patients with Ischemic Stroke or Transient Ischemic Attack from the Stroke Council of the American Heart Association.11 Class and levels of evidence according to American Heart Association Guidelines represented in parentheses as (class-level). ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; bid, twice a day; BMI, body mass index; BP, blood pressure; CAS, carotid artery angioplasty and stenting; CEA, carotid endarterectomy; HTN, hypertension; INR, International Normalized Ratio; LDL-C, low-density lipoprotein cholesterol; NCEP, National Cholesterol Education Project; qd, once a day; TIA, transient ischemic attack.
into improved health outcomes for all citizens. On the topic of translating efficacy into effectiveness, the current director of the National Institutes of Health, Dr. Zerhouni, states; “It is the responsibility of those of us involved in today’s biomedical research enterprise to translate the remarkable scientific innovations we are witnessing into health gains for the nation.”13,14 Large-scale quality improvement (QI) programs in cardiovascular disease have demonstrated the power of this patient-focused approach to increasing rates of adherence to national guidelines.15–17 These successful QI programs have helped provide the rationale for a more organized and sustained approach to stroke QI efforts.
Preliminary Experience with Stroke QI The original criteria for stroke QI organizations into operational definitions of performance measures.6,18 Several small-scale stroke QI programs have demonstrated significant improvements in treatment rates and health outcomes in cerebrovascular disease when evidence-based guidelines are applied in-hospital.19 –22 Because much of the science around prevention addresses ischemic stroke and transient ischemic attack (approximately 85% of all stroke),23 QI programs largely target the ischemic stroke subtype; however, some programs have also included interventions that address hemorrhagic stroke. These programs have
S252 American Journal of Preventive Medicine, Volume 31, Number 6S2
www.ajpm-online.net
shown that a coordinated stroke QI program can lead to substantial clinical and statistically significant improvements in: (1) use of intravenous (IV) tissue-type plasminogen activator; (2) early dysphagia screening, antithrombotics, and deep vein thrombosis prevention; and (3) discharge interventions of smoking-cessation counseling, antithrombotic agents in general with anticoagulation for atrial fibrillation, and lipid-lowering agents for dyslipidemia. In addition, there have been several major efforts by national healthcare organizations to define performance measures for high-quality stroke care. These include the Centers for Disease Control and Prevention’s Paul Coverdell National Acute Stroke Registry (www.cdc.gov/cvh/stroke_registry.htm); the Joint Commission’s Primary Stroke Center certification program (www.jcaho.org/dscc/psc/); the Center for Medicare and Medicaid Services, the Medicare Quality Monitoring System reports (www.cms. hhs.gov/QualityInitiativesGenInfo/); the American Heart Association’s “Get with the Guidelines–Stroke” program (www.strokeassociation.org); and the American Academy of Neurology’s Stroke Performance Improvement Network (www.aan.com/professionals/ patient/spin.cfm). Of these, the current largest experience is within the Get with the Guidelines–Stroke program, with over 130,000 patient admissions captured as of October 2005 (Tammy Gregory, American Heart Association, personal communication, December 1, 2005). Experience from these various QI programs has identified some important challenges: To date, these programs are largely unfunded, requiring significant investment of resources at the local hospital level. They also rely on data that must be collected prospectively or manually extracted from the medical record. In addition, abstraction of extensive, overlapping data on stroke care and attempts to achieve complete case ascertainment can distract from the more complex task of improving care delivery, especially when the same human resources are expected to perform both functions. Most importantly, a universally accepted set of stroke performance measures and data-element definitions is still lacking. Finally, concerns about privacy and the definition of a class of “protected health information,” have placed new federal restrictions on the acquisition of patient information for the purposes of medical research. A conservative institutional interpretation of these regulations has had the unfortunate consequence of inhibiting both in-hospital and postdischarge QI activities. This has prevented collection of QI data, especially postdischarge health outcomes, even though QI activities are explicitly excluded from the purview of the new federal privacy rule.24 –26 December 2006
Tension Between Data-Collection Requirements for Disease Registries Versus QI-Based Systems Although the Centers for Disease Control and Prevention’s Paul Coverdell National Acute Stroke Registry is referred to as a registry, its main function is quality improvement, because its primary goals are to track the delivery of care to hospitalized stroke patients and to guide and monitor improvements in the quality of acute stroke care.27 The distinction between a pure disease registry versus a QI project that incorporates ongoing surveillance activity has important ramifications in terms of its data-collection requirements. For true disease registries, such as cancer or birth defects registries, much attention is appropriately given to collecting “research-grade” data with very high rates of case ascertainment, inter-rater reliability, and accuracy. A large proportion of resources (both time and money) are therefore allocated toward data quality assurance activities—such as training, development of detailed protocols, and ongoing data checks in order to ensure that the collected data are of the highest quality. In addition, these registries’ activities are often supported at the hospital level as an institutional function or by reimbursement on a per-case basis by federal or state agencies. By contrast, QI-based activities do not usually have a continuous source of funding, and so the issue of sustainability is paramount. Given the high costs of data collection, particularly for data that must be collected in real time or by the review of medial records (rather than extracted from administrative sources), there is frequently a need to create a more focused data-collection strategy compared to that used by pure disease registries. Data collection within a QI project must be aligned with the goal of creating a measurable and sustainable change in performance, and, therefore, only those data elements necessary to make such measurements should be collected. Without such a change in emphasis, the costs of data collection can consume an excessive amount of the available resources and negatively affect the primary goal of the QI project, which is to improve care through sustainable changes in care processes and systems. However, despite this change of emphasis, it is absolutely critical that QI-based registries collect data in a consistent and reliable fashion, for there to be comparability both across time within a particular system (so that trends may be documented reliably) and across different systems or sites (so that variations among systems or sites can be identified accurately). Even when a QI project is appropriately focused on measuring only those variables that relate to its mission, data-collection activities still often require a high level of resources. For example, the complexity associated with defining the appropriate eligible population for a given stroke performance indicator often requires the collection of several variables in order to adequately Am J Prev Med 2006;31(6S2)
S253
define the denominator and the numerator. The accurate and reliable construction of performance indicators, therefore, requires development of detailed definitions for data-element and standardized data-collection methods. In turn, these requirements dictate that QI projects provide for adequate staff training at each site and that they demonstrate adequate data reliability and validity early in their development, which is then maintained throughout the course of the project. This also underscores the need for competing agencies to work together to agree on common standards for data elements and performance indicators, so that data-collection requirements are kept to a minimum. National organizations such as the National Quality Forum are ideal bodies to vet these proposed definitions and generate consensus among the various sponsors. Once initial data reliability and validity have been demonstrated (ideally through a standardized methodologic approach) and progress is made toward measurable improvements in care, then the focus of the QI efforts should shift to system change and sustainability. Sustainability can be met by performing ongoing data collection with periodic data audits designed to ensure the maintenance of data accuracy and completeness. Alternatively, data-assurance systems can be sustained by collecting in-depth data periodically through random sampling of the pool of eligible cases.28 The challenge to this sampling approach is to make sure that there is no selection bias that could result in cases with better outcomes being more likely to be included; and to avoid the potential for a Hawthorne effect, which might occur when heightened awareness because of the presence of measurement activities produces a higher quality of care that is not sustained between measurement periods.
Building Infrastructure to Support Data-Collection Activities The adoption of electronic medical records (EMR) and computerized order entry has been recognized by many influential healthcare policy groups as a critical step in transforming the United States’ healthcare system and improving the safety and reliability of care,12 (e.g., Leapfrog Group, www.leapfroggroup.org). These electronic systems hold the promise of providing decision-support algorithms to increase adherence to evidence-based guidelines and avoid preventable medical errors, such as unfavorable medication interactions or inaccurate transcription of orders. These systems also move further toward the goal of integrating highquality data collection into the routine delivery of medical care so that the parallel activity of manual data reabstraction is no longer required.
The ability to collect data required for QI registries in an ongoing fashion through EMR systems will be an enormous step toward ensuring a registry’s long-term sustainability. Until recently, electronic capture of medical data has focused largely on the collection and analysis of administrative data sets that are of very limited utility in understanding the nature of inadequate care delivery, and of even less value in providing insights into how to fix the identified problems. To be useful, these electronic data-capture systems will have to meet the same data quality standards of the current manual data-collection methods in terms of providing clinically relevant data that have high validity and reliability. In addition, these systems will need to interface with larger external data systems for the purposes of uploading data to create aggregate national samples while at the same time providing local practitioners with readily accessible feedback on performance. Although the challenges of developing a fully integrated and functional electronic-based data-collection system are clearly massive, they are nevertheless achievable given sufficient financial and intellectual investment, as reflected in the high priority placed on EMR creation and adoption by the Institute of Medicine report and the new Presidential Consolidated Health Informatics initiative (www.whitehouse.gov/omb/egov/c-3-6-chi. html).
Calculating Costs and Benefits of Stroke QI Activities The costs associated with data collection and quality improvement are difficult to quantify. The huge burden that stroke places on the U.S. and world economy, however, suggests that there are significant opportunities for financial savings through improved prevention. There are no published data on the cost and benefits of stroke QI activities. Development of a model of costs associated with these activities needs to take into account both the direct and indirect costs associated with these activities. These include costs for data collection and analysis (data abstraction based on the average abstraction time per chart, data validation, data cleaning, data entry, data analysis, report generation and distribution), as well as costs associated with provider education and feedback, system and process reengineering, active executive-level sponsorship, and reassignment of staff away from other activities that likely generate more profit for the healthcare institution. There can be significant cost and data quality differences when comparing approaches to QI implementation, such as the more expensive but more reliable method of concurrent case ascertainment and data collection versus a retrospective International Classification of Diseases, ninth edition-based model. Higher quality care in the acute stroke setting might lead to measurable cost savings. For example, if QI
S254 American Journal of Preventive Medicine, Volume 31, Number 6S2
www.ajpm-online.net
efforts increased use of IV tissue-type plasminogen activator for ischemic stroke, which has been shown to generate a cost savings of $4000 per treated patient,29 these savings could offset the direct costs of the QI program. QI programs should also help increase the efficiency within the healthcare system by promoting the use of more cost-effective preventive strategies, such as antithrombotic and antihypertensive therapy. In addition, hospital and provider involvement in QI activities should result in some direct increases in reimbursement. For example, the increased reimbursement under the newly released DiagnosisRelated Group 559 can offset costs associated with QI that increases the rates of thrombolytic administration in eligible patients. Adoption by federal, state, and third-party payers of pay-for-performance or pay-for-participation initiatives is critical to sustained QI. Increased reimbursement for evidence-based interventions in acute stroke care and secondary prevention has perhaps the greatest potential to improve patient outcomes by aligning long-term financial incentives and good clinical practice through structured stroke QI activities.30
The Need for Long-Term Health Outcomes and the Impact of Privacy Regulations Ultimately, any successful QI stroke registry will need to show improvement not just in surrogate measures, such as rates of guideline adherence by providers at discharge, but in actual patient health outcomes. This is critical because long-term stroke recurrence rates range from 4% to 14% annually, and many evidencebased interventions are designed to reduce this risk.31,32 Because the posthospitalization period is an important time for control of risk factors to minimize these recurrent vascular events, strategies that maximize patient adherence with risk reduction need to be identified and implemented. Modifiable risk factors for stroke include hypertension (the highest attributable stroke risk), smoking, excessive alcohol consumption, obesity, cardiac disease, and hyperlipidemia. Unfortunately, the relationship between risk factors and outcomes remains incompletely understood, particularly in populations that are disproportionately affected by stroke. More information is needed on how to translate the successful initiation of interventions in the hospital into successful long-term adherence to these interventions in the community. This is particularly important if disparities in healthcare outcomes seen among ethnic, racial, and socioeconomic groups are to be eliminated, which is why actual health outcomes, rather than just surrogate outcomes, are still needed. A national stroke QI registry must have the ability to determine the impact of implementing primary and secondary in-hospital prevention strategies in December 2006
order to focus resources in areas that need them the most. Feedback on the effectiveness of public health, community, and hospital- or physician-practice– based stroke QI initiatives require representative samples of patient-level data that will by necessity contain some protected health information. But collecting postdischarge data about stroke patients in a large disease-specific registry is challenging, particularly when new privacy rules and stricter local institutional review board policies require patient written informed consent with increasing frequency. The Canadian Stroke Network, a prospective registry based at 20 major stroke centers in Canada, recently published their experience with the impact of new consent rules on patient participation in their registry.25 They reported that they received the required consent for only 50% of potentially eligible patients, and that this created a consent-election bias such that the patients in the registry were no longer representative of the stroke patients in their target population. Privacy rules had a similar impact on enrollment in an acute coronary syndrome registry in Michigan, where consent for follow-up declined from 96.4% in the pre-Health Insurance Portability and Accountability Act of 1996 (HIPAA) period to only 34.0% in the post-HIPAA period (p⬍0.01); those who were included in the post-HIPAA period were older, more likely to be married, and had higher survival rates. Incremental costs associated with complying with the HIPAA Privacy Rule were substantial; for the first year they were $8704.50, and $4558.50 annually thereafter.26 Thus, the HIPAA Privacy Rule has a significant negative impact on the feasibility and sustainability of all QI-based registries. This rule decreases the number of patients available for outcomes research, introduces selection bias, and increases costs, all of which clearly limit the utility of the data for improving the monitoring and planning of healthcare delivery. Clarification of prior legislation, or the creation of new legislation, or regulations on patient privacy and policies permitting waivers of informed consent for minimal-risk observational research is clearly needed, so that the original goals of a national stroke registry can be achieved.
Need for Collaboration To design a sustainable national registry for stroke quality improvement, there is a need for collaboration among key partners in medical research and policy, including patient advocates, healthcare providers, healthcare systems, QI organizations, health payers, and state and federal health agencies. First and foremost, the scientific basis of the evidence must be solidly established and agreed on so that there are no controversies regarding the appropriateness of recommended interventions. These recommendations must be in the form of established patient-care guidelines that are well Am J Prev Med 2006;31(6S2)
S255
Table 2. Characteristics of an ideal hybridization of quality improvement projects and disease registries Focus or % effort
QI project
Disease registry
Ideal hybrid “QI registry”
Time spent on data collection activities Predominant data sources
Moderate
High
Moderate
Paper or electronic medical record (EMR)
Administrative data
Type of case ascertainment
Concurrent clinical identification
Retrospective case identification
Type of abstraction
Concurrent case review
Retrospective case review
Data quality assurance Completeness of case ascertainment Quality improvement Available funding support Type of support
Minimal Minimal
High High
High Minimal Nonprofit or federal
None or minimal High State, federal, or regulatory
Maintaining data quality
Ongoing abstractor training
Type of healthcare change
Provider/system change to improve patient outcomes
Rigorous data cleaning and ongoing data audits Provider/system change to improve data quality
EMR that supports QI and clinical research Concurrent case ascertainment with retrospective backfill Concurrent case review with retrospective data capture as needed Moderate Moderate with spread over time High High Multipayer pay for performance Random sampling or periodic audits System change to simplify data collection and improve patient outcomes
QI, quality improvement.
publicized. The medical research community needs to provide up-to-date, evidence-based information and keep guidelines current, based on emerging data from controlled clinical trials. The medical research community also needs to become much more active in the process of translating guidelines into performance measures—that is, measurable patient-care objectives that are linked to specific interventions that can be used to improve the process of care and patient health outcomes. QI organizations must provide expertise in the crafting of performance measures and work to foster common definitions of key performance variable and indicators. This should occur through educating the healthcare community about the scientific methods of quality improvement and emphasizing collaboration among competing organizations in diseasespecific areas. Just as physician leaders have been trained and recruited to become physician managers and hospital executives, so too must physician leaders embrace QI methodology and become champions for system change at the hospital and national levels. In addition, these efforts cannot happen in isolation. Ideally, they should be integrated with state and federal health agencies to bring uniformity and comparability to the process. Initially, this could take the form of pilot grants to state departments of public health (such as
has occurred in the Paul Coverdell National Acute Stroke Registry program), but eventually this should be administered as a national-level, stroke-specific disease registry program under uniform governmental or nonprofit administration. This program must ultimately be recognized by the major health-payer organizations, so that stroke performance will move from voluntary measurement to incentive-based adherence, with the ultimate goal of the public reporting of verifiable data. Hand-in-hand with this, better performance should be rewarded with peer recognition, stroke center designation/certification with routing of selected patients to higher performing sites, and higher rates of reimbursement for better performance. In summary, the design and implementation of a sustainable national registry for stroke quality improvement requires: (1) the endorsement of a single set of data definitions and relevant performance measures, (2) the provision of adequate funding for data-collection and QI activities, (3) the improvement of electronic capture of clinical data for QI purposes, (4) the removal of barriers to accessing critical patient information both in the hospital and after discharge, and (5) the creation of a vibrant culture of quality in which innovative solutions can be fostered to improve the health outcomes of patients with cerebrovascular disease.
S256 American Journal of Preventive Medicine, Volume 31, Number 6S2
www.ajpm-online.net
Funding for Drs. Schwamm (U50/CCU120238 –1), Frankel (U50/CCU420275– 01), and Reeves (U50/CCU520272– 01) was provided under Paul Coverdell National Acute Stroke Registry Pilot Prototype Grant cooperative agreements from the Centers for Disease Control and Prevention (CDC). No financial conflict of interest was reported by the authors of this paper.
References 1. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Bethesda MD: U.S. National Heart, Lung, and Blood Institute, National Institutes of Health, 2003. 2. The Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Bethesda MD: U.S. National Heart, Lung, and Blood Institute, National Institutes of Health; May 2001; Publication No. 01-3670. 3. Smith SC Jr, Blair SN, Bonow RO, et al. AHA/ACC Scientific Statement: AHA/ACC guidelines for preventing heart attack and death in patients with atherosclerotic cardiovascular disease: 2001 update: a statement for healthcare professionals from the American Heart Association and the American College of Cardiology. Circulation 2001;104:1577–9. 4. Adams HP Jr, Adams RJ, Brott T, et al. Guidelines for the early management of patients with ischemic stroke: a scientific statement from the Stroke Council of the American Stroke Association. Stroke 2003;34:1056 – 83. 5. Jencks SF, Cuerdon T, Burwen DR, et al. Quality of medical care delivered to Medicare beneficiaries: a profile at state and national levels. JAMA 2000;284:1670 – 6. 6. Holloway RG, Benesch C, Rush SR. Stroke prevention: narrowing the evidence-practice gap. Neurology 2000;54:1899 –906. 7. Rogers WJ, Canto JG, Lambrew CT, et al. Temporal trends in the treatment of over 1.5 million patients with myocardial infarction in the U.S. from 1990 through 1999: the National Registry of Myocardial Infarction 1, 2, and 3. J Am Coll Cardiol 2000;36:2056 – 63. 8. Reeves MJ, Arora S, Broderick JP, and the Paul Coverdell Prototype Registries Writing Group. Acute stroke care in the U.S.: results from 4 pilot prototypes of the Paul Coverdell National Acute Stroke Registry. Stroke 2005;36:1232– 40. 9. Quilliam BJ, Lapane KL, Eaton CB, Mor V. Effect of antiplatelet and anticoagulant agents on risk of hospitalization for bleeding among a population of elderly nursing home stroke survivors. Stroke 2001;32:2299 –304. 10. Hachinski V. Stroke: the next 30 years. Stroke 2002;33:1– 4. 11. Sacco RL, Adams R, Albers G, et al. American Heart Association; American Stroke Association Council on Stroke; Council on Cardiovascular Radiology and Intervention; American Academy of Neurology. Guidelines for prevention of stroke in patients with ischemic stroke or transient ischemic attack: a statement for healthcare professionals from the American Heart Association/American Stroke Association Council on Stroke. Stroke 2006; 37:577– 617 and Circulation 2006 Mar 14;113:e409 – 49. 12. Committee on Quality of Health Care in America, Institute of Medicine. Crossing the quality chasm: a new health system for the 21st century. Washington DC: National Academies Press, 2001. 13. Zerhouni E. The NIH roadmap. Science 2003;302:63–72. 14. Zerhouni EA. Translational and clinical science—time for a new vision. N Engl J Med 2005;353:1621–3.
December 2006
15. Eagle KA, Montoye CK, Riba AL, et al. American College of Cardiology’s Guidelines Applied in Practice (GAP) Projects in Michigan; American College of Cardiology Foundation (Bethesda [MD]) Guidelines Applied in Practice Steering committee. Guideline-based standardized care is associated with substantially lower mortality in Medicare patients with acute myocardial infarction: the American College of Cardiology’s Guidelines Applied in Practice (GAP) Projects in Michigan. J Am Coll Cardiol 2005;46:1242– 8. 16. Bhatt DL, Roe MT, Peterson ED, et al. CRUSADE Investigators. Utilization of early invasive management strategies for high-risk patients with non–STsegment elevation acute coronary syndromes: results from the CRUSADE Quality Improvement Initiative. JAMA 2004;292:2096 –104. 17. LaBresh KA, Ellrodt AG, Gliklich R, et al. Get with the guidelines for cardiovascular secondary prevention: pilot results. Arch Intern Med 2004;164(2):203–9. 18. Holloway RG, Vickrey BG, Benesch C, Hinchey JA, Bieber J. National Expert Stroke Panel. Development of performance measures for acute ischemic stroke. Stroke 2001;32:2058 –74. 19. Hinchey JA, Shephard T, Furie K, Smith D, Wang D, Tonn S, Stroke Practice Improvement Network Investigators. Formal dysphagia screening protocols prevent pneumonia. Stroke 2005;36:1972– 6. 20. Schwamm LH, LaBresh KA, Albright DG, et al. “Get with the Guidelines— Stroke” improves acute intervention in patients hospitalized with ischemic stroke or TIA. [abstract]. Stroke 2005;36:436. 21. LaBresh KA, Schwamm LH, Albright DG, et al. Does Get with the Guidelines improve secondary prevention in patients hospitalized with ischemic stroke or TIA? [abstract]. Stroke 2005;36:454. 22. Stoeckle-Roberts S, Mullard A, Choate L, Wehner S, Reeves MJ, for The MASCOTS Writing Group. Closing the gap between evidence-based acute stroke care and hospital-based practice using a collaborative approach. 6th Scientific Forum on Quality of Care and Outcomes Research in Cardiovascular Disease and Stroke. Washington DC, May 2005 [abstract]. Circulation 2005;111(20):E331. 23. American Heart Association. Heart disease and stroke statistics—2005 Update. Dallas TX: American Heart Association, 2004. 24. The Health Insurance Portability and Accountability Act of 1996 (HIPAA), Public Law 104 –191, enacted on August 21, 1996. 25. Tu JV, Willison DJ, Silver FL, et al. Investigators in the Registry of the Canadian Stroke Network. Impracticability of informed consent in the Registry of the Canadian Stroke Network. N Engl J Med 2004;350:1414 –21. 26. Armstrong D, Kline-Rogers E, Jani SM, et al. Potential impact of the HIPAA privacy rule on data collection in a registry of patients with acute coronary syndrome. Arch Intern Med 2005;165:1125–9. 27. Wattigney WA, Croft JB, Mensah GA, et al. Establishing data elements for the Paul Coverdell National Acute Stroke Registry: part 1: proceedings of an expert panel. Stroke 2003;34:151– 6. 28. Reeves MJ, Page C, Zhang H. Design and sample size recommendations for state-based registries participating in the Paul Coverdell National Acute Stroke Registry. Poster presentation. 30th International Stroke Conference 2005, New Orleans LA. February 2005 [abstract]. Stroke 2005;36:479. 29. Fagan SC, Morgenstern LB, Petitta A, et al. Cost-effectiveness of tissue plasminogen activator for acute ischemic stroke. NINDS rt-PA Stroke Study Group. Neurology 1998;50:883–90. 30. Berthiaume JT, Tyler PA, Ng-Osorio J, LaBresh KA. Aligning financial incentives with “Get With The Guidelines” to improve cardiovascular care. Am J Manag Care 2004;10:501– 4. 31. Sacco RL, Wolf PA, Kannel WB, McNamara PM. Survival and recurrence following stroke: the Framingham Study. Stroke 1982;13:290 –5. 32. Petty GW, Brown RD Jr, Whisnant JP, Sicks JD, O’Fallon WM, Wiebers DO. Survival and recurrence after first cerebral infarction: a population-based study in Rochester, Minnesota, 1975 through 1989. Neurology 1998; 50:208 –16.
Am J Prev Med 2006;31(6S2)
S257