Quality of Acute Stroke Care Improvement Framework for the Paul Coverdell National Acute Stroke Registry

Sustaining Stroke Registries: Controversy, Challenge, and Opportunity

Quality of Acute Stroke Care Improvement Framework for the Paul Coverdell National Acute Stroke Registry Facilitating Policy and System Change at the Hospital Level Kenneth A. LaBresh, MD Abstract:

The Paul Coverdell National Acute Stroke Registry prototypes baseline data collection demonstrated a significant gap in the use of evidenced-based interventions. Barriers to the use of these interventions can be characterized as relating to lack of knowledge, attitudes, and ineffective behaviors and systems. Quality improvement programs can address these issues by providing didactic presentations to disseminate the science and peer interactions to address the lack of belief in the evidence, guidelines, and likelihood of improved patient outcomes. Even with knowledge and intention to provide evidenced-based care, the absence of effective systems is a significant behavioral barrier. A program for quality improvement that includes multidisciplinary teams of clinical and quality improvement professionals has been successfully used to carry out redesign of stroke care delivery systems. Teams are given a methodology to set goals, test ideas for system redesign, and implement those changes that can be successfully adapted to the hospital’s environment. Bringing teams from several hospitals together substantially accelerates the process by sharing examples of successful change and by providing strategies to support the behavior change necessary for the adoption of new systems. The participation of many hospitals also creates momentum for the adoption of change by demonstrating observable and successful improvement. Data collection and feedback are useful to demonstrate the need for change and evaluate the impact of system change, but improvement occurs very slowly without a quality improvement program. This quality improvement framework provides hospitals with the capacity and support to redesign systems, and has been shown to improve stroke care considerably, when coupled with an Internet-based decision support registry, and at a much more rapid pace than when hospitals use only the support registry. (Am J Prev Med 2006;31(6S2):S246 –S250) © 2006 American Journal of Preventive Medicine

T

he Paul Coverdell National Acute Stroke Registry prototypes were charged with the task of creating data elements and a data-collection system, as well as analyzing and using the data to improve stroke care. The registries from the initial four prototype states—Georgia, Massachusetts, Michigan, and Ohio— demonstrated the ability to collect baseline data for clinically important quality measures (Table 1) and revealed significant opportunities to improve the delivery of evidenced-based care represented by the measures.1 Much emphasis has been placed on the acute treatment of stroke,2 yet ⬍5% of eligible acute stroke patients in this four-state sample received thrombolytic therapy. The literature not only emphasizes community interventions to reduce delays in presentation, but also From MassPro, Waltham, Massachusetts Address correspondence and reprint requests to: Kenneth A. LaBresh, MD, MassPRO, Inc., 235 Wyman Street, Waltham MA 02451. E-mail: [email protected].

points out the need to substantially improve hospital systems to provide efficient, timely, and safe treatment.3,4 The Coverdell data also suggest that measures for less-complex interventions, including dysphagia screening, lipid profile measurement, and smokingcessation counseling, also have substantial opportunity for improvement, with between one third and one half of eligible patients receiving these interventions.1 Despite considerable work with hospitals by the Centers for Medicare and Medicaid Services through the Quality Improvement Organizations,5 the use of antithrombotic medication at discharge and anticoagulation for patients with atrial fibrillation are still omitted in at least 10% of patients who have no contraindications. In the quality improvement phase of the Coverdell Registry and in the subsequent American Stroke Association’s “Get with the Guidelines–Stroke” (GWTG– Stroke) program, a quality improvement framework was developed and tested. This framework can be divided into three domains: (1) diagnosing barriers to

S246 Am J Prev Med 2006;31(6S2) © 2006 American Journal of Preventive Medicine • Published by Elsevier Inc.

0749-3797/06/$–see front matter doi:10.1016/j.amepre.2006.08.012

Table 1. Treatment rates for selected measures from four pilot prototype registries, the Paul Coverdell National Acute Stroke Registry, 2001–2002 Measure (eligible patients) rtPA-treated Dysphagia screening Lipid profile Smoking-cessation counseling Antithrombotics at discharge Anticoagulation for atrial fibrillation

Georgia (%)

Massachusetts (%)

Michigan (%)

Ohio (%)

3.0 38.5 34.2 21.2

8.5 50.7 38.2 28.2

3.4 50.1 39.4 34.1

3.2 47.2 28.4 16.5

88.1 72.8

94.3 89.8

97.7 91.0

87.7 64.1

rtPA, recombinant tissue plasminogen activator.

the delivery of care, (2) system redesign process, and (3) collaborative model.

Diagnosing Barriers to the Delivery of Care The process of improving the quality of care relies on an accurate diagnosis of the barriers to improved care, and an appropriate “treatment plan” or quality improvement framework based on that diagnosis. Barriers to care can be classified into three general areas: knowledge, attitudes, and behavior (both individual and organizational).6 Knowledge of clinical trial results and guideline recommendations is a necessary prerequisite for delivering evidence-based care. Traditional physician education has focused on presenting this evidence. Too often, however, the dissemination of guidelines was assumed to be sufficient to produce high levels of adherence. As Davis et al.7 point out, traditional didactic presentations are not necessarily associated with changes in physician behavior. Attitudinal barriers include a lack of belief in the data or guidelines, or the perception that evidencebased care is “cookbook” medicine. Such barriers can be addressed through interventions such as peer-topeer communication with trusted opinion leaders. Some attitudinal barriers can, in fact, also represent an underlying deficit in knowledge. For example, hospital physicians’ lack of familiarity with data that support use of cholesterol-reducing medications to reduce vascular events in patients with prior stroke8,9 showing the effectiveness of increased patient adherence to therapy after discharge.10,11 This can result in their resistance to initiating lipid therapy in the hospital because they expect that the primary care physician should do it after discharge. After the literature supporting these concepts is presented, their resistance to treatment initiation during hospitalization fades.12 Even with the knowledge and intention to deliver evidence-based care, however, performance still might not be optimal because of organizational barriers, primarily lack of well-designed systems to ensure reliable delivery of desired care. A goal of these systems might be identification of patients with stroke or transient ischemic attack and the reliable, automatic genDecember 2006

eration of pharmacologic treatment and lifestyle recommendations for all identified, eligible patients. In such a system, physicians can focus on the small subset of patients with contraindications to each intervention and execute an opt-out strategy. Truly automatic systems, often termed forcing functions, avoid reliance on memory by using standard order sets or computerized systems that require the documentation of specific contraindications whenever a recommended therapy is omitted. In the very short hospital stays of patients with acute events, there may be simply too many urgent issues that need to be addressed to also remember the less urgent, but critically important, secondary prevention interventions. Highly reliable systems require a team approach. Emergency department physicians, neurologists, and nurses can work together to provide early treatment, such as thrombolysis, perform dysphagia screening, and obtain early risk assessments, such as a lipid profile via a defined protocol or preprinted order system. An integrated team of people can provide the redundancy needed to avoid an unintentional omission of needed care. Routine use of preprinted orders that specify the use of deep-vein thrombophlebitis prophylaxis, aspirin, warfarin for atrial fibrillation, and statins in all eligible patients (unless a documented contraindication is present) are examples of tools that increase the likelihood of appropriate treatment. Previous work in cardiovascular disease patients has demonstrated a significantly higher performance level within hospitals that engage in quality improvement initiatives in patients in whom these tools were used, compared to those without the use of these tools.13 The collection and feedback of measures from a minimal data set to address the critical assessment and treatment strategies can serve as a reminder to all those involved in the patient’s care. Such systems can lead to substantial improvement in treatment regimens.12,15,16

The System Redesign Process The redesign of clinical care systems is the essence of quality improvement. It requires a strategy that employs clinical and quality improvement professionals involved Am J Prev Med 2006;31(6S2)

S247

in the care of stroke patients executing a continuous quality improvement framework, such as the Model for Improvement. While there are many approaches to clinical quality, this approach is based on continuous quality improvement tools developed by Shewhart, Deming, and Durand in the 1920s, and further refined for healthcare applications during the last 20 years. This model asks three questions: (1) What are we trying to accomplish? (2) How will we know that a change will result in improvement? and (3) What change can we make that will result in improvement?17,18 A multidisciplinary team, consisting of a clinical leader, a system leader (usually a nurse manager or director), and a day-to-day leader, meet to define the scope of their effort. Typically, they will start with a defined pilot population, such as all patients with acute stroke who are admitted to a stroke unit. To answer the first question, they define numeric goals for several measures of treatment (e.g., use of aspirin, statins). The team then begins to address the second question by designing a series of tests to evaluate and adapt specific strategies to improve care in their environment. Teams can refine their order sets, protocols, and other strategies by testing ideas on a small scale at first. At this level, a number of new ideas can be tested over very short periods of time while maintaining the old system. Some ideas will be adopted for testing on a wider scale; some will be adapted based on feedback from the team members who are trying them out; and some will be abandoned. “Failures” on this small scale can be considered as lessons learned; they are often useful in adapting an idea into a successful system change. The willingness to explore new possibilities on this scale opens up creative approaches in a safe setting and can lead to truly innovative thinking. As it expands in scale, the testing process provides an opportunity to include physicians and other staff who have a direct hand in creating and modifying the new tools and processes. This, in turn, cultivates additional opinion leaders to convince peers and aid in the inevitable committee approval process. The third question addresses the implementation stage. Only after specific tools, protocols, or other changes have been adapted to the local environment through a series of tests over an increasingly larger scale should these changes be instituted. Implementation, which is done after demonstrated evidence of success, requires widespread communication and support. These strategies need to become part of the standard system of care. Here, there is no room for failure. New systems need to be monitored and adjusted by the ongoing collection of key performance measures, and they need to be adapted as needed over time. With a propensity to maintain the status quo, what accounts for the adoption and spread of innovations? Everett Rogers19 describes five criteria that predict the successful adoption of change that address this issue.

The demonstration of relative advantage implies that the new system is better than the system that is being replaced. Because the process of changing routines is difficult, the new system must perform better and be easier to work in than the old system in order to make the effort worthwhile. Successful change should not be overly complex and its advantage should be observable. Visible successful improvements in one institution can provide confidence to others seeking to improve the quality of care they deliver, by knowing that higher performance levels are achievable. Trialability implies that the innovation should have the capacity for testing on a small scale before full-scale implementation. Because hospital environments tend to differ from each other in a number of ways, both physically and culturally, the ability to adapt solutions to a specific hospital culture provides a sense of ownership, the feeling of changing, rather than being changed. Successfully adopted innovation needs to be compatible with individual or institutional values, priorities, and resources. The ability to test a system change and customize it is one way to help address this issue. If stroke is not an important part of a hospital’s strategic priorities, it is much less likely that a system change to improve stroke care will be successfully adopted. A system change that uses emergency department physicians to administer recombinant tissue plasminogen activator for an acute stroke without the onsite evaluation of a neurologist can be successful in some hospital environments and not feasible in others.

The Collaborative Model Collaboratives, like the systems they target for redesign, require multiple supporting components to be successful. By themselves collaboratives have had mixed results.20 When the additional elements are incorporated, including community support, technology, patient support, and effective teams as modeled by Bodenheimer et al.,21,22 the collaborative model can produce improvement at a relatively rapid pace.12–16,23 In early reports, rates of tissue-type plasminogen activator use for patients with ischemic stroke presenting within 2 hours of symptom onset, have increased threefold.15 Engaging key stakeholders to support the improvement in stroke care, as measured by the Coverdell performance measures, can help to make stroke quality improvement compatible with the strategic priorities of hospitals by bringing community organizations, agencies, and hospital leadership together. The enactment of state regulation to designate acute stroke centers is an example of a profound influence on compatibility with institutional priorities. Multidisciplinary teams from multiple hospitals are brought together in a series of workshops (Figure 1). The initial meeting addresses the knowledge barrier by providing didactic presentations and materials to dis-

S248 American Journal of Preventive Medicine, Volume 31, Number 6S2

www.ajpm-online.net

Participants (hospital teams)

Baseline data prework

overdell data set

Stakeholder meeting

P A Develop framework and changes

D

A

S

LS 1

Planning group

P

P D

A

S

LS 2

D S

LS 3

Team support Barrier /solution conference calls E-mail Online discussion groups Visits Phone assessments Monthly team rep Figure 1. The collaborative framework. LS, learning session; P, plan; D, do; S, study; A, act.

seminate clinical science and guidelines that may be thought of as the necessary explicit knowledge needed for change.24 This is the first step and is coupled with the introduction of quality improvement skills, such as the use of Plan-Do-Study-Act cycles to evaluate system changes on a small scale. Hospital teams have the opportunity to demonstrate successful changes, serving to make relative advantage observable. In early sessions, presentations typically come from the early adopters, generally about one in six of any large group that embrace change early.19,25 This demonstration of success encourages the next one third, the early majority, to participate based on the demonstration of success. Facilitated breakout sessions provide the opportunity to exchange the practical knowledge of how ideas were developed and tested. This “how-to” is essential to reduce the complexity of system change and demonstrate that new ideas can be refashioned as “our ideas” and thus more readily accepted.24 In these meetings for stroke as well as other collaboratives across a variety of disease states,14 participants invariably leave with new concepts or ideas to test when they return to their own institution, buoyed by the knowledge that other hospitals have been successful and the belief that “if they can do it, we can do it.” The increasing number of participants working on the changes developed in the program creates a “tipping point,” at which time the new way becomes the thing to do.19,25,26 The conversation and sharing are continued among settings by the use of e-mail lists and conference calls and other team support strategies (Figure 1). “Share December 2006

openly and steal shamelessly,” a motto adhered to in the collaborative environment, appears to significantly speed up the improvement process. In the American Stroke Association’s GWTG–Stroke collaborative, significant improvement in many of the same measures collected in the Coverdell Registries occurred within the first 6 months of participation.12,15 In subsequent sessions, more of the time is devoted to presentations by hospital teams and interactive sessions among teams as more develop successful changes. Maintaining the improvement in performance becomes an increasingly important concern as more hospitals participate. The Rogers criteria provide a framework for the adoption of innovation and its long-term stability. The quality-improvement training at this time focuses on holding the gains that have been made by rewriting job descriptions, reallocating resources, and eliminating past versions of preprinted orders and other forms. The goal now is to make the new, better system easy and reversion to the old system difficult. Measurement of care is an essential ingredient in improving care, but it is not sufficient to improve care, particularly when collected retrospectively. Collected data should be useful and relevant to the quality of care, including timely feedback to have impact on the care of patients during their hospitalization. Data collected concurrently, with feedback at the individual patient level during a hospitalization, has a greater likelihood of improving care when it can be used as a reminder, allowing for correction of an unintended Am J Prev Med 2006;31(6S2)

S249

omission. The time needed to collect data should therefore be minimized by limiting data elements to those needed to produce quality measures and the demographics needed to characterize the patient population, thus freeing time to focus on providing optimal care.14 Using feedback with decision support, such as an accessible review of guidelines and evidence available at the point of care, also has the potential to address both knowledge and system barriers while the opportunity to change care for that patient is still available, particularly for subacute and prevention measures. Concurrent systems using the Internet14 or electronic medical records27 have been demonstrated to play important roles in changing care. Two of the Coverdell prototype programs and GWTG–Stroke use an Internet-based patient management tool to provide an inexpensive system for data collection with embedded point-of-care decision support and online, realtime reporting, and benchmarking of performance. This system does not require hospital electronic medical records and has been demonstrated to work effectively to support system change in stroke, as well as coronary artery disease.16 GWTG–Stroke has shown significant improvement in acute care and hospitalinitiated stroke prevention care measures in less than 1 year using the collaborative methodology and this data-collection methodology.12,15

Conclusion Tools and processes to support data collection and the use of those data, as well as the necessary system and cultural changes needed to utilize the data, are critically dependent on a framework to support hospitals to incorporate these elements. Didactic presentations, hospital team sharing, and interactions among teams are all necessary components to accelerate the rate of improvement. Bringing together resources and expertise to support active quality- improvement programs from diverse organizations, including the Centers for Disease Control and Prevention, the American Stroke Association, Quality Improvement Organizations, and state departments of health, will play a key role in reaching our shared goal of providing the right care for every stroke patient every time. Funding for the Paul Coverdell National Acute Stroke Registry was provided by a cooperative agreement from the Centers for Disease Control and Prevention. No financial conflict of interest was reported by the authors of this paper.

References 1. Reeves MJ, Arora S, Broderick JP, et al. 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.

2. Marler JR, Tilley BC, Lu M, et al. for the NINDS rt-PA Stroke Study Group. Early stroke treatment associated with better outcomes. The NINDS rt-PA Stroke Study. Neurology 2000;55:1649 –55. 3. NINDS rt-PA Stroke Study Group. A systems approach to immediate evaluation and management of hyperacute stroke: experience at eight centers and implications for community practice and patient care. Stroke 1997;28:1530 – 40. 4. Tilley BC, Lyden PD, Brott TG, et al. Total quality improvement method for reducing delays between emergency department admission and treatment of acute ischemic stroke. Arch Neurol 1997;54:1466 –74. 5. Jencks SF, Cuerdon T, Burween DR, et al. Quality of medical care delivered to Medicare beneficiaries: a profile at state and national levels. JAMA 2000;284:1670 – 6. 6. Cabana MD, Rand CS, Powe NR, et al. Why don’t physicians follow clinical practice guidelines? A framework for improvement. JAMA 1999;282: 1458 – 65. 7. Davis D, O’Brien MAT, Freemantle N, et al. Impact of formal continuing medical education: do conferences, workshops, rounds, and other traditional continuing education activities change physician behavior or healthcare outcomes? JAMA 1999;82:867–74. 8. Heart Protection Study Collaborative Group. Effects of cholesterol lowering with simvastatin on stroke and other major vascular events in 20,536 people with cerebrovascular disease or other high-risk conditions. Lancet 2004;363:757– 67. 9. The Stroke Council. Statins after ischemic stroke or transient ischemic attack: an advisory statement from the Stroke Council of the American Stroke Association. Stroke 2004;35:1023. 10. Fonarow GC, Gawlinski A, Moughrabi S, Tillisch JH. Improved treatment of coronary heart disease by implementation of a Cardiac Hospitalization Atherosclerosis Management Program (CHAMP). Am J Cardiol 2001; 87:819 –22. 11. Cannon CP, McCabe CH, Bentley J, et al. Early statin therapy is associated with markedly lower mortality in patients with acute coronary syndromes: observations from OPUS–TIMI 16. J Am Coll Cardiol 2001;37(suppl A):334A. 12. LaBresh KA, Schwamm L, Albright D, et al. Does Get With The Guidelines improve secondary prevention in patients hospitalized with ischemic stroke or TIA? Stroke 2005;36:454. 13. Mehta RH, Montoye CK, Gallogly M, et al. Improving the quality of care for acute myocardial infarction: the Guidelines Applied in Practice (GAP) initiative. JAMA 2002;287:1269 –76. 14. LaBresh KA, Gliklich R, Liljestrand J, et al. Using Get With The Guidelines to improve cardiovascular secondary prevention. Jt Comm J Qual Saf 2003;29:539 –50. 15. Schwamm L, LaBresh KA, Albright D, et al. Does Get with the Guidelines– Stroke improve acute intervention in patients hospitalized with ischemic stroke or TIA? Stroke 2005;36:436. 16. LaBresh KA, Ellrodt AG, Glicklich RG, et al. Get with the Guidelines for cardiovascular secondary prevention: pilot results. Arch Intern Med 2004;164:203–9. 17. Berwick DM. A primer on leading the improvement of systems. BMJ 1996;312:619 –22. 18. Kilo CM. Improving care through collaboration. Pediatrics 1999;103(1 suppl E): 384–93. 19. Rogers E. Diffusion of innovations. New York: The Free Press, 1995. 20. Mittman BS. Creating the evidence body for quality improvement collaboratives. Ann Intern Med 2004;140:897–901. 21. Bodenheimer T, Wagner EH, Grumbach K. Improving primary care for patients with chronic illness. JAMA 2002;288:1775–9. 22. Bodenheimer T, Wagner EH, Grumbach K. Improving primary care for patients with chronic illness: the chronic care model, part 2. JAMA 2002;288:1909 –14. 23. LaBresh KA, Tyler PA. A collaborative model for hospital-based cardiovascular secondary prevention. Qual Manag Health Care 2003;12:20 –7. 24. Dixon NM. Common knowledge. Boston MA: Harvard Business School Press, 2000. 25. Berwick DM. Disseminating innovations in health care. JAMA 2003; 289:1969 –75. 26. Gladwell M. The tipping point: how little things can make a big difference. Boston MA: Little Brown and Company, 2000. 27. Kaushal R, Shojania KG, Bates DW. Effects of computerized physician order entry and clinical decision support systems on medication safety: a systematic review. Arch Intern Med 2003;163:1409 –16.

S250 American Journal of Preventive Medicine, Volume 31, Number 6S2

www.ajpm-online.net