Use of a Registry to Improve Acute Stroke Care --- Seven States, 2005−2009

Each year, an estimated 795,000 persons in the United States experience a stroke. In all, an estimated 7 million U.S. residents have had a stroke, and the cost of cerebrovascular disease in 2007 was estimated at $41 billion (1). During 2004--2009, CDC funded seven state health departments for different periods to implement stroke care improvement registries. As part of the quality-improvement activities conducted by the Paul Coverdell National Acute Stroke Registry (PCNASR), CDC monitors adherence to 10 quality-of-care measures. This report documents the status of adherence to these 10 performance measures of stroke care through these seven state health departments for the period 2005--2009. The number of participating hospitals in the seven states ranged from 112 to 229 per year, with a total of 303 participating hospitals during the 5-year period. Average annual improvement in adherence to stroke care measures ranged from zero in one measure (prescription of antithrombotic at discharge) to 11% (use of intravenous tissue plasminogen activator [IV tPA]); five measures showed average annual improvements of at least 6%. The proportion of patients receiving "defect-free care" (a descriptor that indicates whether patients received all of the quality-of-care measures for which they were eligible) also improved. These results demonstrate the ability of state health departments to collaborate with hospitals to monitor and improve the delivery of high-quality care for acute stroke patients. This role for public health and state health departments in hospital quality improvement for acute stroke care is a viable approach to improving the quality of stroke care in the United States.

PCNASR was established by Congress in 2001. Its goals are to 1) measure, track, and improve the quality of care and access to care for stroke patients from the onset of symptoms through rehabilitation and recovery; 2) decrease rates of premature death and disability from acute stroke; 3) eliminate disparities in care; 4) support the development of systems of stroke care that emphasize quality of care; and 5) improve access to rehabilitation and opportunities for recovery after stroke. PCNASR is funded by CDC through a cooperative agreement with state health departments. It requires state health departments to recruit hospitals that are representative of all hospitals within a state in terms of size, status as an academic teaching hospital, rural or urban setting, geographic location, and racial/ethnic populations served. The state health departments work with participating hospitals to track the care of hospitalized stroke patients to improve the quality of acute stroke care from the onset of stroke through hospital discharge. Specifically, the health departments provide hospitals with an interactive program of quality-improvement activities through conference calls, in-person meetings, and individual hospital site visits. Topics include education on data-driven quality-improvement methodology and overcoming barriers to improving stroke care. The methods for data collection and the types of data collected on patients and hospitals have been described previously (2).

In 2008, CDC, the American Heart Association, and the Joint Commission (an independent organization that accredits and certifies health-care organizations and programs) developed a set of 10 performance measures of acute stroke care (Box) (3). These measures are derived from published care guidelines and clinical trials; eight of the 10 measures are endorsed by the National Quality Forum as important measures of care (4). Nine of the measures were bundled into groups that define defect-free care* (5).

CDC examined 2005--2009 data on all patients aged ≥18 years who were admitted to a participating hospital with any of four clinical diagnoses†: 1) acute hemorrhagic stroke (intracerebral hemorrhage or subarachnoid hemorrhage); 2) ischemic stroke; 3) ill-defined stroke (not classified as hemorrhagic or ischemic); or 4) transient ischemic attack (TIA). In-hospital mortality was calculated for each stroke type as the percentage of patients who died during admission for acute stroke. Cochran-Armitage tests were used to test for a trend in improvement over time.

During the 5-year period, 139,260 patients with one of the four diagnoses were admitted to a participating hospital in the seven states. Overall, 18,921 (13.6%) patients had hemorrhagic stroke; 82,066 (58.9%) had ischemic stroke; 8,236 (5.9%) had ill-defined stroke; and 30,037 (21.6%) had TIA. Median age of patients was 72 years (range: 18−108 years); 53% were female, and 74% were non-Hispanic white.

In-hospital mortality for hemorrhagic stroke (intracerebral or subarachnoid hemorrhage) increased from 21.0% in 2005 to 23.7% in 2009 (p=0.02). However, in-hospital mortality for ischemic stroke decreased significantly, from 6.2% in 2005 to 5.1% in 2009 (p<0.001). The in-hospital mortality for ill-defined stroke declined (but not significantly), from 6.2% to 4.9%, and the in-hospitality mortality rate for TIA ranged from 0.1% to 0.3% with no significant trend.

Adherence to nine of the 10 performance measures increased significantly during 2005--2009 (Table). The one measure that did not show improvement was provision of antithrombotic therapy at discharge, but this measure was at 98% in 2005 and remained at 98% in 2009. The greatest overall improvements during the 5-year period were adherence to the use of thrombolytic therapy (IV tPA) (average annual improvement of 11%), counseling on smoking cessation (9.2%), and lipid testing and/or treatment (7.6%).

Defect-free care improved significantly (p<0.001) for patients in each of the four categories: adherence to measures for patients with TIA showed the greatest average annual improvement (21%, from 28% to 57%), followed by inpatient measures for patients with ischemic stroke (17%, from 37% to 69%), measures for patients with hemorrhagic stroke or ill-defined stroke (17%, from 31% to 57%), and discharge measures for patients with ischemic stroke (8%, from 51% to 72%) (Table).

Reported by

MG George, MD, X Tong, MPH, PW Yoon, ScD, Div for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, CDC.

Editorial Note

Results from the initial (prototype) phase of PCNASR (i.e., from the period 2001--2004) indicated a range of adherence to recommended guidelines for care of stroke patients: 45% adherence to screening for dysphagia, 34% adherence to lipid profile measurement, 21% adherence to smoking-cessation counseling, 91% for providing antithrombotic medication at discharge, and 74% for anticoagulation for atrial fibrillation (6). The results described in this report for 2005--2009 document substantial and significant progress in performance, and compare favorably with those of other major stroke care quality-improvement programs involving predominantly academic hospitals. One such program, Get with the Guidelines --- Stroke (GWTG-Stroke) (7), documented improvement in hospitals participating during the same period as PCNASR (i.e., 2005--2009). Improvements in the PCNASR measures were similar to GWTG-Stroke for seven identical performance measures reported (lipid testing and/or treatment, deep venous thrombosis/venous thromboembolism prophylaxis, antithrombotics at discharge, antithrombotics at 48 hours, anticoagulation for atrial fibrillation, and use of IV tPA) (7). The progress shown for PCNASR is especially encouraging considering that the majority of PCNASR hospitals are small and/or nonacademic and have fewer resources to devote to stroke care quality-improvement programs.

The decrease demonstrated in ischemic stroke in-hospital mortality might be a result of improvement in defect-free care and increased use of IV tPA, although other factors, such as increased use of discharge to hospice, shortened length of stay, or increased use of stroke-prevention medications also might have affected mortality rates. Although a small increase was observed in mortality for hemorrhagic stroke, this unadjusted rate is consistent with previous reports documenting 7-day mortality rates >30% (9). The increase might also reflect changes in characteristics of participating hospitals over time, and this finding warrants further study. Measures of acute-phase care (exclusive of discharge measures) for hemorrhagic stroke patients in this registry include only dysphagia screening and deep venous thrombosis/venous thromboembolism prophylaxis. These patients often develop complex neurologic and hemodynamic conditions associated with their stroke that are not addressed by current performance measures. The GWTG-Stroke program reported comparable rates of in-hospital mortality for ischemic stroke (5.5%) and hemorrhagic stroke (20%--25%) (7).

Little improvement was observed in three of the defect-free care measures through 2008, but in 2009 the three showed substantial improvement. For stroke education, improvement was observed during 2005--2007, declining in 2008 and improving in 2009. The stroke education measure changed in 2008, when stricter requirements were established for reporting adherence; this change likely explains the decline observed in 2008. Dysphagia screening and lipid testing and/or treatment showed little improvement until 2008 and 2009. PCNASR programs worked with hospitals to improve adherence to dysphagia screening and lipid management guidelines. Although use of anticoagulation for atrial fibrillation has improved modestly over time, further improvement is warranted.

The findings in this report are subject to at least four limitations. First, improvement in the quality of care over time most likely is multifactorial, and thus establishing a causal relationship between participation in PCNASR and improvement in quality of care is difficult. Dysphagia screening and provision of stroke education are complex processes. Improvement in stroke education and dysphagia screening has been the focus of intensive quality-improvement efforts recommended to hospitals by state health departments. Because hospital recruitment occurs on a different schedule for each state, this study assessed overall improvement of the program by calendar year rather than assessing improvement over time by length of hospital participation in the program. Second, although state health departments are encouraged to recruit hospitals that are representative of hospitals throughout their state by size, geographic location, and teaching status, participation by hospitals is voluntary, which limits generalizability of these results to other states and hospitals. Third, this report did not include modeling to assess trends in stroke care during the study period. Missing data excluded cases from entering into a performance measure, and this might lead to bias in performance measure estimates. Finally, although PCNASR is intended to include all stroke cases, the completeness of case abstraction varies by hospital and state (state registries report abstraction of 75%--100% of eligible cases). Sampling of cases at the hospital level might lead to biased estimates of performance measures.

Additional efforts in primary prevention, including prevention and control of hypertension and hypercholesterolemia, smoking-cessation counseling, provision of smoke-free environments, and promotion of healthy lifestyle behaviors (e.g., regular physical activity and a healthy diet), are needed to reduce the substantial burden of stroke. PCNASR addresses secondary prevention to reduce the recurrence of stroke. The importance of initiating secondary prevention among ischemic stroke patients at the time of discharge from the hospital was shown by the Preventing Recurrence of Thromboembolic Events through Coordinated Treatment (PROTECT) program. PROTECT was designed to improve early initiation of secondary prevention and resulted in significantly fewer adverse vascular events and improved clinical outcomes through better adherence to the use of medications for secondary prevention (8). The Joint Commission recently added stroke care measures to its core hospital quality measures; in 2009, the Centers for Medicare & Medicaid Services added participation in a stroke care quality-improvement registry as a quality indicator and has added stroke care measures to its core quality measures for reporting on meaningful use of electronic health records. These efforts will assist in improving care for stroke throughout the United States. PCNASR provides a strong model for state health departments as they work to develop a comprehensive, coordinated approach to responding to the challenge of stroke care from prevention, through prehospitalization emergency services and acute care, to rehabilitation and recovery.

Improving the quality of stroke care requires a multidisciplinary, systems-focused effort that engages both public and private entities. State health departments perform a vital role by providing leadership in elevating the level of stroke care and promoting access to high-quality, organized care for stroke through policy development and quality assurance (10).

Acknowledgments

This report is based, in part, on data provided by the Paul Coverdell National Acute Stroke Registry in seven U.S. states (Georgia, Illinois, Massachusetts, Michigan, Minnesota, North Carolina, and Ohio).

References

1. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke statistics---2011 update: a report from the American Heart Association. Circulation 2010;123:e18--209.
2. CDC. Paul Coverdell National Acute Stroke Registry Surveillance---four states, 2005--2007. MMWR 2009;58(No. SS-7).
3. CDC. Consensus stroke performance measures. Atlanta, GA: US Department of Health and Human Services, CDC; 2010. Available at http://www.cdc.gov/dhdsp/docs/pcnasr_performance_measures.pdf. Accessed February 16, 2011.
4. National Quality Forum. National voluntary consensus standards for stroke prevention and management across the continuum of care: a consensus report. Washington, DC: National Quality Forum; 2009.
5. Institute for Health Care Improvement. Bundle Up for Safety. Cambridge, MA: Institute for Health Care Improvement; 2011. Available at http://www.ihi.org/ihi/topics/criticalcare/intensivecare/improvementstories/bundleupforsafety.htm. Accessed February 16, 2011.
6. Reeves MJ, Broderick JP, Frankel M, et al. The Paul Coverdell National Acute Stroke Registry: initial results from four prototypes. Am J Prev Med 2006;31(6 Suppl 2):S202−9.
7. Fonarow GC, Reeves MJ, Smith EE, et al. Characteristics, performance measures, and in-hospital outcomes of the first one million stroke and transient ischemic attack admissions in Get With The Guidelines---Stroke. Circ Cardiovasc Qual Outcomes 2010 [Epub ahead of print].
8. Rahiman A, Saver JL, Porter V, et al. In-hospital initiation of secondary prevention is associated with improved vascular outcomes at 3 months. J Stroke Cerebrovasc Dis 2008;17:5−8.
9. Flaherty ML, Haverbusch M, Sekar P, et al. Long-term mortality after intracerebral hemorrhage. Neurology 2006;66:1182−6.
10. George MG, Matters MD, McGruder HF, Valderrama AL, Xie J. The role of public health in promoting quality improvement in care for stroke and heart disease. Prev Chronic Dis 2008;5:A62.