Background: The Center for Medicare and Medicaid Services (CMS) tracks annualized 30 day mortality data for patients with ischemic stroke across hospitals. Due to the increased complexity and severity of stroke patients transferred to tertiary centers, mortality rates would be expected to be higher than at institutions which transfer these patients. Publicly available mortality data may not adequately account for this referral pattern, making tertiary care centers appear to provide inferior care.
Methods: We retrospectively analyzed all recorded deaths within 30 days of ischemic stroke from 2009-2011 at our Comprehensive Stroke Center. Data was abstracted from patient charts, including the NIHSS as recorded or calculated retrospectively based on documented exam, admission time, time of first palliative care consult, time of comfort care orders and time of death. Descriptive and nonparametric statistics were used.
Results: There were 42 CMS patient deaths reviewed, median age (interquartile range, IQR) was 82 (75-87), median NIHSS = 22.5 (IQR 19-27), 55% were transferred in, 24% received IV tPA, 36% died within 72 hours of admission, 36% had Palliative Care consults and 90% were placed on comfort care. Transfer patients that died did not differ from direct admit deaths by age, NIHSS or death within 72 hours; NIHSS scores averaged 21 in the transferred group vs. 23 in the direct admits, death within 72 hours was 39% for transfers vs 31% for direct admits. There was a trend toward increased use of tPA (35% transfers vs. 11% direct admits, P = 0.07). Patients that died within 72 hours of admission had significantly higher NIHSS scores (26 (IQR 24-31) vs. 20 (IQR 13-25), P = 0.0025) and were less likely to have received a Palliative Care consult (13% vs. 48%, P = 0.024); there were no differences in age or other factors.
Conclusions: Over half of the deaths counted in our institutions 30 day ischemic stroke mortality rate were from transfers. Almost all of these patients were placed on comfort care. Many passed away within 72 hours of admission. Early withdrawal of care for severe ischemic strokes in elderly patients may reflect higher, rather than lower, quality of care. Including early deaths with comfort care may bias the CMS mortality measure against tertiary referral centers.