• DOI:10.1161/str.46.suppl_1.tp280
• Corpus ID: 73821374

Abstract T P280: Do-not-resuscitate Orders Are Not Increasing Observed or Risk Adjusted In-hospital Mortality in Patients With Intracranial Hemorrhage

@article{Hemmen2015AbstractTP,
  title={Abstract T P280: Do-not-resuscitate Orders Are Not Increasing Observed or Risk Adjusted In-hospital Mortality in Patients With Intracranial Hemorrhage},
  author={Thomas M. Hemmen and Eema Hemmen},
  journal={Stroke},
  year={2015}
}

• T. Hemmen, E. Hemmen
• Published 1 February 2015
• Medicine
• Stroke

Background:
Patients with Intracranial Hemorrhage (ICH) have a significant in-hospital mortality. Health systems are compared using risk adjusted mortality indices, which factor in Do Not Resuscitate (DNR) status. It is possible that hospitals overuse DNR and affect mortality and/or influence adjusted outcome measures. We explored a large inpatient database, the University HealthSystem Consortium® (UHC) to assess the frequency of DNR orders and effect on outcomes in patients with ICH
.

Methods:
We included all hospitals within the UCH UHC database that had more than 50 ICH discharges (ICD-9 code 431) between 01/2012 and 12/2013. Hospitals were grouped by frequency of DNR order status: Group 1: 0-22.9% of all ICH discharges, Group 2: 23.0-40%. The groups were compared by frequency of DNR order (ICD-9 v49.86), Mortality and Case Mix Index (CMI), and in-hospital mortality. We used unpaired t-test for comparing means (continuous variables).

Results:
We analyzed 25,273 discharges from 120 hospitals (63 in Group 1, 57 Group 2). There was no significant difference in volume (209.2±137.0 versus 212±115.8) or in-hospital observed mortality (22.2%±5.2% versus 24.1%±6.2%). The Mortality Index (1.07±0.62 versus 0.92±0.24) and CMI (3.39±0.78 versus 3.02±0.66) were higher in Group 1.

Conclusion:
There is no evidence that frequent use of DNR orders affects in-hospital mortality in patients with ICH, or the DNR orders are used to influence risk adjustment models in the UHC database. Further nationwide analyses using diverse health systems are required to confirm these findings to assure generalizability.