Objectives:
In 2009, the VHA reported inpatient stroke quality indicators based on chart abstractions of fiscal year (FY) 2007 data at all VA medical centers (VAMCs). Prior to a randomized trial of a Systems Redesign-based intervention, we re-measured VA stroke quality indicators (QIs) from 2009 data in 11 of the largest volume VAMCs. The purpose of this analysis is to examine whether any significant changes occurred in inpatient stroke care in these sites between 2007 and 2009.
Methods:
Data for 10 Joint Commission (JC) inpatient stroke QIs were obtained by experienced external VA chart abstractors via review of FY 2007 electronic medical records. We abstracted 2009 data at 11 sites as baseline data for a quality improvement randomized study. We calculated eligibility and passing rates for ten inpatient stroke QIs defined similarly to the 10 JC indicators from the FY 2007 study. We compared patient demographics, clinical variables, and passing rates for each QI between the FY 2007 and CY 2009 data at the 11 sites using Student’s t-test and Chi-square tests.
Results:
Comparing 2007 (N =750) to 2009 (N =817) data, mean age (66.3, 66.6), % male (97%, 96%), and % Black (34%, 33%) were similar but mean NIH Stroke Scale score was increased in 2009 (4.2, 5.9, p < 0.001). Three QIs were unchanged over time: DVT prophylaxis, anticoagulation for atrial fibrillation, and antithrombotic at discharge (Table). Performance on four indicators was significantly improved: dysphagia screening (16%, 45%), receipt of rehabilitation consultation (62%, 89%), stroke education (17%, 31%), and receipt of tPA (17%, 47%). Performance on three indicators was significantly reduced: antithrombotic by hospital day two (98%, 87%), cholesterol lowering medication at discharge (90%, 72%), and receipt of smoking cessation counseling (100%, 89%).
Conclusions:
Prior to VHA national quality improvement efforts, both positive and negative shifts in performance occurred for common inpatient stroke QIs. Future work should examine whether focusing efforts on one aspect of stroke care can lead to reduction in quality in other areas, and on whether consistent reporting of these QIs can promote maintenance of high quality stroke care across a large national healthcare system.
Method for developing national quality indicators based on manual data extraction from medical records
