Using a Vascular Data Base to Change Surgeon Behavior in Managing Patients Undergoing Carotid Endarterectomy: a 5-year prospective study of 598 patients

BACKGROUND: Several data bases for carotid endarterectomy exist (e.g. SVS and NSQIP). However, they are not linked with explicit quality improvement methodology. Over a five-year period, we tested whether collecting data and actually using it methodically could improve carotid endarterectomy outcomes in a nonacademic heart and vascular center with several disparate private practice groups. If successful, we theorized that such a model, based on the Vascular Study Group of Northern New England (VSGNNE), could encourage creation of a Southern Vascular Study Group for Quality Improvement.
METHODS: With permission of the VSGNNE, we applied their data collection system and quality-improvement methodology to study and improve patient management amongst vascular surgeons in three separate nonacademic practices. Although we studied all major vascular procedures, the first study was focused on carotid endarterectomy. A dedicated study coordinator assisted in data collection and analysis. Problematic outcomes were discussed quarterly to identify methods to improve them. A consensus for improvement was achieved and applied immediately. We focused explicitedly on combined stroke/death rates, postop-bleeding and returns to OR, length of hospital stay (LOS), and managing CV risk factors (i.e. preop and postop antiplatelet agents, statins, and beta-blockers). On a yearly basis, we informed our referring physicians of our outcomes.
RESULTS: Between 2004-2009, 598 patients underwent endarterectomy for symptomatic (31%) and asymptomatic (69%) carotid stenosis. Combined stroke/death rate remained consistently low at 1.0% (6/598). Significant improvement was achieved in several areas. Preop and postop antiplatelet use increased from 93% to 100%. Preoperative coverage with statins went from 48% to 83%. Appropriate beta-blockade also improved (31% t0 56%). Postoperative hematomas requiring return to OR were essentially eliminated. Length of stay was reduced from 2.45 days to 1.2 days (p<0.01) in four years. As outcomes became transparent to referring physicians, annual surgical volume rose from 102 CEAs in 2005 to 161 CEAs in 2008.
CONCLUSIONS: A systematic program for both data collection and quality improvement, based on the previously successful VSGNNE, is applicable and effective in another geographic region and amongst nonacademic practices. This study encourages the formation of a Southern Vascular Study Group for Quality Improvement for both academic and nonacademic vascular surgeons.