RRP 07-295 – HSR&D Study
Career Development Projects
Optimizing Glycemic Control of Hospitalized Stroke Patients
James P. Walsh MD
Richard L. Roudebush VA Medical Center, Indianapolis, IN
August 2008 -
There is evidence that effective management of blood glucoses with insulin after ischemic stroke improves neurologic outcomes and reduces morbidity, mortality, and length of stay. This pilot study tested the impact of an evidence-based, systematic approach for early detection and management of hyperglycemia in acute stroke patients admitted to tertiary care VA Medical Center.
The specific objectives were to: (1) detect hyperglycemia promptly when acute ischemic stroke patients are admitted to the hospital; (2) implement a Computerized Patient Record System (CPRS)-based Decision Support System (DSS) to improve management of hyperglycemia; and (3) conduct an evaluation to determine the barriers and facilitators of the implementation strategy.
An evidence-based protocol for prompt detection and treatment of hyperglycemia in patients admitted with acute stroke or suspected stroke was developed and introduced at our facility. Recommended interventions include blood glucose tests for the first three hospital days and initiation of conservative doses of basal and nutritional insulin if glucose exceeded 150 mg/dL. During the study period, a diabetes specialist met with providers at the start of their VA rotations and provided information on the protocol, the CPRS-based decision support tool, and the rationale for good glycemic control in stroke patients.
A prospective pre-post cohort design is being used to examine the specific objectives. The primary outcome is the mean blood glucose during the 72 hours after the first high glucose measurement. The decline in mean blood glucose during the 72 hours period after the first high glucose and the percent of patient days with poor glycemic control will also be evaluated. To evaluate the barriers and facilitators of protocol adherence, providers caring for stroke patients were contacted by a study coordinator and asked the rationale for the insulin orders they wrote.
Fifty patients with acute stroke or suspected stroke were followed through their hospitalization. Of these 50 patients, 32 had orders written for blood glucose monitoring as recommended by the protocol and 34 became hyperglycemic during their hospital admission. Of the 34 who became hyperglycemic, 10 had orders for basal insulin written as recommended. While the protocol clearly influenced glucose management in some cases, it was not followed in the majority. The major barrier to protocol adherence cited by the providers was a perception that blood glucose management is a low priority and, as a result, numerous other responsibilities often get in the way. Secondary barriers included concerns about insulin use in patients without a diagnosis of diabetes and concerns about hypoglycemia.
Based on results of this study, we are revising the CPRS-based decision support system for insulin administration. Educational presentations for providers have been revised to directly address concerns raised. We are also piloting use of mid-level practitioners to assist physicians with insulin ordering and assessment of glucose data. Demonstration of an effective and safe strategy for glucose management of stroke patients would facilitate the design of a larger, randomized trial to further clarify the benefits of post-stroke glucose lowering. No other results at this time.
None at this time.
Treatment - Observational