Session number: 1090
Abstract title: Adoption of New Medication Recommendations for Diabetes
Author(s):
MJ Pugh - Center for Health Quality, Outcomes, & Economic Research
JA Anderson - Boston University School of Medicine
LM Pogach - Chief, Endocrinology, VA New Jersey Health Care System
University of Medicine and Dentistry of New Jersey/New Jersey Medical School
DR Berlowitz - Center for Health Quality, Outcomes, & Economic Research
Boston University School of Medicine, Boston University School of Public Health
Objectives: Understanding adoption patterns of new therapies by clinicians is crucial to improving quality of care. This is particularly true in diabetes, where research demonstrates that intensive treatment using newer medications and multi-medication regimes improves glycemic control and decreases complications. We examine prescribing patterns to gauge adoption of novel regimes recommended by 1997 VA guidelines and clinical studies.
Methods: We linked FY97-FY99 pharmacy, laboratory, and administrative databases (VISN1 and VISN8) to identify veterans with diabetes who were on hypoglycemic medications at some point between FY97 and FY99 (N=20,639). Medication regimes and clinic visits were classified each year. “Traditional” regimes included insulin and sulfonylurea mono-therapies, and “Novel” regimes included multiple oral medications, insulin-oral combinations, and newer “glitazone” medications. Veterans seen in a diabetes clinic had a specialty visit coded for current and subsequent years. Bivariate analyses examined prescribing patterns. Logistic regression models fit using generalized estimating equations determined if adoption of new regimes occurred, and if adoption varied by generalists and specialists. Age, gender, race, VISN, glycemic control (HbA1c), and presence of diabetes complications were controlled.
Results: Between FY97 and FY99, the use of novel regimes increased from 27% to 39% (p<.001). A decrease in sulfonylurea monotherapy (44% to 33%) drove decreases in traditional regimes, while all categories of novel regimes experienced significant gains. Further, veterans seen by specialists at some point were more likely (p<.001) to receive novel regimes than did those seen only in primary care (40% vs. 31%). Logistic regression models indicated that novel regimes were more likely to be prescribed in FY99 (OR: 2.41, CI 2.29-2.55). Veterans seen by a specialist were more likely (OR: 1.21, CI 1.14-1.28) to receive novel regimes, despite controlling for disease severity. Veterans from VISN8 were more likely (OR: 1.47, CI 1.40-1.60) to receive novel regimes than VISN1 veterans.
Conclusions: Novel regimes are being adopted. However, patterns suggest a delay in adoption of recommendations by primary care providers and geographic variations.
Impact statement: As care is increasingly provided by generalists, it is critical to re-examine processes by which new information is disseminated. Research examining methods to facilitate more rapid adoption of recommendations may improve quality of care.