Stafford RS, Misra B; Association for Health Services Research. Meeting.
Abstr Book Assoc Health Serv Res Meet. 1999; 16: 380.
Institute for Health Policy, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114, USA.
RESEARCH OBJECTIVE: Patterns of chronic disease treatment have a substantial impact on health care costs. National patterns of hypertension treatment (HTN) have been noted to emphasize newer, more costly medications without evidence of improved outcomes relative to less costly alternatives. As a quality improvement initiative to monitor the cost and outcomes of primary care practice, we used pharmacy claims data to investigate physician practice patterns in a large integrated delivery system. Our goals were to evaluate the usefulness of these data as a method of efficiently monitoring practice patterns and to determine patterns of HTN treatment and their cost implications. STUDY DESIGN: We employed 1997 claims data available on 126,184 members of a capitated, employer-based risk contract population in Eastern Masschusetts. Data on 619,530 pharmacy claims were merged with ICD-9 diagnostic information from 1,361,234 medical claims. Medications potentially used in the treatment of HTN were identified by National Drug Codes (NDCs). PRINCIPAL FINDINGS: An ICD-9 code for HTN was noted for 9% of adult members. Of the 7% of adult members taking anti-HTN medications, 50% did not have a diagnosis of HTN noted on medical claims. Leading diagnoses in these members were chest pain, diabetes and hyperlipidemia. Beta blockers (BBs, 31%) and ACE inhibitors (ACEIs, 29%) were the most common anti-HTN medications, followed by calcium channel blockers (CCBs) and diuretics (both 19%). ACEIs (32%) and BBs (27%) remained the most common agents for those patients that had a HTN diagnosis. The mean cost of anti-HTN prescriptions was $32 (including co-payments) or $2.2 million in aggregate. Anti-HTN medications represented 9% of aggregate drug costs for all conditions. Because of their high cost per prescription, CCBs ($57) and ACEIs ($44) accounted for 73% of all anti-HTN costs and 6% of all drug costs. Compared to patients with established therapy, patients identified as being newly treated for HTN had less frequent use of CCBs (8%), more frequent use of diuretics (28%), and unchanged use of ACEIs (30%). CONCLUSIONS: Using claims data to monitor anti-HTN drug use has several weaknesses, particularly those imposed by the limitations of diagnostic coding on medical claims. Pharmacy claims by themselves, however, provided useful data on practice patterns and their costs. These patterns suggest that current practice patterns may not be consistent with national practice guidelines. Practices for newly diagnosed patients appear to be more consistent with these guidelines, perhaps because of increasing cost-consciousness among physicians in their decisions regarding medication initiation. IMPLICATIONS FOR POLICY, DELIVERY, OR PRACTICE: Despite their limitations, claims data can provide useful information to quality improvement initiatives seeking to improve the cost-effectiveness of prescribing. These data provide a relatively inexpensive means of supplying feedback to physicians regarding their practices, an intervention that may help produce needed improvement in the management of hypertension.
Publication Types:
Keywords:
- Adrenergic beta-Antagonists
- Adult
- Angiotensin-Converting Enzyme Inhibitors
- Calcium Channel Blockers
- Cost-Benefit Analysis
- Diabetes Mellitus
- Diuretics
- Drug Costs
- Health Care Costs
- Humans
- Hypertension
- Monitoring, Physiologic
- Pharmaceutical Preparations
- Physician's Practice Patterns
- Practice Guidelines as Topic
- Prescriptions, Drug
- Primary Health Care
- economics
- hsrmtgs
Other ID:
UI: 102194755
From Meeting Abstracts