Performance Improvement 1999. Evaluation of Drug Use Review Demonstration Projects: Final Report

02/01/1999

Highlights
An external evaluation was conducted on drug use review (DUR) demonstration programs within the Medicaid agencies of two States: Washington and Iowa. The States of Georgia and Maryland were included in the study for experimental and comparison purposes. Specifically, the study evaluated the efforts of Project C.A.R.E., a joint project of Washington's Department of Social and Human Services and the University of Washington's School of Pharmacy. The Washington effort was designed to test the effectiveness of payments to pharmacists for the provision of cognitive services. The Iowa Medicaid online prospective drug use review (OPDUR) demonstration project was also evaluated and tested. Both evaluation efforts were designed to improve drug prescription methods by influencing pharmacists’ behavior.

The findings from the evaluations of Project C.A.R.E. and the Iowa OPDUR indicate that prospective drug use review did not affect the frequency of drug problems. According to the study, expenditures on and frequencies in the use of prescription drugs and other medical services also were not measurably affected. Further, analysis of data collected from cognitive services (CS) provision also revealed no evidence of changes in the behavior of pharmacists who received OPDUR messages.

This research study found support for making no changes in DUR requirements. The analysis indicates a more effective use of Medicaid DUR funds would be to encourage more retrospective DUR, establish a way of compensating pharmacists for counseling and educating patients, and to support pharmacists by improving their capacity to prescribe.

Purpose
The purpose of this study was to test the effectiveness of the drug use review (DUR) demonstration projects in improving drug use and in reducing the negative consequences of drug misuse in the Medicaid programs of Washington and Iowa. If either of these State demonstrations proved effective, the evaluators were also expected to further study the costs of achieving these improvements compared to other policies and programs designed to improve drug use and to reduce the consequences of drug misuse.

Background
Since 1990, the Omnibus Budget Reconciliation Act has required all State Medicaid programs to implement drug use review (DUR). Also required were demonstrations of innovative models of DUR. Therefore, in 1992, the Health Care Financing Administration (HCFA) funded two experiments managed by two State Medicaid agencies: Project C.A.R.E. in the state of Washington, which tests the effects of paying pharmacists to provide cognitive services (CS); and the Iowa Medicaid OPDUR Demonstration Project, which tested online prospective drug use review (OPDUR). This is the final report of a 1993 HCFA-funded external evaluation of both demonstrations.

Methods
Two methodological tools were used for this evaluation study: The Penn State University (PSU) screener and a bibliographic database. The PSU screener is a knowledge-based research tool used for conducting epidemiologic and economic studies on potentially inappropriate drug therapy. It is capable of examining millions of drug claims records and is intended to be used with large pharmaceutical assistance and Medicaid programs. For this project, the screener was used to compute the occurrence of screen failures for a consistent set of DUR screening criteria across the two demonstration States, as well as the co-experimental and comparison States.

The DUR Outcomes Bibliographic Database was developed as a way of comprehensively and definitively linking DUR screening failures to expected clinical outcomes. The bibliographic review was then used in the selection and evaluation of candidate clinical outcomes.

To test the effects of Project C.A.R.E. on prescription drug use and costs, the evaluation team compared statistical reductions in drug cost and use of Medicaid recipients obtaining prescriptions from demonstration pharmacies against a control group of pharmacies. The same data set, comprised of geographic clusters, was used for the experimental and control groups. This involved random assignments of pharmacies. The study population was the same, with Medicaid recipients who had dual eligibility for Medicare. The sample population included 7,809 elderly subjects and 8,090 non-elderly subjects. Multiple regression models were used to test the study hypotheses.

To implement the Iowa demonstration, pharmacies had to meet a set of eligibility criteria and were recruited through newsletters, letters and articles in the Iowa Pharmacists Association Journal. All the pharmacies participating in the OPDUR demonstration were linked to an online DUR screener that reviewed all medical prescription drug claims as they were submitted electronically for payment. The pharmacies in the experimental group received feedback that identified potential prescription problems. The control group pharmacies received no messages. Although the pharmacies in both groups were told to document all cognitive services provided, they were not offered payment to do so. Iowa also adapted commercial screening software to local conditions by customizing the ProDUR systems, a commercially available online prospective DUR software. The project defined the syntax for modifications and developed software to test the syntax.

Findings
Eight years of drug claims for Medicaid programs were screened in Washington, Iowa, Georgia and Maryland. In the study of Washington's Project C.A.R.E., it was found that pharmacists and prescribers agreed on the benefits of increased communication between the two groups. Survey respondents and those who participated in focus groups supported patient counseling for healthier patients and more qualitative prescription methods. The physicians in the study believed that pharmacists could help them best by increasing narcotics monitoring and offering drug cost options. The pharmacists believed that physicians could best help them if they were allowed access to medical records or patients' diagnoses. The pharmacists also believed that they should be compensated for patient counseling.

Findings from Project C.A.R.E. demonstrate that CS payment is replicable in other States, that it is useful in providing a clear definition of what CS is, and that the provision of cognitive services to patients has no measurable impact on either drug use or cost. Project C.A.R.E. did not conduct a demonstration comparing OPDUR and cognitive services. Therefore, this study could not determine whether pharmacists had a preference for drug use review.

The Iowa study sought to determine if a system that instantly screened prescriptions for errors would improve pharmacists knowledge, increase their use of cognitive services, and improve optimal drug utilization, thereby decreasing sub-optimal drug use by Medicaid recipients. However, there was no evidence that OPDUR affected pharmacists' behavior. It was concluded from analysis of the Iowa OPDUR Demonstration and Project C.A.R.E. efforts that prospective DUR has no measurable effect on the frequency of drug problems, on how much is spent, how frequently prescriptions are used, or on clinical outcomes.

Use of Results
Since 1990, the Medicaid program has been mandated to operate prospective and retrospective drug use review programs. Now that most Medicaid patients (with the exception of the adult disabled and Medicaid/Medicare eligible elderly) will be folded into managed care operations, the question becomes whether policy makers should advocate for changes in DUR requirements based on current knowledge, fund additional research studies, or leave the system as it is.

The findings gained from this study indicate that a more effective use of Medicaid DUR funds would be to encourage more retrospective DUR, and to establish a way of compensating pharmacists for counseling and educating patients. The findings further indicate that OPDUR can continue to support pharmacists by improving their capacity to prescribe. This might be a more productive way to use public money instead of payments for generalized cognitive service provision, or standardizing and validating OPDUR screening systems.

This evaluation study also supports making no changes in the current system. There was no evidence found to support DUR policy changes that would be cost-effective for the Medicaid program or Medicaid recipients' sub-optimal drug use. There also was no evidence that OPDUR affected pharmacists' behavior. Further conclusions, from analysis of the Iowa OPDUR Demonstration and Project C.A.R.E. efforts, also indicate that prospective DUR has no measurable effect on the frequency of drug problems, how much is spent, how frequently prescriptions are used, or on clinical outcomes.

AGENCY SPONSOR: Health Care Financing Administration, Office of Strategic Planning

FEDERAL CONTACT: Jay P. Bae, Ph.D.

PHONE NUMBER: 410-786-6591

PERFORMER ORGANIZATION: Abt Associates, Cambridge, MA