Evaluating Program Effectiveness Using the Regression Point Displacement Design

Abstract

Most health care initiatives are evaluated using observational study designs in lieu of randomized controlled trials (RCT) due primarily to resource limitations. However, although observational studies are less expensive to implement and evaluate, they are also more problematic in determining causality than the RCT. This trade off is most apparent in the initial planning stage of program development. An RCT is generally preferred though the cost of implementing a pilot program using the RCT might outstrip the potential benefit if the desired results are not obtained. This article describes a simple quasi-experimental model called the regression point displacement (RPD) design, which compares the prepost results of a single or multiple treatment groups to that of a control population. This design has shown great potential in evaluating health care pilot programs or demonstration projects-especially those that are community based-due to its relative ease of implementation and low cost of analysis.