Evaluating disease management programme effectiveness: an introduction to the regression discontinuity design

Area of Science:

• Health Services Research
• Program Evaluation
• Public Health

Background:

• Disease management (DM) programs have existed for over a decade, but their effectiveness in improving health and reducing costs remains uncertain.
• Traditional evaluations often use weak observational designs prone to selection bias and regression to the mean.
• Robust methods are needed to accurately assess DM program impact.

Purpose of the Study:

• To introduce the regression discontinuity (RD) design as a superior alternative to randomized controlled trials for evaluating DM program effectiveness.
• To highlight the methodological advantages of the RD design in overcoming biases inherent in observational studies.
• To emphasize the importance of appropriate participant identification and assignment for successful RD design implementation in DM.

Main Methods:

• The regression discontinuity (RD) design is proposed as an alternative to randomized studies for DM program evaluation.
• This design utilizes a pre-test 'cut-off' score to assign individuals to intervention or control groups.
• Programme effectiveness is assessed by observing changes in the pre-post relationship at the identified cut-off point.

Main Results:

• The RD design allows for the use of varied pre-test measures, not necessarily aligned with outcome measures.
• This flexibility enables DM programs to leverage research-based guidelines and clinical understanding for participant selection.
• The cut-off score can be determined through clinical judgment, empirical data, or resource allocation.

Conclusions:

• The RD design is uniquely suited for evaluating the effectiveness of disease management programs.
• Successful implementation hinges on fundamental changes in how DM programs identify and assign participants.
• Adopting the RD design can lead to more accurate and reliable assessments of DM program outcomes.