[Design and analysis of two groups interrupt time series]

Area of Science:

• Epidemiology
• Biostatistics
• Health Services Research

Background:

• Interrupted time-series (ITS) is a valuable quasi-experimental design for assessing intervention impact.
• Single-group ITS designs may be limited by confounding factors.
• Two-group ITS designs offer improved control for pre-interventional confounders.

Purpose of the Study:

• To elucidate the principles and statistical methodologies of two-group interrupted time-series (ITS) analysis.
• To demonstrate the application of two-group ITS in evaluating vaccine efficacy against disease incidence.
• To highlight the advantages of two-group ITS in controlling for confounding variables.

Main Methods:

• Utilized a two-group interrupted time-series (ITS) quasi-experimental design.
• Applied Prais-Winsten and Newey-West regression methods for model fitting.
• Employed a case study evaluating vaccine effects on disease incidence in two cities.

Main Results:

• The two-group ITS design effectively controls for pre-interventional confounding factors.
• This method provides a more robust evaluation of intervention effectiveness, especially when concurrent interventions exist.
• Detailed comparison and explanation of regression model results were provided.

Conclusions:

• The two-group ITS method offers enhanced control over confounding factors compared to single-group ITS.
• This approach is particularly effective for balancing pre-existing trends when interventions coincide with other confounding factors.
• The two-group ITS design holds significant practical importance for program evaluation and policy-making.