A Mixed-Effects Interrupted Time Series Model for Assessing Complex Health Care Interventions Across Multiple Units

• Activity Number: 18 - New Models, Diagnostics, and Considerations in Evaluating Intervention and Policy Effects
• Type: Topic Contributed
• Date/Time: Sunday, August 7, 2022 : 2:00 PM to 3:50 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Maricela Cruz* and Andrew Holbrook and Daniel L Gillen
• Companies: Kaiser Permanente Washington Health Research Institute and UCLA Biostatistics and University of California Irvine
• Keywords: Interrupted Time Series; Segmented Regression ; Complex Interventions; Patient-centered Data; Change Point Detection; Hierarchical Model

Abstract

Interrupted time series (ITS) designs are aptly situated for studying the impacts of large-scale public health policies, as they borrow from case-crossover designs and serve as quasi-experimental methodology that can retrospectively assess the impact of an intervention. Statistical models used to analyze ITS designs are traditionally single unit segmented regression models, inherently assume a change point exists, and either restrict the intervention’s effect to a preset time point or remove data for which the intervention effects may not be realized. We propose the Mixed-Effects Robust ITS (MERITS) model that hierarchically incorporates multiple units, allows for and estimates unit-specific change points when appropriate, and accounts for changes in temporal dependence post-intervention. Alongside MERITS, we flexibly model the unit-specific change points, quantify uncertainty in these estimates, and conduct empirical type one error and power simulations. We demonstrate the methodology by analyzing multi-unit patient centered data from a hospital that implemented a new nursing care delivery model.