A Bootstrap Method to Estimate Cost of Behavioral Intervention Implementation: A Proof of Concept

Abstract

Objective

To develop a bootstrapping method to augment time-driven activity-based costing (TDABC) analysis intended to allow more realistic cost estimates.

Data Sources

Secondary data from a multisite clinical trial conducted from 2016 to 2018 on an ostomy self-management telehealth intervention for cancer survivors.

Study Design

The intervention cost was newly estimated by incorporating expected patient participation rates calculated via bootstrapping. This cost was compared against the cost estimate obtained via traditional TDABC.

Data Collection

Study personnel self-reported the time spent on each activity associated with the intervention. We also utilized patient participation data collected from the trial.

Principal Findings

The total cost of the telehealth intervention estimated via the bootstrapping method was $210,052.62 (95% CI: 208,652.13, 211,402.51), with an average cost per participant of $1981.63 (95% CI: 1968.42, 1994.36). Traditional TDABC analysis yielded $186,363 or $1758 per participant. Further adjusting assumptions about the cost of the postintervention monitoring phase, our approach yielded an alternative estimate of $176,362.56 (95% CI: 174,962.07, 177,712.45) and an average cost per participant of $1663.80 (95% CI: 1650.59, 1676.53) suggesting both methods yielded similar bottom-line results.

Conclusions

Incorporating bootstrapping into traditional TDABC methodology is feasible and is likely to capture variance in clinical trial data.