Waiting for Dr. Godot: how much and who responds to predicted health care wait times?

TL;DR

This study examines how publicly disclosed predicted wait times influence patient demand in emergency departments, revealing that longer wait times decrease patient numbers and alter patient acuity levels, with effects varying at different wait durations.

Contribution

It provides empirical evidence on the impact of wait time disclosures on patient behavior using a regression discontinuity design and impulse response analysis.

Findings

30-minute increase in wait time reduces ED demand by 2% within 3 hours.

Longer wait times lead to fewer low-acuity patients seeking emergency care.

At very high wait times, even sick patients reduce ED visits.

Abstract

Asymmetric information in healthcare implies that patients could have difficulty trading off non-health and health related information. I document effects on patient demand when predicted wait time is disclosed to patients in an emergency department (ED) system. I use a regression discontinuity where EDs with similar predicted wait times display different online wait times to patients. I use impulse response functions estimated by local projections to demonstrate effects of the higher wait time. I find that an additional thirty minutes of wait time results in 15% fewer waiting patients at urgent cares and 2% fewer waiting patients at EDs within 3 hours of display. I find that the type of patient that stops using emergency care is triaged as having lower acuity and would have used an urgent care. However, I find that at very high wait times there are declines in all acuity patients including sick patients.