Impact of Waiting Time on Patient Flow: An Econometric Analysis of Outpatient Clinics (joint work with Diwas KC)
Research Seminars
Academic Areas Operations Management
Nikolay Osadchiy, Ph.D. Goizueta Business School Emory University
June 14, 2013
| 12:30 PM - 2:00 PM | Friday
AC-2 MLT, hyderabad, India
For ISB Community
Abstract:
We examine the effect of waiting time to appointment on patient throughput in an outpatient clinic setting. Using a novel data set consisting of allotted capacity, scheduled appointments, and patient flow from over 1400 healthcare providers for a period of two years, we conduct a detailed econometric analysis of waiting time, patient choice, and aggregate throughput. We find that waiting time affects throughput in two ways - by reducing the likelihood that a patient books an appointment, and by reducing the likelihood that a patient shows up for a scheduled appointment. To estimate the magnitudes of these effects, we apply a state of the art non-parametric model of patient choice, which generates a distribution of patient patience (or willingness to wait) for an appointment. This in turn allows us to quantify the effect of wait times on lost sales. Next, we estimate the effect of wait times on the patient no-show rate. In general, this estimation poses an econometric challenge because the observed sample represents only patients who book appointments. To circumvent the selection bias, we augment our model of patient choice with our previously estimated patient patience parameters. Overall, we find that a 10% increase in capacity increases throughput on average by 4.2%. 85% of the throughput gain is attributed to the reduction of lost sales and 15% is accounted for by the reduction of no-shows. The tools and the empirical results presented in the paper provide hospital administrators and policy makers with a better understanding of the effects of capacity constraints and waiting times on throughput and access to care.
We examine the effect of waiting time to appointment on patient throughput in an outpatient clinic setting. Using a novel data set consisting of allotted capacity, scheduled appointments, and patient flow from over 1400 healthcare providers for a period of two years, we conduct a detailed econometric analysis of waiting time, patient choice, and aggregate throughput. We find that waiting time affects throughput in two ways - by reducing the likelihood that a patient books an appointment, and by reducing the likelihood that a patient shows up for a scheduled appointment. To estimate the magnitudes of these effects, we apply a state of the art non-parametric model of patient choice, which generates a distribution of patient patience (or willingness to wait) for an appointment. This in turn allows us to quantify the effect of wait times on lost sales. Next, we estimate the effect of wait times on the patient no-show rate. In general, this estimation poses an econometric challenge because the observed sample represents only patients who book appointments. To circumvent the selection bias, we augment our model of patient choice with our previously estimated patient patience parameters. Overall, we find that a 10% increase in capacity increases throughput on average by 4.2%. 85% of the throughput gain is attributed to the reduction of lost sales and 15% is accounted for by the reduction of no-shows. The tools and the empirical results presented in the paper provide hospital administrators and policy makers with a better understanding of the effects of capacity constraints and waiting times on throughput and access to care.
