Predicting Surgery Duration with Neural Heteroscedastic Regression

Nathan H Ng, Rodney A Gabriel, Julian McAuley, Charles Elkan, Zachary C Lipton
Proceedings of the 2nd Machine Learning for Healthcare Conference, PMLR 68(26):100-111, 2017.

Abstract

Scheduling surgeries is a challenging task due to the fundamental uncertainty of the clinical environment, as well as the risks and costs associated with under- and over-booking. We investigate neural regression algorithms to estimate the parameters of surgery case durations, focusing on the issue of heteroscedasticity. We seek to simultaneously estimate the duration of each surgery, as well as a surgery-specific notion of our uncertainty about its duration. Estimating this uncertainty can lead to more nuanced and effective scheduling strategies, as we are able to schedule surgeries more efficiently while allowing an informed and case-specific margin of error. Using surgery records from a large United States health system we demonstrate potential improvements on the order of 20% (in terms of minutes overbooked) compared to current scheduling techniques. Moreover, we demonstrate that surgery durations are indeed heteroscedastic. We show that models that estimate case-specific uncertainty better fit the data (log likelihood). Additionally, we show that the heteroscedastic predictions can more optimally trade off between over and under-booking minutes, especially when idle minutes and scheduling collisions confer disparate costs.

Cite this Paper

BibTeX
@InProceedings{pmlr-v68-ng17a, title = {Predicting Surgery Duration with Neural Heteroscedastic Regression}, author = {Ng, Nathan H and Gabriel, Rodney A and McAuley, Julian and Elkan, Charles and Lipton, Zachary C}, booktitle = {Proceedings of the 2nd Machine Learning for Healthcare Conference}, pages = {100--111}, year = {2017}, editor = {Doshi-Velez, Finale and Fackler, Jim and Kale, David and Ranganath, Rajesh and Wallace, Byron and Wiens, Jenna}, volume = {68}, number = {26}, series = {Proceedings of Machine Learning Research}, month = {18--19 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v68/ng17a/ng17a.pdf}, url = {https://proceedings.mlr.press/v68/ng17a.html}, abstract = {Scheduling surgeries is a challenging task due to the fundamental uncertainty of the clinical environment, as well as the risks and costs associated with under- and over-booking. We investigate neural regression algorithms to estimate the parameters of surgery case durations, focusing on the issue of heteroscedasticity. We seek to simultaneously estimate the duration of each surgery, as well as a surgery-specific notion of our uncertainty about its duration. Estimating this uncertainty can lead to more nuanced and effective scheduling strategies, as we are able to schedule surgeries more efficiently while allowing an informed and case-specific margin of error. Using surgery records from a large United States health system we demonstrate potential improvements on the order of 20% (in terms of minutes overbooked) compared to current scheduling techniques. Moreover, we demonstrate that surgery durations are indeed heteroscedastic. We show that models that estimate case-specific uncertainty better fit the data (log likelihood). Additionally, we show that the heteroscedastic predictions can more optimally trade off between over and under-booking minutes, especially when idle minutes and scheduling collisions confer disparate costs.} }
Endnote
%0 Conference Paper %T Predicting Surgery Duration with Neural Heteroscedastic Regression %A Nathan H Ng %A Rodney A Gabriel %A Julian McAuley %A Charles Elkan %A Zachary C Lipton %B Proceedings of the 2nd Machine Learning for Healthcare Conference %C Proceedings of Machine Learning Research %D 2017 %E Finale Doshi-Velez %E Jim Fackler %E David Kale %E Rajesh Ranganath %E Byron Wallace %E Jenna Wiens %F pmlr-v68-ng17a %I PMLR %P 100--111 %U https://proceedings.mlr.press/v68/ng17a.html %V 68 %N 26 %X Scheduling surgeries is a challenging task due to the fundamental uncertainty of the clinical environment, as well as the risks and costs associated with under- and over-booking. We investigate neural regression algorithms to estimate the parameters of surgery case durations, focusing on the issue of heteroscedasticity. We seek to simultaneously estimate the duration of each surgery, as well as a surgery-specific notion of our uncertainty about its duration. Estimating this uncertainty can lead to more nuanced and effective scheduling strategies, as we are able to schedule surgeries more efficiently while allowing an informed and case-specific margin of error. Using surgery records from a large United States health system we demonstrate potential improvements on the order of 20% (in terms of minutes overbooked) compared to current scheduling techniques. Moreover, we demonstrate that surgery durations are indeed heteroscedastic. We show that models that estimate case-specific uncertainty better fit the data (log likelihood). Additionally, we show that the heteroscedastic predictions can more optimally trade off between over and under-booking minutes, especially when idle minutes and scheduling collisions confer disparate costs.
APA
Ng, N.H., Gabriel, R.A., McAuley, J., Elkan, C. & Lipton, Z.C.. (2017). Predicting Surgery Duration with Neural Heteroscedastic Regression. Proceedings of the 2nd Machine Learning for Healthcare Conference, in Proceedings of Machine Learning Research 68(26):100-111 Available from https://proceedings.mlr.press/v68/ng17a.html.

Related Material