Discharge recommendation based on a novel technique of homeostatic analysis

Abstract

Objective: We propose a computational framework for integrating diverse patient measurements into an aggregate health score and applying it to patient stability prediction.

Materials and methods: We mapped retrospective patient data from the Multiparameter Intelligent Monitoring in Intensive Care (MIMIC) II clinical database into a discrete multidimensional space, which was searched for measurement combinations and trends relevant to patient outcomes of interest. Patient trajectories through this space were then used to make outcome predictions. As a case study, we built AutoTriage, a patient stability prediction tool to be used for discharge recommendation.

Results: AutoTriage correctly identified 3 times as many stabilizing patients as existing tools and achieved an accuracy of 92.9% (95% CI: 91.6-93.9%), while maintaining 94.5% specificity. Analysis of AutoTriage parameters revealed that interdependencies between risk factors comprised the majority of each patient stability score.

Discussion: AutoTriage demonstrated an improvement in the sensitivity of existing stability prediction tools, while considering patient safety upon discharge. The relative contributions of risk factors indicated that time-series trends and measurement interdependencies are most important to stability prediction.

Conclusion: Our results motivate the application of multidimensional analysis to other clinical problems and highlight the importance of risk factor trends and interdependencies in outcome prediction.

Keywords: clinical decision support systems; computer-assisted diagnosis; length of stay; medical informatics; patient discharge.

© The Author 2016. Published by Oxford University Press on behalf of the American Medical Informatics Association. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

Figures

Figure 1.

Representative schematic of S2-type and D2-type analyses. (A) Sample patient physiological data. (B) Heart rate (HR) and systolic blood pressure (SysBP) coevolve over time and (C) ΔHR (change or trend in HR) and ΔSysBP (trend in SysBP) coevolve over time. In panel (A), observations of HR and SysBP, as well as their hourly changes, are tabulated and normalized according to a heuristic lookup table (bolded numbers in parentheses). These normalized values are then mapped into finite discrete hyper-dimensional space as time-parametrized curves in panels (B) and (C).

Figure 2.

Discharge prediction comparison receiver operating characteristic (ROC) curve of AutoTriage, MET, and MEWS. The clinical operating range, which is the region with an acceptable discharge error rate, is highlighted in yellow., At an erroneous discharge rate of 5.5% (vertical dotted line), stable patient discharge rates of AutoTriage, MET, and MEWS are 0.578, 0.185, and 0.141, respectively.

Figure 3.

Average percent contribution of each measurement type on patient score. S1 are individual measurements and D1 are the trends of individual measurements. S2 are the groupings of 2 measurements and D2 are the equivalent groupings of 2 trends. 95% confidence interval error bars are shown.