Real-time prediction of mortality, readmission, and length of stay using electronic health record data

Abstract

Objective: To develop a predictive model for real-time predictions of length of stay, mortality, and readmission for hospitalized patients using electronic health records (EHRs).

Materials and methods: A Bayesian Network model was built to estimate the probability of a hospitalized patient being "at home," in the hospital, or dead for each of the next 7 days. The network utilizes patient-specific administrative and laboratory data and is updated each time a new pathology test result becomes available. Electronic health records from 32 634 patients admitted to a Sydney metropolitan hospital via the emergency department from July 2008 through December 2011 were used. The model was tested on 2011 data and trained on the data of earlier years.

Results: The model achieved an average daily accuracy of 80% and area under the receiving operating characteristic curve (AUROC) of 0.82. The model's predictive ability was highest within 24 hours from prediction (AUROC = 0.83) and decreased slightly with time. Death was the most predictable outcome with a daily average accuracy of 93% and AUROC of 0.84.

Discussion: We developed the first non-disease-specific model that simultaneously predicts remaining days of hospitalization, death, and readmission as part of the same outcome. By providing a future daily probability for each outcome class, we enable the visualization of future patient trajectories. Among these, it is possible to identify trajectories indicating expected discharge, expected continuing hospitalization, expected death, and possible readmission.

Conclusions: Bayesian Networks can model EHRs to provide real-time forecasts for patient outcomes, which provide richer information than traditional independent point predictions of length of stay, death, or readmission, and can thus better support decision making.

Keywords: length of stay; mortality; patient outcome; prediction; readmission.

Figure 1:

An illustration of how the predictive model is updated following the availability of one or more pathology test results in the EHR system. Patient outcomes comprise the probability of staying at hospital, being “at home” (ie, having been discharged alive), or being dead during each of the 7 days following a temporal event. Abbreviation: CT, computed tomography.

Figure 2:

Overview of how the predictive model was built in 5 steps.

Figure 3:

The Bayesian Network model includes 7 target days (in yellow), selected primary features (in purple), and selected secondary features (in blue) as nodes, representing random variables. Arcs were created from each of the target days to their corresponding selected primary features, and from these primary features to their corresponding selected secondary features. Rectangular nodes represent dynamic variables while elliptical nodes represent static variables. Abbreviations: U. Sodium, urine sodium; U. Potassium, urine potassium; HCT, hematocrit; WBC, white blood cell count; Hgb, hemoglobin; UEC, urea, electrolytes, creatinine; Alk. Phos., alkaline phosphatase; pH, potential hydrogen; CRP, C-reactive protein; RBC, red blood cell count; APTT, activated partial thromboplastin time; Tot. Protein, total protein; ED arrival, mode of arrival to emergency department; Triage, triage category; Prev. LOS, cumulative length of stay in previous hospitalizations; Inorg Phos, inorganic phosphate; Test Count, number of laboratory tests performed so far during hospitalization; HOS days, days already in the hospital; Hours since HOS, hours since previous hospitalization.

Figure 4:

Model predictions for selected individual patients. Each line represents a future daily probability of being in the hospital (red), “at home” (blue), or dead (green). Days 1 to 7 represent the future consecutive days relative to the time of prediction. The dotted vertical lines indicate true events. Patients and times of prediction have been randomly selected among examples for which the model correctly classifies patient outcomes for all, or most of, the 7 days. Panel A shows a typical prediction of expected continuing hospitalization; Panel B illustrates a prediction of expected discharge. Panels C and D are typical predictions of expected death, and Panel E predicts possible readmission.