When you can't tell when it hurts: a preliminary algorithm to assess pain in patients who can't communicate

Abstract

Pain is a common but significant problem that is considered a high priority area of care. Although there are many pain assessment scales that can be applied to patients who can communicate, either verbally or non-verbally, pain assessment for minimally responsive patients is limited. In this preliminary work, we developed a novel approach for assessing pain in such patients using a principal component analysis (PCA)-based local detector. Our algorithm produce a single index to indicate the increase in pain level based on unsynchronized, sparse and noisy time series data collected from electronic flowsheets. Among 8032 patient cases collected, 53 cases that satisfied the data requirements for PCA were used in this experiment. Our preliminary results indicate high potential in this approach by yielding an average AUC of 0.76 for the 53 cases.

Figure 1.

Singular Value Decomposition (SVD) for observations in a given period A^(t)

Figure 2.

Calculated projection residual

Figure 3.

Workflow of the record selection and feature selection procedures.

Figure 4.

Sample ROC curves and the box plot of the 53 AUC values

Figure 5.

Transition of c (artificial timestamps are used for patient privacy)