Patrick J Tighe1, Paul Nickerson, Roger B Fillingim, Parisa Rashidi. 1. *Pain Research and Intervention Center of Excellence (PRICE) †Department of Anesthesiology ‡The J. Crayton Pruitt Family, Department of Biomedical Engineering, University of Florida §Department of Community Dentistry and Behavioral Science, University of Florida College of Dentistry, Gainesville, FL.
Abstract
OBJECTIVES: The primary aim was to characterize the temporal dynamics of postoperative pain intensity using symbolic aggregate approximation (SAX). The secondary aim was to explore the effects of sociodemographic and clinical factors on the SAX representations of postoperative pain intensity. MATERIALS AND METHODS: We applied SAX to a large-scale time series database of 226,808 acute postoperative pain intensity ratings. Pain scores were stratified by patient age, sex, type of surgery, home opioid use, and postoperative day (POD), and costratified by age and sex. Cosine similarity, a metric that measures distance using vector angle, was applied to these motif data to compare pain behavior similarities across strata. RESULTS: Across age groups, SAX clusters revealed a shift from low-to-low pain score transitions in older patients to high-to-high pain score transitions in younger patients, whereas analyses stratified by sex showed that males had a greater focus of pain score transitions among lower-intensity pain scores compared with females. Surgical stratification, using cardiovascular surgery as a reference, demonstrated that pulmonary surgery had the highest cosine similarity at 0.855. With POD stratification, POD 7 carried the greatest cosine similarity to POD 0 (0.611) after POD 1 (0.765), with POD 3 (0.419) and POD 4 (0.441) carrying the lowest cosine similarities with POD 0. DISCUSSION: SAX offers a feasible and effective framework for characterizing large-scale postoperative pain within the time domain. Stratification of SAX representations demonstrate unique temporal dynamic profiles on the basis of age group, sex, type of surgery, preoperative opioid use, and across PODs 1 to 7.
OBJECTIVES: The primary aim was to characterize the temporal dynamics of postoperative pain intensity using symbolic aggregate approximation (SAX). The secondary aim was to explore the effects of sociodemographic and clinical factors on the SAX representations of postoperative pain intensity. MATERIALS AND METHODS: We applied SAX to a large-scale time series database of 226,808 acute postoperative pain intensity ratings. Pain scores were stratified by patient age, sex, type of surgery, home opioid use, and postoperative day (POD), and costratified by age and sex. Cosine similarity, a metric that measures distance using vector angle, was applied to these motif data to compare pain behavior similarities across strata. RESULTS: Across age groups, SAX clusters revealed a shift from low-to-low pain score transitions in older patients to high-to-high pain score transitions in younger patients, whereas analyses stratified by sex showed that males had a greater focus of pain score transitions among lower-intensity pain scores compared with females. Surgical stratification, using cardiovascular surgery as a reference, demonstrated that pulmonary surgery had the highest cosine similarity at 0.855. With POD stratification, POD 7 carried the greatest cosine similarity to POD 0 (0.611) after POD 1 (0.765), with POD 3 (0.419) and POD 4 (0.441) carrying the lowest cosine similarities with POD 0. DISCUSSION: SAX offers a feasible and effective framework for characterizing large-scale postoperative pain within the time domain. Stratification of SAX representations demonstrate unique temporal dynamic profiles on the basis of age group, sex, type of surgery, preoperative opioid use, and across PODs 1 to 7.
Authors: M Divella; M Cecconi; N Fasano; N Langiano; M Buttazzoni; I Gimigliano; G Della Rocca Journal: Minerva Anestesiol Date: 2012-02-10 Impact factor: 3.051
Authors: Paul V Nickerson; Raheleh Baharloo; Amal A Wanigatunga; Todd M Manini; Patrick J Tighe; Parisa Rashidi Journal: IEEE J Biomed Health Inform Date: 2017-05-16 Impact factor: 5.772
Authors: Cameron R Smith; Raheleh Baharloo; Paul Nickerson; Margaret Wallace; Baiming Zou; Roger B Fillingim; Paul Crispen; Hari Parvataneni; Chancellor Gray; Hernan Prieto; Tiago Machuca; Steven Hughes; Gregory Murad; Parisa Rashidi; Patrick J Tighe Journal: Eur J Pain Date: 2020-12-04 Impact factor: 3.931