Reliability and validity of the Chinese version of the behavioral pain scale in intubated and non-intubated critically ill patients: Two cross-sectional studies

Abstract

BackgroundSelf-report pain assessment scales may be inappropriate when critically ill patients are incapable of adequate communication because of sedation or mechanical ventilation. The Behavioral Pain Scale (BPS, for intubated patients) and the BPS-non intubated (BPS-NI, for non-intubated patients) measure objective behavioral indicators of pain in non-communicating critically ill patients. ObjectivesTo develop a Chinese version of the BPS combining the original version of the BPS and BPS-NI suitable for pain assessment among critically ill patients and to determine its reliability and validity. DesignTwo cross-sectional studies. SettingsA 15-bed surgical intensive care unit (ICU) in a teaching hospital in Beijing, China. ParticipantsIn the first study, 129 patients (53 intubated and 76 non-intubated) were recruited; in the second study, 83 (43 intubated and 40 non-intubated) were recruited. MethodsThe Chinese version of the BPS (BPS-C) was developed via rigorous translation methods, including double back-translation and content validation involving 13 clinical experts. Internal consistency, discriminative validity, and criterion-related validity were established using the BPS-C and the Numeric Rating Scale (NRS). The BPS-C and NRS were used to assess pain in 53 intubated and 76 non-intubated post-abdominal surgery patients during low pain exposure and increased pain exposure in the first study. To establish interrater reliability, a researcher and a bedside nurse independently performed 172 paired assessments in 43 intubated patients and 160 paired assessments in 40 non-intubated patients with the BPS-C under the same conditions in the second study. ResultsThe BPS-C achieved conceptual and semantic equivalence with the original tool. Internal consistency was established through Cronbach's alpha (α=0.724–0.743 in intubated patients, α=0.701–0.762 in non-intubated patients). Interrater reliability was confirmed through the intraclass correlation coefficients (ICCs), which ranged from 0.962 to 1.000 in both intubated and non-intubated patients with high agreement percentages (95.3–100.0% in intubated and 95.0–100.0% in non-intubated patients). BPS-C scores during increased exposure to pain were significantly higher than those obtained during low exposure to pain, indicating discriminative validity. Criterion-related validity was confirmed by strong positive correlations between BPS-C and NRS scores (Pearson's correlations r=0.815–0.937 for intubated patients, Pearson's correlations r=0.755–0.899 for non-intubated patients). ConclusionsThe Chinese version of the BPS (BPS-C) is appropriate for pain assessment among intubated and non-intubated ICU patients.