Translation, validation and comparison of three behavioural pain assessment tools for patients who cannot communicate verbally
Aim The thesis set out to examine validations of three observational pain assessment tools and establish nurses’ expectations of them and the factors that might influence them within intensive care unit (ICU) settings. Background The guidelines to pain assessment specific to ICU patients have been of great interest to health professionals over the last 20 years. Pain assessment remains a challenge for most ICU patients due to the difficulty of assessing pain with any precision. Evidence suggests that the Behavioural Pain Scale (BPS) and Critical- Care Pain Observation Tool (CPOT) have demonstrated sound psychometric properties. A review of the relevant literature highlighted the fact that no such studies have yet been conducted with a similar homogenous group in Asia. The Wong-Baker Face Pain Rating Scale (FPRS) is currently widely used for nonverbally communicating patients (NVCPs) with pain in ICU settings, and is even recommended for use with children. Valid assessment tools are required for effective pain assessment in ICU settings, particularly in patients who are experiencing communication difficulties. Design An embedded mixed methods design was employed to: 1) translate Chinese versions of BPS and CPOT, 2) test their validity and reliability of in comparison with FPRS, and 3) establish the nurses’ expectations about the three study scales when undertaking pain assessment by using semi-structured focus group interviews. Methods This thesis initially reviews the literature available to select the most appropriate scales for assessing pain in critically-ill NVCPs. The selected scales were then translated into a Traditional-Chinese version using established procedures for the Taiwanese context. Evaluations of the three pain scales were gathered using quantitative measures of pain scores in NVCPs experiencing painless/painful interventions. These were further compared with a few focus groups to establish the feasibility and utility of the three pain scales. The psychometric properties of the pain scales were assessed for reliability by using internal consistency and inter-rater agreement) and for validity by using content validity, concurrent validity, discriminant validity, and responsiveness. The validity was evaluated using ANOVA to compare the changes between the different procedures. The significance level was set at 0.05. As for the analysis of the qualitative data, this study typically follows the path of aggregating the words into themes of information and presenting the diversity of ideas gathered during the data collection. Results For the 2068 observations in 237 patients, there were no statistical differences between the characteristics of the BPS, CPOT, and FPRS groups. Validity was demonstrated by changes from baseline in the scores of the three groups, which were significantly higher during suction (p < 0.001). In regard to the result for the criterion validity, both BPS and CPOT had moderate positive correlations with FPRS. The internal consistency was excellent; the Cronbach’s α was 0.700 for BPS and 0.821 for CPOT when all items were included. The majority of nurses preferred to use BPS to assess pain in their clinical practice. When the nurses were asked how long they needed and how easy they found it to complete the assessments using these tools, they all agreed that each patient assessments were easier and took the least time when they used FRPS. However, the nurses considered that the most effective pain reaction during nociceptive procedures had been assessed by using BPS. Even though all of the participant nurses stated that CPOT provides a detailed item-description about pain behaviour, it also provided the biggest obstacle to use because of its ambiguous indicators. Conclusions BPS, CPOT and FRPS provide potentially useful measurement scales for assessing pain in ICU NVCPs. However, judging from the inconsistencies between the nurses’ replies, the results could reflect a conflict between the need to use a validated measure of pain for NVCPs on the one hand and managing a heavy workload in the ICU on the other. This study opens up an avenue for investigating further the link between the underlying conceptions of pain behaviour and the effectiveness of pain assessments in NVCPs when using an objective pain measurement.