회복실 비언어적 통증 사정을 위한 PAINAD-K척도

Abstract

Purpose Pain is called the fifth vital sign, and self-reporting by the patient is considered the gold standard for pain evaluation. However, accurate pain assessment cannot be performed according to self-reporting by the patient due to the patient’s altered consciousness and changes in cognitive functions such as delirium emergence in the post-anesthesia care unit (PACU). This exploratory study aimed to verify the reliability and validity of the Pain Assessment in Advanced Dementia Scale-Korean version (PAINAD-K) for assessing pain among patients who were unable to communicate in the PACU.

Method The study included 205 patients aged >65 years who were admitted to the PACU following general surgery at a general hospital. Data were collected by the primary author and trained nurse practitioners between December 2021 and July 2022. Internal consistency, scale stability, and inter-observer reliability were assessed. The stability of the scale was assessed using the test-retest reliability of 91 patients who were not administered analgesics. Sensitivity, specificity, and concurrent validity were calculated to examine criterion validity. To test concurrent validity, a numerical rating scale (NRS) was used for verbally communicative patients, while the Algoplus scale was used for patients who were unable to communicate. To evaluate sensitivity and specificity, the patients were divided according to the presence or absence of pain. Convergent and discriminant validity were tested to evaluate the construct validity of the PAINAD-K. The Korean version of the Critical Care Pain Observation Tool (CPOT-K) was used for patients who were unable to communicate to test the convergent validity of the PAINAD-K. The collected data were analyzed using Cronbach’s α, intraclass correlation coefficient (ICC), Cohen’s kappa, spearman rank correlation test, receiver operating characteristic (ROC) curve analysis, and repeated- measures two-way analysis of variance (ANOVA) using SPSS 27.0.

Result The Cronbach’s α values for the PAINAD-K were 0.784. The ICC values ranged from 0.951 to 0.983. The weighted kappa coefficients ranged from 0.906 to 0.966. The test-retest stability of the scale was 0.736 for verbally communicative patients and 0.556 for patients who were unable to communicate, and the overall stability was 0.636. These results indicate good internal consistency, inter-rater reliability, and scale stability. ROC curves were drawn to determine cutoff points. In verbally communicative patients, using the PAINAD-K and NRS, the cutoff was 2.5 points, and the area under the curve (AUC) was 0.952 (sensitivity = 84.3%, specificity = 97.0%) at the baseline pain assessment. Using the PAINAD-K and Algoplus in patients who were unable to communicate, the cutoff was 1.5 points and the AUC was 0.995 (sensitivity = 98.4%, specificity = 97.5%) at the baseline pain assessment, and the cutoff was 1.5 points and the AUC was 0.952 (sensitivity = 96.4%, specificity = 93.2%) at the 30-minute follow-up. These results demonstrate the feasibility of using the PAINAD-K in patients in the PACU who are unable to communicate. For concurrent validity, the correlations of the PAINAD-K with the NRS at baseline and 30-minute follow-up were 0.817 and 0.399, respectively, in verbally communicative patients. The correlations of the PAINAD-K with Algoplus at the baseline and 30-minute follow-up were 0.932 and 0.882, respectively, demonstrating a high degree of correlation with the gold standard for measuring pain level in patients in the PACU who were unable to communicate. These results suggest the criterion validity of the PAINAD-K. Regarding convergent validity in patients who were unable to communicate, the correlations of the PAINAD-K with the CPOT-K at the baseline and 30-minute points were 0.694 and 0.782, respectively, in patients with pain, and the correlations at the baseline and 30-minute time point were 0.881 and 0.957, respectively, in patients without pain. Repeated-measures two-way ANOVA showed discriminant validity with a significant interaction effect between time (baseline vs. 30 minutes) and group (with vs. without pain medication). These results support the construct validity of the two measures in representing the same construct.

Conclusion The findings of this study suggest that the PAINAD-K is a psychometrically valid scale for assessing pain in patients in the PACU who are aged >65 years and unable to communicate. PACU nurses should ensure optimal pain management by selecting an appropriate pain assessment scale, depending on whether the patient is able to communicate.