The Diagnostic Accuracy of Critical Care Pain Observation Tool (CPOT) in ICU Patients: A Systematic Review and Meta-Analysis

Background. The critical care pain observation tool (CPOT) has been widely used to assess pain in ICU patients, and its validity and reliability have been tested in various contexts. Objective. To determine the diagnostic accuracy of the CPOT in critically ill patients, a systematic review of diagnostic studies was performed. Methods. A database search (PubMed, Medline, CINAHL, ProQuest, EMBASE, Cochrane Library, CNKI, Wanfang, COVIP, CBM) was conducted, as was the manual identification of eligible papers from citations. Eligible studies were published between 2006 and February 2020. Quality appraisal of the studies was carried out with the QUADAS-2 checklist, and data extraction was performed in alignment with STARD 2015. Open Meta Analyst was used for the statistical analysis. Results. In total, 25 articles involving 1920 subjects with 3493 experimental results were included. Most of the studies were of fair quality. A high degree of heterogeneity (1 2 = 57.2%, P < 0.001) was discovered. The Youden index values were 1.56 and 1.55 when the CPOT threshold was set at two and three, respectively. The diagnostic performance of the CPOT was affected by the reference standard. The CPOT had a higher diagnostic odds ratio of 11.52 (95% CI: 7.42-17.87) during nociceptive procedures compared with 9.14 (95% CI: 5.38-15.53) at rest or during non-nociceptive procedures. Conclusion. CPOT has moderate diagnostic parameters with a threshold of two or three, suggesting that it is a fair but not excellent tool. More research on the validity of the CPOT in specific subgroups is needed to broaden its applicability in critical care. (C) 2020 American Academy of Hospice and Palliative Medicine. Published by Elsevier Inc. All rights reserved.