Pain assessment with critical-care pain observation tool (CPOT) in severe brain injury patients——A meta-analysis

Objective To evaluate the validity of pain assessment with Critical-care pain observation tool (CPOT) in severe brain injury patients. Methods Literatures regarding pain assessment with CPOT in severe brain injury patients were systematically searched through the PubMed and other databases. Studies that met the inclusion and exclusion criteria were identified, quality was assessed and data were extracted by two independent investigators. Statistical analysisincluding meta-analysis of mean difference (MD) and diagnostic studieswere performed withRevMan 5.3 and meta-DiSc software. Results A total of 9 published and eligible studies involving 944 cases were included in this meta-analysis. In total of 6 studies, CPOT score had statistical significancein the discriminant validity during non-nociceptive and nociceptive procedure (MD=-2.25, P<0.001). Meanwhile, CPOT had statistical significance before and during nociceptive procedure in total of 8 studies (MD=-2.06, P<0.001). To assess the criterionvalidity of CPOT, with the referencestandard of ″patients′ complaint of pain″ during medical procedures in 4 studies, pooled sensitivity and specificity were 0.739 (95% CI: 0.689-0.785, I2=34.7%) and 0.789 (95% CI: 0.737-0.835, I2=0). The area under AUC curve was 0.831. The heterogeneity analysis of calibration validity showed the existence of threshold effect. Conclusion CPOT has a good discriminant validity and criterion validityforpain assessment during nociceptive procedureamong severe brain injury patients. For the criterion validity of CPOT the cut-off value of pain needs further investigation. Key words: Pain; Assessment; Severe brain injury; Critical care; Critical-care pain observation tool; Meta-analysis