ABSTRACT
OBJECTIVES: Evaluation of pain in critically ill intubated patients is difficult and subjective. This study aimed to evaluate the accuracy of oximetry-derived peripheral perfusion index (PPI) in pain assessment in critically ill intubated patients using the behavioral pain scale (BPS) as a reference. MATERIALS AND METHODS: This prospective observational study included 35 adult mechanically ventilated surgical patients during their first 2 postoperative days in the intensive care unit. Values of PPI, BPS, Richmond Agitation Sedation Scale (RASS), heart rate, and blood pressure were obtained before and after a standard painful stimulus (changing the patient position) and the ratio between the second and the first reading was calculated to determine the change (Δ) in all variables. The outcomes were the correlation between ΔBPS and ΔPPI as well as other hemodynamic parameters. The ability of the PPI to detect pain (defined as BPS ≥6) was analyzed using the area under receiver operating characteristic curve. RESULTS: Paired readings were obtained from 35 patients. After the standard painful stimulus, the PPI decreased while the BPS and the Richmond agitation sedation scale increased. The Spearman correlation coefficient (95% confidence interval) between Δ PPI and Δ BPS was 0.41 (0.09-0.65). PPI values showed poor accuracy in detecting pain with area under receiver operating characteristic curve (95% confidence interval): 0.65 (0.53-0.76), with best cutoff value of ≤2.7. CONCLUSION: The PPI decreased after application of a standard painful stimulus in critically ill intubated patients. ∆PPI showed a low correlation with ∆BPS, and a PPI of ≤2.7 showed a low ability to detect BPS ≥6. Therefore, PPI should not be used for pain evaluation in critically ill intubated surgical patients.