Analgesia nociception index for the assessment of pain in critically ill patients: a diagnostic accuracy study

Background. Behavioural pain tools are used in Intensive Care Unit (ICU) patients unable to self-report their pain-intensity but need sustained efforts to educate and train the ICU team because of the subjective nature of these clinical tools. This study measured the validity and performance of an electrophysiological monitoring tool based on the spectral analysis of heart rate variability, the Analgesia Nociception Index (ANI) which varies from 0 (minimal parasympathetic tone, maximal stress-response and pain) to 100 (maximal parasympathetic tone, minimal stress-response and pain). Methods. Mean-ANI (ANIm) and Instant-ANI (ANIi) were continuously recorded then compared with the Behavioral Pain Scale (BPS) before, during and after routine care procedures in critically-ill non-comatose patients. Results. 969 assessments were performed in 110 patients. ANIi was the most discriminative pain tool. It was significantly correlated with BPS (r=-0.30; 95% CI-0.37 to -0.25; P<0.001). For an ANIi threshold of 42.5, the sensitivity, specificity, positive and negative predictive values were respectively 61.4%, 77.4%, 37.0%, and 90.4%. Compared with the BPS, ANIi had no significantly different ability to change during turning and tracheal-suctioning but changed significantly more during dressing change. ANIi increased independently with age, obesity and severity of illness, and controlled mechanical-ventilation, vasopressors use and analgesia. ANIi decreased independently when vigilance status and respiratory rate increased. ANIm demonstrated poor psychometric properties to detect pain. Conclusions. Despite low sensitivity/specificity, ANIi >= 43 had a Negative-Predictive-Value of 90%. Hence ANIi may be of highest benefit for excluding significant pain. A randomized controlled trial should compare sedation-analgesia protocols based on ANIi to presently recommended behavioural-pain-tools.