Reliability of the surgical Pleth index for assessment of postoperative pain: a pilot study

Postoperative pain control is essential and may have a beneficial effect on postoperative outcome and morbidity. Analgesia quality is controlled using tools such as a Numerical Rating Scale (NRS). These tools require cooperation and often fail in the presence of reduced awareness. The Surgical Pleth index (SPI) has been introduced as a monitoring tool for intraoperative pain under general anaesthesia.

We investigated the correlation between SPI and pain intensity, analgesic consumption and fitness for discharge in the postanaesthesia care unit.

An observational study.

The central postanaesthesia care unit of our tertiary care hospital.

Written informed consent was obtained from 100 patients scheduled for elective surgery under general anaesthesia. Patients below the age of 18 years and those with an abnormal cardiac rhythm were excluded from the study.

Patients were interviewed every 10 min for 2 h.

Pain intensity measured by NRS, discomfort and Aldrete and Post-Anaesthetic Discharge Scoring System (PADSS) scores were noted. SPI and total dose of opioids administered were recorded.

A total of 1300 pain measurements were recorded; 482 (37%) reflected no or mild pain (NRS 0 to 3), 532 (41%) moderate pain (NRS 4 to 6) and 286 (22%) severe pain (NRS 7 to 10). Both NRS (r = 0.62, P < 0.001) and SPI (r = 0.38, P < 0.001) correlated significantly with total opioid consumption. SPI showed a moderate correlation with NRS (r = 0.49, P < 0.001). Receiver operating characteristic analysis showed moderate sensitivity and specificity for discrimination between low and moderate pain (NRS ≤3) (sensitivity 67%, specificity 69% for SPI ≤45), and between moderate and severe pain (NRS >6) (sensitivity 72%, specificity 72% for SPI ≥57). SPI and NRS showed weak negative correlations with Aldrete and PADSS scores.

Sensitivity and specificity of SPI to discriminate between low, moderate and severe pain levels was moderate. Both NRS and SPI correlated significantly with total opioid consumption.