In the Intensive care unit (ICU), pain is usually underestimated due to the difficulty of its assessment in critically ill patients. It can evolve from many sources, for e.g., postoperative surgical incisions, penetrating chest tubes, and even ICU procedures as bedside debridement. It was shown that alleviating pain effectively in both intubated and non-intubated ICU patients has been associated with improved outcomes [1]. Scales such as the Visual analogue scale (VAS) and numeric rating scale (NRS) are used to assess pain intensity postoperatively. In order to use these scales, patients need to be able to understand what is said to them and express themselves. But this cannot be carried out for individuals with communication problems [2]. Also, the use of haemodynamic changes has been demonstrated to be neither valid nor reliable as it is affected by many other aetiologies, and guidelines recommend that vital signs should not be used to evaluate pain in ICU patients [3]. This phenomenon has led to the construction of categorical and numerical methods of pain assessment in critically ill patients. The behavioural pain scale (BPS) -whether intubated or non-intubated forms- has been reported as a valid and reliable tool for pain assessment in ICU patients with recommendations of its use to assess the presence of pain in adult ICU patients when self-reporting is not possible [4]. But unfortunately, the use of Behavioural pain scale non-intubated scoring system (BPS-NI) scale requires sustained efforts to educate and train the ICU team regarding the scale because of its subjective nature [5]. Also, BPS-NI is time-consuming with multiple points of assessment, making it non-practical [6]. The need for simple, non-invasive, rapid, and objective tools for pain evaluation represents a present gap in the literature. The Masimo device could be a promising indirect tool for pain assessment. The Masimo set pulse oximetry system can measure the perfusion index (PI) at the monitored site by calculating the relation between pulsatile and static blood in peripheral tissues. In contrast to the conventional pulse oximeter which measures O2 saturation, Masimo Signal Extraction Technology depends upon the amount of blood at the monitoring site, not upon blood oxygenation. Therefore, PI is considered as an indirect, non-invasive, and continuous measure of peripheral perfusion. It ranges from 0.02% (very weak pulse strength) to 20% (very strong pulse strength) [7]. Pain induces vasoconstriction due to sympathetic nervous system stimulation with a subsequent decrease in PI [8]. This direct relation between pain and sympathetic stimulation raises the hypothesis that the PI can be used as an indirect objective tool for pain assessment. The current study aimed primarily to evaluate the correlation between perfusion index and other clinical indicators of pain after rescue analgesia administration and so detecting its usefulness as an objective indicator of pain assessment in ICU.

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The work was approved by the Ethics committee of Ain Shams University hospital (FMASU R 05/ 2019) on 23/1/2019. The study was prospectively registered with Pan African Clinical Trial Registry (PACTR) with Registration Number PACTR201901839969911 in accordance with WHO and ICMJE standards. Written informed consent was obtained from all subjects or their legal surrogate.

This is a prospective observational study that was conducted in Ain Shams university hospital intensive care unit through the period from January 2019 to October 2019. The study comprised 40 patients. Eligibility criteria for this study included patients with American Society of Anaesthesiologists (ASA) physical status I to III, of either sex, 18-80 years of age, non-sedated non-intubated patients that were admitted postoperatively to ICU after major abdominal surgery. We categorized the participants into two “age groups” according to age: an elderly group (> 60 years) and a young group (< 60 years). Then we further classified each “age group” into a male group and a female group. Now we had a total of 4 groups: an elderly male group (> 60 years) = OM group, an elderly female group (> 60 years) = OF group, a young male group (< 60 years) = YM group and a young female group (< 60 years) = YF group. Exclusion criteria involved patients with fever, hypothermia, history of a neurological, psychiatric, dementia or chronic pain disorder. Patients with unstable haemodynamic status and unconscious patients were also excluded. Patients who had combined general epidural anaesthesia or Transversus abdominis plane block were also excluded.

We studied 40 age-matched patients, 20 males and 20 females, with a mean ± SD age of 48 ± 19 years. The characteristics of the study population and operative details are shown in Table 1.

Table 2 shows a comparison of pain score, PI and other indicators of pain before and after analgesic administration. There was a statistically significant reduction in the BPS-NI score, MAP and heart rate after analgesic administration (P-values, <0.001, 0.039 and 0.001, respectively). On the other hand, there was a statistically significant increase in the PI after analgesic administration (P-value, 0.004). Regarding the difference in axillary temperature, there was no statistically significant difference between the measured axillary temperature at T1 and that at T2 (P-value, 0.442). There was no statistically significant correlation between the PI and the pain score or other clinical indicators of pain either before or after administration of analgesic. There was no statistically significant correlation between the change in PI and the change in pain score or other clinical indicators of pain (Table 3). Studying the correlation of PI with other clinical variables before and after administration of analgesic showed a weak inverse correlation between the PI after administration of analgesic and the RASS score at 1 h (rho, -0.378; P-value, 0.016) and moderate inverse correlation between the change in PI and the RASS score at 1 h (rho, -0.409; P-value, 0.009). There was no statistically significant relationship between the age or sex and the PI either before or after the administration of analgesic. Neither there was a statistically significant relationship between the change in PI and the age or sex (Table 4).

Table 5 shows the results of multivariable regression analysis for the effect of age or sex on the change in PI (Δ PI) with adjustment for other confounding factors. After adjustment for the effect of American Society of Anaesthesiologists - Physical Status, operative time and intraoperative opioid consumption, there was no statistically significant relationship between the change in PI and the patient’s age (P-value, 0.269) and sex (P-value, 0.482).

In our study, there was a statistically significant increase in the PI after analgesic administration. Also, there was a statistically significant reduction in the BPS-NI score, MAP and heart rate after analgesic administration. But, we did not find any statistically significant correlation between the absolute value of “PI and other examined clinical pain indicators (the BPS-NI, MAP, and HR)” before or after rescue analgesia administration as well as with their changes. The relationship between analgesia and PI is the basis of our hypothesis in this study. The PI is a non-invasive and easy method that can be used for evaluating pain and monitoring the effectiveness of analgesia. It can also eliminate psychological factors such as fear, anxiety, depression, and anger [11]. This benefit can be more valid in patients suffering from cognitive impairment and dementia especially because common pain behaviour scales are very difficult, require training of the ICU staff and are time-consuming [11]. There are multiple studies exploring the relationship between PI and pain, whether in awake patients [12, 13] or those under general anaesthesia [8]. And they all proved that PI decreased due to painful stimulus. On the other hand, other studies explored the relationship between PI and analgesia whether under general anaesthesia [2, 11, 14] or epidural analgesia [15] or transforaminal block [16]. And they all proved that PI increased after analgesic administration. All of these studies explored different types of pain as postoperative surgical pain [2, 11, 14, 15], intensive care procedural pain [13], electric stimulation pain [8, 12] and finally, chronic radicular pain [16].

In agreement with the current study, Tapar and colleagues [2] showed that there was a statistically significant difference between pre-analgesic and post-analgesic PI, VAS scores and haemodynamics with no correlation between PI absolute values & VAS scores absolute values at pre- and post-analgesic measurements. Also, there was a detected weak negative correlation between the change in PI and the change of pain score (VAS score). It was a prospective observational study that was done on 89 patients that had undergone minor to moderate surgical procedures and were observed in Post Anesthesia Care Unit (PACU) postoperatively. They used morphine increments for post-operative analgesia and the subjective pain score used was VAS score. Another study confirming our findings was carried out by Mohammed and colleagues [11], in which a Masimo pulse co-oximetry perfusion index was attached to 70 American Society of Anaesthesiologists-Physical Status I adult patients at PACU, who underwent lumbar spine discectomy. The PI was significantly higher at post-analgesic timing than at pre-analgesic timing. This increase was associated with a statistically significant decrease in other measured parameters. This means that the PI increases with adequate relief from pain, as indicated by a decrease in VAS, HR, and MAP. A decrease in VAS was associated with an increase in PI, but the correlation was not statistically significant. Also, the correlation between change in PI and change in VAS score & change in MAP was not statistically significant and this is consistent with our study. It is to be noted that there was a statistically significant negative correlation between change in HR and change in PI. For all patients, analgesia was achieved with i.v. morphine and i.v. 1 g paracetamol and subjective pain scale used was the VAS score. In correspondence to the current study, Nishimura and colleagues [12] studied the changes in perfusion index in response to noxious electrical stimulation in awake healthy subjects. They measured the PI and pulse rate in 70 healthy volunteers exposed to increasing electrical stimulation until they reached their pain tolerance threshold. They observed a significantly decreased PI in response to electrical stimulation but with no increase in the pulse rate due to its very small intensity. They concluded that the PI may be an independent parameter reflecting the perception of noxious stimuli and offers a non-invasive option for objectively evaluating pain perception. Finally, in a study done by Hasanin and colleagues [13], they reported a difference between PI values, Systolic blood pressure, Diastolic blood pressure, HR, and pain intensity before and after the pain created by positioning in ICU patients. BPS-NI has been used for subjective pain assessment especially as all patients were sedated (but not-intubated), which might affect their communication with the medical staff. There was a significant increase in the Systolic blood pressure, Diastolic blood pressure, heart rate and BPS-NI post-positioning values compared with pre-positioning values. Also, a significant decrease in PI was also observed at post-positioning values compared with pre-positioning values. Also, no correlation was found between the PI values and any other variable (Systolic blood pressure, Diastolic blood pressure, HR, and BPS-NI) before or after the patient positioning. Hasanin’s study differs from the current study in that the change in BPS-NI showed a good correlation with the change in PI. On the other hand, there are two studies which showed a weak correlation between different parameters. In a retrospective observational study done by Chu and colleagues [14], the correlation between the PI and VAS score together with their delta change and their percentage change showed weak correlations. They enrolled 80 female patients postoperatively, with a different age range, who were observed in PACU before and after intravenous morphine analgesic administration. The second study was done by Kupeli & Kulhan [15]. They investigated the relationship between labour pain level and PI in 30 women undergoing spontaneous vaginal delivery under epidural analgesia. They noticed that upon activation of the epidural blockade with 10 mL 0.25% bupivacaine, the PI increased. Also, they noticed a gradual decrease in PI with a fade of epidural analgesia (manifested by a gradual increase in labor pain). They concluded that PI could offer a non-invasive option to objectively assess pain perception and this is in accordance with our study findings. But in opposition, there was a significant negative association between PI and VAS absolute values at the 10th, 30th, 60th minutes and 2nd hour after epidural blockade activation. Also, there was a significant negative association between PI and HR absolute values before the procedure and at the time of administration of epidural analgesia and 5 minutes later. They noted that perfusion index had no significant correlation with both systolic and diastolic blood pressures.

Few studies had explored the effect of age or sex on the change in PI after analgesic administration or painful stimulation. In the current study, there was no statistically significant relationship between the age or sex and the PI either before or after the administration of analgesic. Neither was there a statistically significant relationship between the change in PI and the age or sex. Supporting our findings, Chu and colleagues [14] stated the same findings in their study. On the other hand, Nishimura and colleagues [12] observed that the old women group did not show any changes in PI before or after electrical stimulation when compared to other age and sex groups that showed a decrease in PI.

PI measurements are very sensitive to patients’ movements. The rapid fluctuation and sensitivity of PI are its weakness as well as strength in the clinical field. To compensate for this limitation, PI monitoring should be done after ensuring position stability.

Perfusion index can be added to other indicators of pain assessment in ICU. It is easy, non-invasive, free of subjective interpretation, less time-consuming and finally, not affected by age or sex related factors.