Analgesia nociception index and systemic haemodynamics during anaesthetic induction and tracheal intubation: A secondary analysis of a randomised controlled trial.

BACKGROUND AND AIMS: Direct laryngoscopy and tracheal intubation is a noxious stimulation that induces significant stress response. Currently, this nociceptive response is assessed mainly by haemodynamic changes. Recently, analgesia nociception index (ANI) is introduced into anaesthesia practice and provides objective information about parasympathetic (low nociceptive stress) and sympathetic (high nociceptive stress) balance, which reflects the degree of intraoperative nociception/analgesia. This study evaluated the changes in ANI and haemodynamics during anaesthetic induction and intubation, and their correlation during tracheal intubation.
METHODS: Sixty adult patients scheduled for elective brain tumour surgery under general anaesthesia were studied for changes in ANI, heart rate (HR) and mean blood pressure (MBP) during anaesthetic induction and intubation. This was a secondary analysis of a previously published trial. Linear mixed effects model was used to evaluate changes in ANI, HR and MBP and to test correlation between ANI and haemodynamics.
RESULTS: Anaesthetic induction reduced ANI (but not below the critical threshold of nociception of 50) and MBP, and increased the HR (P < 0.001). Direct laryngoscopy and tracheal intubation resulted in increase in HR and MBP with decrease in ANI below the threshold of 50 (P < 0.001). A linear negative correlation was observed between ANI and HR; r = -0.405, P < 0.001, and ANI and MBP; r = -0.415, P = 0.001.
CONCLUSION: Significant changes are observed in ANI during anaesthetic induction and intubation. There is a negative linear correlation between ANI and systemic haemodynamics during intubation.

INTRODUCTION

Direct laryngoscopy and tracheal intubation (DLTI) is an essential component of general anaesthesia for neurosurgery. Despite the therapeutic dose of opioid analgesic and adequate depth of anaesthesia, significant stress response is common during noxious stimulation from DLTI.[1] This response can be detrimental in at-risk patients such as those with cardiovascular co-morbidities and intracranial pathologies.[2] Clinically, changes in the haemodynamic parameters such as tachycardia and hypertension are considered as indicators of nociceptive response to DLTI. Other surrogate tools such as catecholamine levels,[3] heart rate variability (HRV),[4] surgical pleth index (SPI),[5] difference in the state and response entropy,[6] have been explored to assess nociception during DLTI. Analgesia nociception index (ANI) has recently been explored to assess nociception during perioperative period. The ANI is an electrophysiological monitoring tool which provides a 0 to 100 score based on the spectral analysis of HRV; where 0 reflects minimal parasympathetic tone with maximal stress response and nociception, and 100 represents maximal parasympathetic tone with minimal stress response and nociception.[7] There are no previous studies which have examined specifically and in detail, the changes in ANI during DLTI. Previous studies lacked information about time points studied (both before and after intubation),[89] or examined ANI only at one time point after intubation[10] necessitating the need for this detailed study.

The primary objective of this study was to evaluate changes in ANI during anaesthetic induction and DLTI in patients undergoing craniotomy for supra-tentorial brain tumours. The secondary objectives were to evaluate changes in haemodynamics; heart rate (HR) and mean blood pressure (MBP), and correlate changes in the ANI with changes in the haemodynamics during DLTI.

METHODS

This is a secondary analysis of a previously published randomised controlled trial evaluating ANI-guided fentanyl consumption in patients undergoing craniotomy with scalp block (n = 30) and incision site infiltration (n = 30).[11] In brief, the current study is a pooled data analysis of all patients (n = 60) undergoing anaesthetic induction followed by DLTI for craniotomy at a tertiary neurosciences centre in India. The institutional ethics committee approved the study and the trial was registered with the Clinical Trial Registry of India - CTRI/2018/01/011299.

All consecutive consenting patients aged between 18 and 65 years of either sex were recruited if they were scheduled for elective craniotomy for brain tumours over an 18 months period (May 2015 to Oct 2016). Presence of diabetes mellitus, systemic hypertension, significant arrhythmias, chronic pain, allergy to local anaesthetics, coagulopathy, scalp infection, previous craniotomy, pacemaker, pregnancy, and medications affecting the autonomic system were exclusions for this study.

The ANI provides measurement of analgesia/nociception balance, with higher values reflecting increased parasympathetic activity (analgesia) and lower values corresponding to sympathetic activation (nociception).[12] The ANI monitor (MetroDoloris Medical Systems, Lille, France) displays two parameters, the ANIi which is ANI instantaneous (single value) and the ANIm, which is the mean ANI obtained by a 2 minute averaging of ANIi. The ANI > 50 predicts adequate analgesia.[13]

After the patients were wheeled into the operating room, standard monitors (electrocardiogram, pulse oximeter and non-invasive blood pressure) were established. The ANI electrodes were applied at V1 and V5 electrocardiographic positions as per the recommendations of the manufacturers. All patients received fentanyl 2 μg/kg intravenous (IV) and thiopentone 5 mg/kg IV for anaesthetic induction followed by vecuronium 0.15 mg/kg IV for facilitating intubation. This anaesthetic induction protocol is followed in our institution for neurosurgeries in this population and we did not deviate from this practice for the purpose of this study. The DLTI was performed 3 minutes later by one anaesthesiologist with more than 4 years of experience using an appropriate size (3 or 4) Macintosh laryngoscope blade at 1 minimum alveolar concentration (MAC) of sevoflurane with 50% nitrous-oxide in oxygen. Sevoflurane was started soon after intravenous anaesthetic induction and maintained at 1 MAC (between 1.8 to 2.2 end expired sevoflurane concentration) till the completion of intubation. Following completion of data collection regarding DLTI, and 8 minutes before skull pin fixation, local anaesthetic scalp block or pin site infiltration was performed.

We collected data regarding HR, MBP and ANI just before the thiopentone administration and at 1, 2, and 3 minutes after thiopentone to assess the effect of anaesthetic induction on these parameters. Similarly, data regarding HR, MBP and ANI were collected at following time points: just before insertion of laryngoscope (pre-laryngoscopy), during laryngoscopy (intubation 0 minute), and at 1, 2, 3, 4 and 5 minutes after intubation.

No formal sample size was estimated for this explorative secondary analysis. Data was collated offline on a Microsoft Excel spreadsheet for analysis and SPSS version 17 was used for statistical analysis. Interval scale variables are represented as mean ± standard deviation and categorical variables as percentages and frequencies. Preliminary data visualization as line trends for all variables for individual patients demonstrated variation in slopes of change over time and hence, linear mixed effects models incorporating random intercepts and slopes to account for variation in dependent variable due to between patient variability were chosen for analysis. The estimation and hypothesis testing for fixed effect of time on the variables was done in two sets – 4 time points for anaesthetic induction using thiopentone and 7 time points for DLTI. Maximum likelihood method was used, assuming scaled identity covariance matrix structure, with random slopes and intercepts incorporated. The same procedure was used for observing correlations between the variables, with use of HR and MBP as predictor variables for prediction of ANIi. A P < 0.05 was taken as level of statistical significance.

RESULTS

The complete data was available for 57/60 patients and was analysed. The data in 3 patients was lost due to technical issue with ANI sensors and use of vasopressor for managing hypotension. There was no difficulty with mask ventilation in the study population. All patients were successfully intubated in first attempt by the same operator with either 3 or 4 size Macintosh laryngoscope blade. The mean age (years) of the study population was 38.83 ± 14.79, weight (kg) was 59.63 ± 9.13 and 29 (50.9%) patients belonged to the male gender.

There were significant changes (P < 0.001) in the measured parameters- HR, MBP, ANIi and ANIm after anaesthetic induction as compared to baseline [Table 1]. After anaesthetic induction, the HR increased immediately with trend returning towards baseline value at 3 minutes. The MBP decreased with anaesthetic induction and remained significantly lower than baseline value at 3 minutes after induction. Both ANIi and ANIm decreased with anaesthetic induction but the values remained above the threshold of nociception (50).

Table 1

Changes in the study variables in 57 patients during anaesthetic induction (mean±standard deviation)

Time point	HR (bpm)	MBP (mmHg)	ANI Mean	ANI Instantaneous
Pre-induction	77.23±13.15	91.91±20.87	66.25±11.46	65.75±12.84
Thiopentone 1 min	82.91±13.78	81.89±11.45	62.33±18.07	51.82±21.82
Thiopentone 2 min	80.88±10.56	81.88±13.98	57.14±20.09	49.32±16.04
Thiopentone 3 min	79.32±10.73	76.84±14.46	50.21±10.60	50.12±15.06
P	<0.001	<0.001	<0.001	<0.001

P<0.05 is statistically significant

Significant changes (P < 0.001) were observed for all four parameters- HR, MBP, ANIi and ANIm after DLTI as compared to baseline values [Table 2]. The HR increased immediately and significantly with DLTI, was maximal at 2 minutes and gradually decreased without reaching the baseline value. The MBP increased starting with DLTI, reached the maximum at 1 minute and returned to baseline value at 5 minutes after DLTI. The ANIm decreased significantly with DLTI but remained above the critical threshold of 50 throughout the measured time-points. However, ANIi decreased immediately and significantly, remaining below 50 till 2 minutes and increased gradually to reach close to the baseline value by 5 minutes after DLTI.

Table 2

Changes in the study variables in 57 patients during laryngoscopy and tracheal intubation (mean±standard deviation)

Time point	HR (bpm)	MBP (mmHg)	ANI Mean	ANI Instantaneous
Pre-laryngoscopy	74.02±9.99	74.51±12.41	57.77±13.71	59.98±16.15
Intubation 0 min	86.23±14.68	85.64±18.14	55.37±13.32	47.68±24.16
Intubation 1 min	89.11±17.53	94.67±25.45	53.16±14.66	39.60±11.05
Intubation 2 min	89.81±16.03	90.12±22.07	49.95±13.10	43.74±13.22
Intubation 3 min	87.16±16.05	82.09±14.96	49.65±13.61	50.35±19.74
Intubation 4 min	82.56±16.56	79.18±13.03	52.96±17.56	53.00±17.89
Intubation 5 min	81.53±14.83	76.82±14.32	53.89±19.00	55.11±17.95
P	<0.001	<0.001	<0.001	<0.001

P<0.05 is statistically significant

There was a linear negative correlation between ANIi and HR over all time-points for all patients [correlation estimate = −0.405, standard error = 0.052, P < 0.001 [Figure 1]. Similarly, there was a linear negative correlation between ANIi and MBP over all time-points for all patients [correlation estimate = −0.415, standard error = 0.045, P = 0.001 [Figure 2].

Figure 1

Scatter plot between heart rate and ANI instantaneous over all time-points for all patients. B – Coefficient estimate, SE – standard error, P– P value. P < 0.05 is statistically significant

Figure 2

Scatter plot between mean blood pressure and ANI instantaneous over all time-points of all patients. MBP – mean blood pressure, B – coefficient estimate, SE – standard error, P– P value. P < 0.05 is statistically significant

DISCUSSION

We observed increase in HR, and reduction in MBP and ANI values with anaesthetic induction. Our observations are in agreement with the findings of previous studies. An earlier study demonstrated reduction in total HRV with greater depression of HRVhigh component as compared to HRVlow after anaesthetic induction with thiopentone 4 mg/kg with 60% nitrous-oxide in oxygen indicating a greater depressant effect on parasympathetic reflexes (vagolytic effect) as compared to sympathetic system with this technique.[14] Similarly, a significant reduction in high frequency (HF) vis-à-vis low frequency (LF) power was observed after propofol or thiopentone induction in 47 patients.[15] Another study involving 100 patients observed that fentanyl administered during anaesthetic induction decreased total and LF power indicating greater suppression of sympathetic activity while thiopentone increased LF power demonstrating vagolytic effect (increased sympathetic activity).[16] In our study, thiopentone was administered immediately after fentanyl and the overall effect predominantly reflected the vagolytic effect with decrease in ANI.

We observed significant decrease in ANIm and ANIi with increase in HR and MBP during DLTI despite clinically acceptable depth of anaesthesia and analgesia. Similar findings were noted by Ledowski et al. during airway manipulation in 30 patients with ANIi decreasing from 52 to 33 (P < 0.001) at 30 s after intubation.[10] No further time point other than at 30 s was studied to understand the magnitude and pattern of change. Further, data regarding ANI was missing in 43% (13/30) of patients studied resulting in small sample of 17 patients. In a recent study involving 21 neurosurgical patients, the mean ANI decreased from 68 at induction to 52 after intubation.[8] Here again, the changes specific to DLTI were not studied and the time point of measurement of ANI is not clear for both the pre-intubation value, and the post-intubation value. Another study published in non-English language observed ANI of 44 and 39 after intubation in propofol and sevoflurane group respectively, both below the threshold of 50.[9] In this study too, the pre- and post-intubation time point of ANI measurement was not reported. Our study specifically addressed the limitations of these previous studies with regards to the time pattern of change in the ANI and haemodynamics starting from just before DLTI, through the intubation process and then every minute for five minutes after intubation which provided comprehensive impact (magnitude and direction of change) of DLTI on ANI. Boselli et al. also observed a significant decrease in ANI from 72 to 46 (P < 0.01) during suspension laryngoscopy as compared to baseline during anaesthesia with propofol-remifentanil anaesthesia.[17] In this study no intubation was performed and no neuromuscular blocking drug was used making it different from our study population. However, in our study too, the ANI decreased below the threshold of 50 despite 1 MAC of sevoflurane and adequate analgesia (ANI >50) before intubation.

We observed a significant negative linear correlation between ANI and HR, and ANI and MBP during DLTI in this study. A similar negative correlation is documented between ANI and systemic haemodynamics during periods of noxious stimulation in the intraoperative period,[8] and between ANI and postoperative pain as assessed by numerical rating scale score after general anaesthesia.[91819]

The strength of this study is that this study specifically and in detail evaluated the effect of DLTI on ANI and haemodynamic parameters during anaesthesia, unlike previous studies. ANI monitoring provides objective assessment of the balance between pain and analgesia during periods of noxious stimulus such as laryngoscopy and intubation unlike the changes in the systemic haemodynamics, which can manifest from other causes. We also noted that the conventional dose of potent opioid analgesic (2 μg/kg fentanyl IV) is inadequate in ablating the nociceptive response to DLTI as assessed by ANI. The major limitation of this study is the inability to assess the potential impact of transitional state from spontaneous respiration to apnoea to controlled ventilation on ANI during anaesthetic induction and DLTI. Secondly, we excluded patients with likely affection of autonomic nervous system from drugs or diseases such as diabetes mellitus and hypertension. The ANI changes and haemodynamic response might vary differently in these populations for the similar noxious stimulus of DLTI. Lastly, we did not explore the impact of certain confounders such as Cormack Lehane grade, experience of the operator performing the intubation and duration of laryngoscopy. These are important parameters that determine laryngoscopy response which this secondary analysis did not capture. Poor Cormack Lehane grade, less experience of intubation and prolonged duration of laryngoscopy are likely to result in more nociception. These aspects need to be evaluated in future studies.

Use of ANI as a monitoring modality helps assess the magnitude of pain and adequacy of analgesia objectively unlike changes in the heart rate and blood pressure during noxious stimulation of laryngoscopy and intubation. Further studies are needed to evaluate optimal dose of potent opioids in ablating nociceptive response to DLTI and to assess impact of pre-determined airway characteristics on ANI during intubation.

CONCLUSION

Significant increase in heart rate, and decrease in blood pressure and ANI were observed after anaesthetic induction as compared to baseline. Heart rate and blood pressure increased significantly and ANIi and ANIm decreased significantly during tracheal intubation with 2 μg/kg of fentanyl dose. There was a negative linear correlation between ANI and systemic haemodynamics during intubation.

Financial support and sponsorship

Departmental.

Conflicts of interest

There are no conflicts of interest.