Objective Monitoring of Pain Using High Frequency Heart Rate Variability—A Narrative Review

Introduction

In clinical medicine, a patient’s individual perception of pain heavily impacts the course of treatment that is most appropriate for their needs. A challenge clinicians face when managing pain is interpreting the level of pain a patient is experiencing as well as managing pain when a patient is unable to communicate. To assess pain intensity, scales such as the Visual Analogue Scale (VAS), Graphic Rating Scale (GRS), Numerical Rating Scale (NRS), Verbal Rating Scale (VRS), and many others are widely used [1]. For anesthetized patients, hemodynamic parameters are closely monitored to manage patient responses to procedures and surgeries. These methods and pain intensity scales allow clinicians to manage pain appropriately and aid in achieving a favorable postoperative recovery.

The use of hemodynamic parameters has significantly contributed to the understanding of nociception in patients under anesthesia, but changes in these parameters are not completely dependent on nociception and therefore could be misleading. Devices more specific to the surveillance of nociception have been created to better assess pain in anesthetized patients. Some of the more common nociceptive monitors utilized include surgical plethysmographic index, pupillary pain index, nociception level index, quantum consciousness index, and skin conductance algesimeter [2–6]. However, the approaches of each monitor in detecting pain have limitations. For instance, intraoperative medications used in the operating room can confound the readings in many of these devices [6]. One specific example includes pupillary constriction in opioid analgesia, which can confound the pupil diameter measured in the pupillary pain index.

A novel and commercially available device developed by MDoloris, the high-frequency heart rate variability index (HFVI) monitor, also known as the analgesia nociception index (ANI) monitor, assesses pain and nociception in patients under anesthesia through electrocardiographic (ECG) data and measurement of autonomic nervous system parameters. This device’s ability to monitor pain in anesthetized patients could play a major role in the future of pain management in many settings. There are numerous studies exploring the use of high frequency heart rate variability (HRV) monitoring in the operating room (OR), post-anesthesia care unit, emergency room, and non-hospital areas. These studies have also explored its use in different demographics of patients encountered in these settings. This narrative review explicates the observations and results made in these studies that seek to analyze HFVI/ANI in both clinical and non-clinical settings. To provide a comprehensive summary for clinicians considering the use of novel pain monitoring technologies, we have structured this review as follows: first, we discuss the mechanisms of multiple novel pain monitors. We then examine the HFVI/ANI’s ability to accurately detect nociception in response to artificial tetanic stimuli, surgical stimuli, and in non-clinical settings. We evaluate these devices’ potential clinical benefits, including intra-operative opioid sparing and post-operative pain management. Finally, we discuss its use across various patient groups, including palliative, septic, maternal, and pediatric patients.

Heart rate variability in monitoring pain

One benefit of using HFVI/ANI is its simplicity of use on patients. It is a non-invasive device that uses two electrodes, one placed inferior to the right clavicle and the other at left-mid-axillary line (Figure 1). The device interpreta the ECG signals captured to assess sympathetic and parasympathetic changes during anesthesia and management of nociceptive stimuli during surgery [7]. This is determined by calculating R-R wave interval fluctuations on ECG, indicating heart rate HRV. Through the spectral analysis of HRV, ANI analyzes the predominance of the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS) [8, 9]. Specifically, the high frequency HRV represents the activity of the PNS, whereas the low frequency HRV represents the combined activity of the PNS and the SNS. As such, the main principle of detecting pain and nociception is based on the balance between high and low frequency HRV, with a higher frequency HRV representing PNS activity and thereby anti-nociception and a lower frequency HRV representing nociception. The device then converts the spectral analysis into a score between 0–100 to link and quantify the parasympathetic and sympathetic activity and ensuing analgesic need (Figure 2). Scores below 50 signify high sympathetic activity and therefore an increased need for analgesia, while scores above 50 signify high parasympathetic activity and therefore no need for changes in analgesia. Scores between 50–70 are considered to represent adequate analgesia by the manufacturer. Utilizing this method, HFVI/ANI can be used for adjusting analgesia before physiologic responses in hemodynamic parameters.

Other commercially available pain monitors

While this review focuses on the HFVI/ANI device, it is worth noting the existence of other pain monitors and the similarities and differences in their methods of interpreting nociception. The surgical plethysmographic index analyzes the photoplethysmographic amplitude and pulse interval to monitor a patient’s sympathetic tone as a reflection of intraoperative nociception and hemodynamic response to surgical stimuli [2]. Similar to the HFVI/ANI, the advantage of using this monitor is its ease of use; it just requires a finger pulse oximeter. The pupillary pain index monitors the pupillary reflex response to surgical stimulation as an indicator of nociception (i.e. a larger pupil diameter indicates greater nociception) [3]. Due to its reliance on pupils, it is a quick and easily accessible method to monitor nociception, as many surgeries are done in supine position. Limitations include surgeries in the prone position or medications that can confound pupil diameter (e.g. opioids). The skin conductance algesimeter utilizes electrodes placed on palms or soles to monitor the activity of sympathetic nerve firing in the skin as a reflection of sympathetic drive (and therefore nociception) [6]. Another device, the nociception level index, combines many of the parameters utilized by other devices simultaneously to measure nociception, including HRV, photoplethysmography, and skin conductance [4]. Similar to the surgical plethysmographric index, its advantage is its ease of use, as it only requires a single finger clip. Finally, the quantum consciousness index monitor provides a unique methodology in assessing nociception by analyzing autonomic activity through an EEG. Since many other monitors rely on cardiopulmonary parameters, this monitor is less susceptible to influence by cardiovascular drugs used intraoperatively [5].

Given the novel concept of intraoperative nociception monitoring, each device has their own studies that critically analyze their strengths and limitations. Moreover, each device has varying methods of analyzing physiologic data and interpreting nociceptive response, hence the efficacy of each device should be analyzed separately in various settings (e.g. intra-operatively, ICU setting, pediatric patients with an overall different autonomic tone). This review focuses on summarizing the current findings on the HFVI/ANI device in the context of both clinical and non-clinical pain monitoring into a centralized narrative review.

Evaluating HFVI/ANI through tetanic nociceptive stimulation

Researchers have utilized several methods to administer nociceptive stimuli to test HFVI/ANI’s ability to detect pain and hemodynamic responses. A decrease in HFVI/ANI scores following nociceptive stimuli would suggest an ability to detect pain and a decrease in scores prior to changes in hemodynamic parameters would suggest that the device has a potential use in predicting these changes. Such studies are valuable to assess HFVI/ANI’s validity and improve patient outcomes.

Funcke et al. used tetanic and intracutaneous electric stimulation as nociceptive stimuli and observed ANI scores to determine its ability to detect the stimuli [10]. They found that decreased HFVI/ANI scores had a strong association with painful stimuli administered with a predictive probability (P_K) = 0.98, sensitivity of 87.9%, and specificity of 98.5%. They also monitored scores with changes in heart rate and blood pressure to evaluate the device’s ability to predict hemodynamic responses and determined that increases greater than 5 mmHg for blood pressure, 5 beats/min for heart rate, or 10% in either indicate a hemodynamic response and tested HFVI/ANI’s ability to predict these changes. The P_K to detect these changes was 0.70; sensitivity was 20.6% and specificity was 46.8%. These results by Funcke et al. suggest that HFVI/ANI predicts the presence of nociceptive stimuli but cannot reliably predict the occurrence of a hemodynamic response. Gruenwald et al. monitored changes in HFVI/ANI scores in 25 adult patients who were administered a nociceptive stimulus after they were given gradually increasing doses of remifentanil [11]. They found that the scores after stimulation significantly decreased for all remifentanil doses given. When monitoring heart rate and mean arterial pressure (MAP), they found little change for all doses of remifentanil used. Susano et al. identified decreases in HFVI/ANI scores after administration of electrical stimulation [12]. The results of these three studies suggest that ANI appropriately predicts the presence of nociceptive stimuli. However, their inability to consistently detect hemodynamic responses upon administration of nociceptive stimuli and the low predictive values found by Funcke et al. do not support the device’s validity in predicting hemodynamic responses [10–12].

HFVI/ANI use in conscious, awake patients outside clinical settings

While HFVI/ANI was initially intended for use during surgery, few studies investigated the ability of the device to measure differences in sympathetic and parasympathetic tones in conscious and awake patients. In 2012, DeJonckheere et al. tested HFVI/ANI’s response to negative emotional stimulus in human subjects (n = 25) outside the hospital by presenting these subjects with a video that strongly elicited negative-emotions while recording their HFVI/ANI scores [13]. In this study, they found that the average scores decreased immediately after watching the film, and that these new scores after the film were significantly different than the subjects’ average scores before the film was presented (p = 0.023).

Some studies reported no evidence supporting the clinical relevance of HFVI/ANI in conscious patients. In a randomized stimuli and placebo-controlled, single-blinded study, Jess et al. applied non-painful and painful stimuli to participants (n = 20) in a darkened and silenced room while measuring their HFVI/ANI values and asked participants to rate their pain levels after each stimulus to obtain an NRS value [14]. This study revealed that any type of stimuli, whether painful or non-painful, decreased participants HFVI/ANI values, and no significant correlation was found between HFVI/ANI values and NRS values (all Pearson correlative values were between 0 and −0.1). Similarly, Issa et al. investigated whether HFVI/ANI and NRS values were correlated when providing increasing electric stimulus to the wrist of conscious, supine patients (n = 23) in an operating room [15] and determined a very weak negative correlation between HFVI/ANI and NRS values (Pearson coefficient = −0.089; p = 0.045).

The studies discussed in this section have their own limitations, such as low sample sizes and the large differences in their methodologies. Furthermore, these studies are not representative of the small number of existing studies on HFVI/ANI and conscious patients. Ultimately, a generalized conclusion about HFVI/ANI use in conscious patients cannot be made due to the limited studies on this topic.

Evaluating HFVI/ANI through surgical and procedural nociceptive stimulation

Many researchers have conducted studies on HFVI/ANI during surgeries and other procedures as well. Studies have monitored HFVI/ANI scores during procedures for changes in relation to nociceptive stimuli, hemodynamic parameters, and analgesia management. Ledowski et al. studied the changes in scores during airway management or incisions and in relation to hemodynamic responses in 30 patients [16]. They found that scores decreased during airway management and incisions; this was followed by increases in heart rate and systolic blood pressure in under 30 s. In these cases, the use of fentanyl led to increased HFVI/ANI scores and stabilization of heart rates and systolic blood pressure in under 5 min. This suggests that the device can detect the nociceptive stimuli before they lead to hemodynamic responses and identify when analgesia was optimal. Another study monitored 27 patients during total knee replacements and found that 20 patients had decreased scores prior to at least a 20% increase in heart rate or systolic blood pressure [17]. Boselli et al. monitored HFVI/ANI scores and hemodynamic responses in 50 patients during suspension laryngoscopies [18]. They applied increases of at least 20% in heart rate or systolic blood pressure to determine when a hemodynamic response has occurred and tested the device’s abilities to predict a response within 5 min. They found that scores below 55 predicted a hemodynamic response with a sensitivity of 88%, specificity of 83%, and AUC ROC = 0.88. In another study, HFVI/ANI was used to predict hemodynamic responses in 128 ear-nose-throat and orthopedic procedures [19]. It was found that when scores decreased by at least 19%, hemodynamic responses were predicted with a sensitivity of 85%, specificity of 85%, and AUC ROC = 0.90. Other studies have similarly found HFVI/ANI scores to decrease upon nociceptive stimuli during procedures such as intubations, supratentorial craniotomies, trocar insertion, and propofol injections [20–23]. Unlike the studies conducted with artificial stimuli, these results suggest that the HFVI/ANI device could be used to detect nociceptive stimuli and predict hemodynamic changes well which supports its use in surgeries and procedures. However, it is notable that one study that supported the device’s ability to detect nociceptive stimuli also demonstrated that it had a poor predictive performance for hemodynamic adverse events [22]. Furthermore, a more recent study did not find a significant association between HFVI scores and postoperative numerical rating scores in patients just before extubation [24]. With the relatively small sample sizes in many of these studies, unaccounted factors such as fluid status, and differing conclusions, there remains some doubt regarding this device’s validity for clinical use. The clinical studies outlined mostly support HFVI/ANI’s use as an aid in anesthesia and analgesia management, but more work is required to determine its overall validity.

HFVI/ANI and intraoperative opioid sparing

Researchers have also studied HFVI/ANI’s use in intraoperative opioid administration. Specifically, studies have been conducted to determine if the use of HFVI/ANI in opioid administration could minimize opioid use during surgeries compared to conventional practice. Soral et al. used HFVI/ANI to assist opioid management in patients undergoing colonoscopies and compared opioid use to patients undergoing the same procedure without the device [25]. The group of patients that were not using the device were managed by standard methods of opioid management with remifentanil, while HFVI/ANI group were administered remifentanil to achieve a score of 50–70 and adjusted as needed. In these 102 patients, they found that significantly less remifentanil was used in the HFVI/ANI group compared to the control group. Dundar et al. also observed that significantly less remifentanil was used when utilizing HFVI/ANI to track analgesia in a study of 44 patients in breast surgeries [26]. In a study of 60 patients undergoing bariatric surgery, significantly less sufentanil was used when analgesia was guided by HFVI/ANI [27]. Few studies have shown conflicting results when using the device to guide analgesia. In a study of 72 patients undergoing spinal procedures, patients with HFVI/ANI guided analgesia did not have less sufentanil use compared to traditional analgesia management [28]. However, they did have different time courses of sufentanil administration, but this did not lead to changes in complications, cortisol, or pain. A study of 120 patients undergoing laparoscopic cholecystectomies found no significant difference in fentanyl use when using the device [29]. HFVI/ANI also did not lead to a difference in fentanyl use in a study of 60 mastectomies [30].

Studies regarding HFVI/ANI guided opioid administration have shown conflicting results. Some findings support the use of HFVI/ANI to decrease intraoperative opioid use [25–27]. Others have found no significant differences in opioid administration with use of the device [28–30]. While there is conflicting evidence regarding the device’s ability to minimize intraoperative opioid use, it should be noted that the use of HFVI/ANI facilitates individualized intraoperative opioid dosing, rather than opioid avoidance. Additionally, it is important to point out the differing variables amongst these studies that could impact HFVI/ANI scores. For example, these studies monitored patients undergoing different types of procedures. Differing results amongst studies analyzing various procedures suggest that the results cannot be generalized to intraoperative opioid management as a whole. The studies also utilized different opioids to manage analgesia which could also have impacted the confounding results. Age and gender have also been found to impact HRV, which can confound HFVI/ANI scores [31]. Specifically, researchers have found minor differences in HRV amongst different age groups and larger differences based on gender, with males having higher cardiac sympathetic activity and response to pain when compared to females. Overall, more research is needed to support the idea that HFVI/ANI guided analgesia could lead to less opioid use, but there is promise for this application.

Evaluating postoperative pain using HFVI/ANI

Researchers have considered the use of HFVI/ANI to detect pain postoperatively along with VRS, VAS, GRS, and NRS. Several studies have been conducted to determine if correlations exist between HFVI/ANI and widely used pain intensity scales. A study of 200 postoperative and endoscopy patients analyzed the correlation of NRS and HFVI/ANI scores during periods of pain [32]. They found a significant negative linear correlation between NRS and HFVI/ANI scores within 10 min in the post-anesthesia intensive care unit. HFVI/ANI scores below 57 were correlated with NRS scores above 3 with a sensitivity of 78%, specificity of 80%, and AUC ROC of 0.86. When analyzing HFVI/ANI scores below 48 and NRS scores above 7, they determined a sensitivity of 92%, specificity of 82%, and AUC ROC = 0.91. Likewise, Abdullayev et al. looked at the correlation between NRS and HFVI/ANI scores in 107 patients in the post-anesthesia intensive care unit who underwent procedures requiring fentanyl or remifentanil [33]. They determined a negative linear correlation between NRS and HFVI/ANI scores with a R² = −0.312 and p = 0.001. While R² = −0.312 does not represent a strong correlation, the mentioned relationship is still statistically significant. A study of 120 post-surgery patients that assessed the correlation of NRS and HFVI/ANI scores found a statistically significant but weak correlation [34]. They also found a low specificity and sensitivity when identifying NRS scores of 0 versus 6–10 with HFVI/ANI. Another recent study even combined the use of the HFVI/ANI device with machine learning algorithms to predict postoperative pain as well [35]. Similar to other instances, these studies demonstrate that there is potential for the use of this device in clinical settings, but also show some limitations. Few studies found strong correlations between HFVI/ANI and the widely used NRS to quantify the intensity of pain [32, 33], while others found a correlation but not a strong one [34]. Studies at larger scales may be needed before the device’s efficacy to detect postoperative pain can be determined, but few current studies suggest that it could play a role in such settings.

The effect of intraoperative HFVI/ANI-guided analgesia on postoperative pain

Studies have also explored HFVI/ANI use in intraoperative setting to reduce postoperative pain. In one study, 50 patients scheduled for lumbar spine-related procedures were assigned to either a control group or HFVI/ANI managed group [36]. Postoperative outcomes such as NRS scores were then monitored in the two groups. They found that the HFVI/ANI managed group had significantly lower postoperative opioid use and NRS scores than the control group. Another study of 180 surgeries utilizing HFVI/ANI to aid in analgesia management found that 86% of patients did not require postoperative opioids [37]. After one day post-surgery, these patients reported a max NRS rating of 2. Several previously mentioned studies also monitored postoperative pain in cases utilizing HFVI/ANI intraoperatively through NRS and VAS [27–30]. There was no significant difference between patients who had HFVI/ANI-guided analgesia and patients in control groups. Due to these conflicting results, no inferences are made regarding intraoperative HFVI/ANI-guided analgesia in decreasing postoperative pain.

Other potential clinical utilities of HFVI/ANI in adults

A number of studies looked into potential utilities of HFVI/ANI other than its intended use. For instance, Tanrikulu et al. highlighted the clinical benefit of using the device for continuous pain monitoring in the cardiovascular ICU [38]. Because ANI can be used to assess pain and nociception, other studies attempted to utilize this when monitoring the comfort levels of palliative care patients. In a study by Bauschert et al., a total of 58 clinical evaluations were conducted on 33 patients undergoing end of life care [39]. These clinical evaluations included both HFVI/ANI and various other observational scales, including but not limited to the patient comfort score, critical care pain observation tool, and behavioral pain score. Bauschert et al. reported that the clinical assessment and HFVI/ANI values were consistent with each other’s readings of comfort/discomfort 45 out of 58 times (77.58%), and inconsistent for 13 out of 58 evaluations (22.41%). In this study, patients whose clinical evaluations and HFVI/ANI values were discordant (clinical evaluation was comfort; HFVI/ANI evaluation was discomfort) gave the opportunity for clinicians to increased therapeutic treatment. In another palliative care study, Six et al. compared the subjective caregiver assessments on pain with the objective measurement of nociception from HFVI/ANI [40]. In this prospective observational study, Six et al. observed that there was poor agreement between the subjective caregiver assessments on pain and consciousness and the measures provided by HFVI/ANI. As such, they concluded that including objective monitoring of nociception in end of life patients may be beneficial in improving patient comfort.

In studies on sepsis, literature has shown that the onset of sepsis is characterized by decreases in HRV[41]. As such, HFVI/ANI was used to demonstrate that poorer outcomes in patients with sepsis were associated with lower HRV[42, 43]. One study utilized this device to measure HRV and found that lower HRV has predictive power for mortality in septic patients [42]. Similarly, another study utilized HFVI/ANI to measure HRV and found that HRV has predictive capabilities in determining if patients with sepsis would develop multiple organ dysfunction syndrome [44].

In the field of obstetrics, one study found that patients who underwent caesarean sections had significantly higher HFVI/ANI scores (decreased pain sensation) immediately after making skin contact with their newborns [43]. Another study found that the device accurately reflected pain during uterine contractions in expecting patients, highlighting another potential area where the device could be used [45]. In women undergoing breast tumor excisions, one study utilized the device to evaluate the effectiveness of their pectoralis muscle fascia blocks and found that scores <50 were strongly correlated with the need for postoperative analgesia [46].

Despite these studies, there is still a large gap in the literature regarding HFVI/ANI use outside the surgical setting. With these presented findings, however, there may be undiscovered potential for clinically relevant benefits through HRV monitoring. For instance, future studies on sepsis could assess whether early HFVI/ANI intervention to assess HRV in septic patients would improve mortality outcomes. Future studies in obstetrics could also bridge this gap by comprehensively assessing the validity of this device in labor patients, observing how scores change in response to epidural anesthesia, or whether HFVI/ANI overall improves obstetrics outcomes. But at this present time, more research is necessary before making generalized conclusions on the other potential clinical utilities of HRV monitoring for patients at the end of life, patients with sepsis, and patients undergoing labor.

HFVI/ANI in children

Several studies assessed the potential clinical relevance of HFVI/ANI in children undergoing surgery as well. For example, Avez-Couturier et al. recorded HFVI/ANI scores prior to surgical incision and after surgical incision in 26 children undergoing a muscle biopsy [47]. After incision, they observed a significant decrease in scores (indicative of nociceptive stimulation) when compared to scores prior to incision. Likewise, in a study with 58 children, Migeon et al. found that HFVI/ANI scores decreased significantly when there was a 10% or greater increase in heart rate during skin incisions compared to groups whose heart rate did not increase in response to the incision [48]. These results from Migeon et al. also support the device’s utility in reflecting not only pain and the body’s autonomic response, but also the changes in hemodynamic status. In another study, 42 out of 49 children had decreased scores within 80 s following surgical incision [49]. Furthermore, the AUROC value for HFVI/ANI to detect pain from the skin incision was greater than 0.75. Other values typically used to reflect the body’s response to surgery, such as heart rate and blood pressure, had lower AUROC values in this study compared to the HFVI device.

Other studies had different methods for assessing the device’s utility in children. Instead of using surgical stimuli and observing if HFVI/ANI scores decreased to indicate nociception, one study did the opposite. In an observational study with 131 pediatric patients, researchers demonstrated the utility of HFVI/ANI to indicate sufficient analgesia. In this study, anesthesiologists blinded to the HFVI/ANI monitor would announce when they decided to administer an analgesic, and researchers would observe the change scores at that time [50]. Researchers in this study observed that when HFVI/ANI scores were less than 50 while analgesia was given, the scores increased, suggesting the clinical relevance of ANI for detecting anti-nociception. Another unique methodology of a study was assessing the device in the post-anesthesia care unit (PACU) instead of the operating room. In the PACU, Gall et al. confirmed the device’s reflection of pain: pediatric patients who underwent surgery (n = 32) had significantly decreased HFVI/ANI values whereas the control group (no surgery; n = 30) had significantly higher HFVI/ANI values [51].

While ANI may have demonstrated clinical utility in the previous studies, it is still uncertain whether ANI improves surgical outcomes in pediatric patients. A recent study found a significantly decreased intra-operative sufentanil consumption when HFVI was used compared to the standard [52]. Conversely, another outcome previously analyzed was postoperative emergence agitation; however, there was no statistically significant difference found between the device treatment and no device treatment groups [53]. Given this existing gap in the literature, further research is required for any HFVI/ANI effects on outcomes in this patient population.

Conclusions

Monitoring pain and nociception could provide significant benefits for patient care and outcomes. Intraoperatively, it has the potential to be an additional objective source of information to help guide analgesia use, reduce opioid consumption, and lower postoperative complications. Additionally, it can be used to monitor pain and comfort levels in non-anesthetized patients outside surgical settings, such as the PACU, and the emergency department. In both surgical and non-surgical settings, the ability to monitor nociception can bridge the inability of patients to communicate their pain due to factors such as being anesthetized, intubated, or simply being critically ill. Furthermore, due to its ease of use, the device has potential in non-clinical settings as demonstrated by the studies conducted on conscious, awake patients outside a hospital. As such, future research can overcome the numerous confounding variables discussed in the presented studies and the High Frequency Heart Rate Variability Index demonstrates a promising avenue for clinical and non-clinical pain monitoring.

Authors’ contributions

Hum, Bill: conceptualization, investigation, writing - original draft, review, and editing. Shibly, Yusef: writing - original draft, review, and editing. Christophides, Alexa: conceptualization, investigation. Jin, Zhaosheng: writing - review and editing. Elias, Murad: writing - review and editing. Bergese, Sergio: conceptualization, supervision, and writing - review and editing.

Funding

This research did not receive external funding from any agencies.

Ethical Statement

Not Applicable.

Data Availability Statement

Source data is not available for this article.

Conflict of Interest

The authors declare no conflict of interest and have no financial disclosures to make.

Appendices and nomenclature