SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. References

The objective of this study was to compare linear and non-linear analysis of heart rate variability (HRV) in terms of correlation with haemodynamic fluctuation during induction of general anaesthesia. Pre-operatively, HRV was estimated by the MemCalc method in 114 patients scheduled for general anaesthesia. After anaesthesia was induced with propofol, fentanyl and vecuronium bromide, tracheal intubation was performed. Haemodynamic fluctuations during induction of anaesthesia were recorded and the correlation between pre-operative HRV and haemodynamic fluctuation was examined using logistic regression analysis. As an index of non-linear analysis of HRV, ultra short-term entropy (UsEn) correlated better with blood pressure fluctuation than did the ratio of the power of low frequency component of HRV to that of high frequency component (LF/HF). In contrast, although LF/HF significantly correlated with heart rate increase caused by tracheal intubation, the correlation between UsEn and heart rate fluctuation was not significant.

Induction of general anaesthesia is a time of great haemodynamic fluctuation. A variety of interventions [1–6] have been examined as possible stabilisers of the cardiovascular response, but none of these interventions have resulted in stable induction of anaesthesia because of either the inadequacy of their effect or unwanted side-effects. It would be of great help if we could predict the haemodynamic response to induction of general anaesthesia.

Recently, some investigators have found linear analysis of heart rate variability (HRV) to predict the incidence of hypotension caused by induction of general [7] or spinal anaesthesia [8, 9]. While frequency-domain measures of HRV, such as spectral analysis, focus on the status of cardiac autonomic activity, it is proposed that non-linear indices may provide insight into the overall structure of the heart rate regulating system. Although it was reported that pre-operative point dimension correlation (PD2), as a non-linear index of HRV, predicted hypotension after spinal anaesthesia [10], there is no previous report examining the correlation between non-linear index of HRV and circulatory response to induction of general anaesthesia.

The MemCalc method, which is a combination of the maximum entropy method for spectral analysis and the non-linear least squares method for fitting analysis (Tarawa, Suwa Trust, Tokyo, Japan), has recently been developed [11, 12]. This method enables us to estimate reliable HRV from a series of RR intervals over 30 s and recognises the abnormal RR interval of premature beats or artefacts, including noise, and removes it automatically. The MemCalc method does not cause distortion of the power calculation even if the underlying variation is changed, e.g. a change in respiration. When compared to conventional methods of estimating HRV such as Fast Fourier Transform, this method has a great advantage in the clinical situation as it makes it possible to estimate HRV with RR intervals for shorter periods. It also provides information about the entropy of RR interval as a non-linear index of HRV. Although there are numerous entropy formulations, entropy is a concept that addresses system randomness and predictability, with greater entropy often associated with more randomness and less system order [13]. When it is calculated with the MemCalc method, it is normalised from 0, which means no randomness of heart rate (HR), to 100, which means complete randomness, much like Gaussian white noise. We have called this entropy of HRV ultra short-term entropy (UsEn) as it can be estimated from a short series of RR intervals.

Thus, we compared UsEn with LF/HF ratio in terms of the correlation with haemodynamic fluctuation during induction of general anaesthesia.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. References

After IRB approval and written informed consent from all patients were obtained, 114 patients (ASA I–II, aged 41–84 years) scheduled for elective surgery were investigated in this study. Exclusion criteria were as follows: emergency operation, cardiac arrhythmia, autonomic nervous dysfunction, and multi-attempts at tracheal intubation.

Anaesthesia and haemodynamic measurement

All patients were allowed to consume clear fluids until 2 h before entering the operating theatre. Without any premedication, patients entered the operating theatre and lay on an operating table. Measurements of non-invasive blood pressure, pulse oximetry and electrocardiography were then initiated. After 10 min of rest on the operating table, baseline values of systolic blood pressure (SBP), and heart rate (HR) were recorded (SBPbaseline and HRbaseline); meanwhile, the estimations of pre-operative HRV were also started.

Anaesthesia was induced with a target-controlled infusion of propofol so as to produce a blood concentration of 3 μg.ml−1, with infusions of 1.5 μg.kg−1.h−1 of fentanyl and 0.15 mg.kg−1.h−1 of vecuronium bromide. Three minutes later, tracheal intubation was performed and the patient's lungs were mechanically ventilated with 50% oxygen in air. Following induction of anaesthesia, the target concentration of propofol was decreased to 2 μg.ml−1 for maintenance of anaesthesia. Consecutive non-invasive blood pressure measurement using STAT mode was started just before induction of anaesthesia and was continued until 5 min after tracheal intubation. The haemodynamic fluctuation was evaluated as follows.

First, we recorded specific value of SBP or HR at each point of time:

  • • 
    SBPbaseline: baseline SBP;
  • • 
    SBPpostinduction: the lowest SBP between induction and tracheal intubation;
  • • 
    SBPpostintubation: the highest SBP for 5 min after tracheal intubation;
  • • 
    HRbaseline: baseline HR;
  • • 
    HRpostinduction: the lowest HR between induction and tracheal intubation;
  • • 
    HRpostintubation: the highest HR for 5 min after tracheal intubation.

Next, we compared the SBP or HR changes caused by induction of anaesthesia:

  • • 
    ΔSBPinduction: SBPbaseline- SBPpostinduction;
  • • 
    ΔSBPintubation: SBPpostintubation- SBPpostinduction;
  • • 
    ΔHRinduction: HRbaseline- HRpostinduction;
  • • 
    ΔHRintubation: HRpostintubation- HRpostinduction.

Finally, as an overall index of haemodynamic fluctuation, ΔSBPtotal and ΔHRtotal were obtained as follows:

  • • 
    ΔSBPtotal: ΔSBPinduction + ΔSBPintubation;
  • • 
    ΔHRtotal: ΔHRinduction + ΔHRintubation.

HRV and entropy measurement

ECG signals were obtained from a conventional anaesthesia monitor (Hewlett Packard, Model 66S, Palo Alto, USA), digitised at 1000 Hz and transferred to a personal computer (Epson NT2700, Suwa, Japan). After the RR intervals were determined, on-line analysis of the HRV was done using the MemCalc method. Then the averaged HR, UsEn, the logarithms of power of the low (LF; 0.04–0.15Hz) and high (HF; 0.15–0.4Hz) frequency component of HRV and LF/HF ratio during conscious state were calculated. UsEn was estimated from eight RR intervals.

Statistical analysis

Data were analysed using standard software (JMP version 5, SAS Institute Inc., Cary, NC). To determine the independent predictors for the magnitude of haemodynamic fluctuations, logistic regression analysis was performed for the determination of significant correlation between haemodynamic fluctuation and UsEn, logLF, logHF and LF/HF. p values of < 0.05 were considered significant. To optimise the accuracy of the analysis, all dependent variables showing SBP and HR fluctuations were categorised into four groups divided by the 25th, 50th and 75th percentiles.

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. References

Patients' characteristics, the results of HRV analysis and haemodynamic fluctuation during induction are shown in Table 1.

Table 1.   Demographic data of patients and pre-operative HRV.
  1. Values are mean (SD) or number.

Age; year 61 (18)
Gender; male/female 55/59
Height; cm161 (9)
Weight; kg 60 (14)
BMI; kg.m−2 23 (4)
ASA status; I/II 47/67
Pre-operative HRV
 UsEn 45 (12)
 logLF; ms2  2.46 (0.50)
 logHF; ms2  2.11 (0.61)
 L/H  3.5 (0.61)
 SBP; mmHg142 (20)
 HR; beats.min−1 74 (15)

The results of logistic regression analysis for the correlation between pre-operative HRV parameters and variables indicating SBP and HR fluctuation are shown in Tables 2 and 3, respectively. Only UsEn significantly correlated with all variables indicating SBP fluctuation, including both specific SBP and the SBP changes. That is, patients with lower UsEn showed greater SBP fluctuation. In contrast, none of the linear indices correlated with SBPpostinduction. Moreover, LF/HF failed to correlate with ΔSBPinduction. Although LF/HF significantly correlated with ΔSBPintubation and ΔSBPtotal, the p values for the significant correlation were higher than with UsEn. Although all indices correlated with HRpostinduction, correlation between HRV and HR fluctuation was generally weak. Only LF/HF significantly correlated with HR postintubation and ΔHRintubation.

Table 2.   The results of logistic regression analysis for the correlation between HRV and SBP fluctuation.
 UsEnlogLFlogHFL/F
  1. OR, odds ratio; 95% CI, 95% confidence interval. Values in bold represent significant correlation.

SBPpostinduction
 p0.0040.360.120.98
 OR0.96   
 95% CI[0.96, 0.96]   
SBPpostintubation
 p0.0020.0030.0020.007
 OR1.051.0011.0010.82
 95% CI[1.04, 1.05][1.001, 1.001][1.001, 1.001][0.81, 0.83]
ΔSBPinduction
 p< 0.00010.0003< 0.00010.36
 OR1.071.0011.002 
 95% CI[1.06, 1.07][1.001, 1.001][1.002, 1.002] 
ΔSBPintubation
 p< 0.00010.0003< 0.00010.006
 OR1.081.0011.0010.81
 95% CI[1.077, 1.08][1.001, 1.001][1.001, 1.001][0.80, 0.82]
ΔSBPtotal
 p< 0.0001< 0.0001< 0.00010.01
 OR1.091.0021.0020.82
 95% CI[1.08, 1.09][1.002, 1.002][1.002, 1.002][0.81, 0.83]
Table 3.   The results of logistic regression analysis for the correlations between HRV and HR fluctuation.
 UsEnlogLFlogHFL/F
  1. OR, odds ratio; 95% CI, 95% confidence interval. Values in bold represent significant correlation.

HRpostinduction
 p0.0040.0040.0270.03
 OR1.041.0011.0010.85
 95% CI[1.05, 1.040][1.001, 1.001][1.001, 1.001][0.84, 0.87]
HRpostintubation
 p0.450.450.390.003
 OR   0.8
 95% CI   [0.79, 0.82]
ΔHRinduction
 p0.810.860.720.96
 OR
 95% CI
ΔHRintubation
 p0.270.310.810.026
 OR   0.86
 95% CI   [0.85, 0.87]
ΔHRtotal
 p0.730.820.550.079
 OR
 95% CI

As an example, the scatter graph for the relationship between pre-operative UsEn and ΔSBPtotal is demonstrated in Fig. 1. The correlation coefficient for the relationship between UsEn and ΔSBPtotal was 0.47 (95% CI 0.31–0.60).

image

Figure 1.  Scatter graph for the relationship between UsEn and SBP fluctuation during induction of anaesthesia (ΔSBPtotal).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. References

Spectral analysis of HRV is widely accepted as a non-invasive probe for assessment of the autonomic control of the heart [14]. Although there is some controversy in terms of interpretation, it is generally agreed that the LF/HF ratio represents the balance between sympathetic and parasympathetic parts of the autonomic nervous system, LF reflects sympathetic and vagal nerve activity, and HF represents vagal nerve activity. In reality, however, the behaviour of the cardiac system is dynamic, non-linear and non-stationary. Linear analyses may not account for all aspects of cardiac performance [15]. Thus, analytical methods derived from non-linear dynamics have been proposed as a new approach for assessing the dynamics of HRV [13, 16–19].

In this study, we found that HRV estimated by the MemCalc method significantly correlated with SBP fluctuation during induction of general anaesthesia. Both UsEn, as an index of non-linear analysis, and logLF, logHF, LF/HF, as an index of linear analysis, significantly correlated with ΔSBPtotal as an overall index of SBP fluctuation. However, only UsEn significantly correlated with specific SBPpostinduction. Moreover, LF/HF failed to correlate with ΔSBPinduction. We suggested that UsEn was superior to logLF, logHF and LF/HF in terms of correlation with SBP fluctuation caused by induction of anaesthesia. Although correlation between age and HRV has been described, logistic regression analysis revealed that HRV is an independent predictor of ΔSBPtotal. We demonstrated that correlation between HRV and ΔSBPtotal cannot be explained only by age.

In this study, the greatest ΔSBPtotal was 145 mmHg (SBPbaseline, SBPpostinduction, SBPpostintubation were 150, 85, 165 mmHg, respectively). We tried to minimise the noxious stimulus of tracheal intubation by giving fentanyl before intubation. Further effort to attenuate the effect of intubation does not necessarily result in stable haemodynamics because of side-effects or inadequate efficacy. On the other hand, the correlation between haemodynamic fluctuation during induction of anaesthesia and the prevalence of peri-operative complications has not been established. Thus, we believe that this kind of haemodynamic fluctuation is not avoidable but does not increase the risk to patients.

Recently, it was reported that LF/HF predicts blood pressure decrease after spinal anaesthesia [9, 10]. The effect of spinal anaesthesia on circulation is mainly attributed to alteration of the autonomic nervous activity caused by preganglionic sympathetic block. This could be the reason why pre-operative LF/HF strongly correlated with blood pressure decrease after spinal anaesthesia. On the other hand, the effect of general anaesthesia on circulation is more complex. In addition to the sympathovagal effect, general anaesthetics, including propofol, have negative inotropic effects on the heart as well as direct vasodilatory effects. While frequency-domain measures of HRV such as spectral analysis focus on the status of cardiac autonomic activity, it has been proposed that non-linear indices including entropy may provide insight into the overall structure of the heart rate regulating system. Some previous reports suggest that non-linear indices may represent circulatory healthiness or robustness [20, 21]. Rather than sympathovagal balance, overall robustness of the heart rate regulating system may be more important in maintaining stable blood pressure during induction of anaesthesia.

Recently, it was demonstrated that there is a significant relationship between HRV pre-operatively measured at rest and blood pressure stability during anaesthesia induction in diabetics [7]. Knüttgen et al. reported that blood pressure after tracheal intubation was significantly higher in patients with normal coefficient of variation (CV) values. Although CV is an index of time-domain analysis of HRV, it represents the overall magnitude of HRV. Their results may also suggest that the overall magnitude of HRV, rather than the balance of autonomic nervous activity, is more important in predicting SBP fluctuation during induction of anaesthesia.

On the other hand, the correlation between HRV and HR fluctuation was generally weak. UsEn, logLF and logHF were not significantly correlated with any indices demonstrating HR fluctuation during induction of anaesthesia. In contrast to SBP fluctuation, only LF/HF significantly correlated with HR changes caused by tracheal intubation. That is, patients with pre-operative high LF/HF tended to develop tachycardia after tracheal intubation. HR fluctuation caused by tracheal intubation is an acute response to the mechanical stimuli to the larynx and trachea, which is mediated by both divisions of the autonomic nervous system. The balance of autonomic nervous activity may play a significant role in the determination of the HR response to tracheal intubation.

Previously, a significant reduction in HF has been shown in patients who developed relative bradycardia after a sufentanil-vecuronium induction [22], results which are not in agreement with ours. This study used only sufentanil and vecuronium for induction of anaesthesia, whereas we employed propofol in addition to fentanyl. Moreover, that study compared HR at 4 min after tracheal intubation, whereas we compared the difference between baseline HR and maximum HR after tracheal intubation. These differences of study protocol may account for the discrepancy between their and our study.

There is one limitation of our study. The MemCalc method, which is a relatively new method of estimating HRV, may be vulnerable to the criticism that it has not yet been validated. However, Sawada et al. [12] have demonstrated the validity of MemCalc method compared with autoregressive modelling. Moreover, there have been reports investigating autonomic function with this new technique in the field of anaesthesia [23, 24]. According to the recommendations of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, the comparison between this method and fast Fourier transformation will be needed in the future.

In conclusion, HRV measured using the MemCalc method significantly correlated with SBP fluctuation during induction of anaesthesia. UsEn correlated with SBP fluctuation better than LF/HF ratio. In contrast to SBP fluctuation, the correlation between HRV measured by the MemCalc method and HR fluctuation during induction of anaesthesia was weak. Only LF/HF significantly correlated with HR increase caused by tracheal intubation.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. References
  • 1
    Maguire AM, Kumar N, Parker JL, Rowbotham DJ, Thompson JP. Comparison of effects of remifentanil and alfentanil on cardiovascular response to tracheal intubation in hypertensive patients. British Journal of Anaesthesia 2001; 86: 903.
  • 2
    Thompson JP, Hall AP, Russell J, Cagney B, Robowtham DJ. Effect of remifentanil on the haemodynamic response to intubation. British Journal of Anaesthesia 1998; 80: 4679.
  • 3
    Chraemmer-Jorgensen B, Hoilund-Carlsen PF, Marvin J, Christensen V. Lack of effect of intravenous lidocaine on hemodynamic responses to rapid sequence induction of general anaesthesia: a double-blind controlled clinical trial. Anesthesia and Analgesia 1986; 65: 103741.
  • 4
    Hamill JF, Bedford RF, Weaver DC, Colohan AR. Lidocaine before endotracheal intubation. Intravenous or laryngotracheal? Anesthesiology 1981; 55: 57881.
  • 5
    Derbyshire DR, Smith G, Achola KJ. Effect of topical lignocaine on the sympathoadrenal responses to tracheal intubation. British Journal of Anaesthesia 1987; 59: 3004.
  • 6
    Atlee JL, Dhamee MS, Olund TL, George V. The use of esmolol, nicardipine or their combination to blunt hemodynamic changes after laryngoscopy and tracheal intubation. Anesthesia and Analgesia 2000; 90: 2805.
  • 7
    Knüttgen D, Trojan S, Weber M, Wolf M, Wappler F. Pre-operative measurement of heart rate variability in diabetics: a method to estimate blood pressure stability during anaesthesia induction. Anaesthstist 2005; 54: 4429.
  • 8
    Hanss R, Bein B, Ledowski T, et al. Heart rate variability predicts severe hypotension after spinal anesthesia for elective cesarean delivery. Anesthesiology 2005; 102: 108693.
  • 9
    Hanss R, Bein B, Weseloh H, et al. Heart rate variability predicts severe hypotension after spinal anesthesia. Anesthesiology 2006; 104: 53745.
  • 10
    Chamchad D, Arkoosh VA, Horrow JC, et al. Using heart rate variability to stratify risk of obstetric patients undergoing spinal anesthesia. Anesthesia and Analgesia 2004; 99: 181821.
  • 11
    Ohtomo N. New method of time series analysis and its application to Wolf's sunspot number data. Japanese Journal of Applied Physiology 1994; 33: 282131.
  • 12
    Sawada Y, Ohtomo N, Tanaka Y et al. New techinique for time series analysis combining the maximum entropy method and non-linear least squares method: its value in heart rate variability analysis. Medical and Biology Engineering and Computing 1997; 35: 31822.
  • 13
    Pincus SM, Goldberger AL. Physiological time-series analysis: what does regularity quantify? American Journal of Physiology 1994; 266: H164356.
  • 14
    Akselrod S, Gordon D, Ubel FA, et al. Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat-to-beat cardiovascular control. Science 1981; 213: 2202.
  • 15
    Schumacher A. Linear and nonlinear approaches to the analysis of R-R interval variability. Biology Research for Nursing 2004; 5: 21121.
  • 16
    Katz MJ. Fractals and the analysis of waveforms. Computer in Biology and Medicine 1988; 18: 14556.
  • 17
    Peng CK, Havlin S, Hausdorff JM, Mietus JE, Stanley HE, Goldberger AL. Fractal mechanisms and heart rate dynamics. Long-range correlations and their breakdown with disease. Journal of Electrocardiology 1995; 28 (Suppl.): 5965.
  • 18
    Yamamoto Y, Nakamura Y, Sato H, Yamamoto M, Kato K, Hughson RL. On the fractal nature of heart rate variability in humans: effects of vagal blockade. American Journal of Physiology 1995; 269: R8307.
  • 19
    Rosenstein MT, Collins JJ, De Luca CJ. A practical method for calculating the largest Lyapunov exponents from small data sets. Physica D 1993; 65: 11734.
  • 20
    Kaplan DT, Furman MI, Pincus SM, et al. Aging and the complexity of cardiovascular dynamics. Biophysical Journal 1991; 59: 9459.
  • 21
    Pincus SM, Cummins TR, Haddad GG. Heart rate control in normal and aborted-SIDS infants. American Journal of Physiology 1993; 264: R63846.
  • 22
    Estafanous FG, Brum JM, Ribeiro MP, et al. Analysis of heart rate variability to assess hemodynamic alterations following induction of anesthesia. Journal of Cardiothoracic and Vascular Anesthesia 1992; 6: 6517.
  • 23
    Kanaya N, Hirata N, Kurosawa S, Nakayama M, Namiki A. Differential effects of propofol and sevoflurane on heart rate variability. Anesthesiology 2003; 98: 3440.
  • 24
    Win NN, Fukayama H, Kohase H, Umino M. The different effects of intravenous propofol and midazolam sedation on hemodynamic and heart rate variability. Anesthesia and Analgesia 2005; 101: 97102.