Heart rate variability as a measure of mental stress in surgery: a systematic review

Surgery is one of the most demanding safety–critical professions. The operating theatre can be a stressful environment (Menon et al. 2016; Demirtas et al. 2004). There is ample evidence that when physicians are under stress, quality of care is indeed reduced (Wallace et al. 2009). Stress also affects the physicians themselves. Long-term exposure to stress has been associated with a number of ill-health outcomes such as burn-out (Unterbrink et al 2007), cardiovascular diseases (Peter and Siegrist 2000) and depression (Oskrochi et al. 2018). Detection of mental stress is therefore not only extremely important to detect, reduce and prevent the adverse effects of mental stress on quality of care, but also on the physicians themselves.

While accurate and reliable measurements of stress are important, measuring stress is challenging, as stress is perceived and coped with differently by individuals. Measures of stress vary from questionnaires to biochemical evaluations such as cortisol measurements to heart rate variability. There is an increasing interest into more objective measurements of mental stress, as these cannot easily be manipulated and provide an accurate representation of the stress level (Amirian et al. 2014).

One objective measurement which can be used for measuring mental stress in the surgical setting is heart rate variability (HRV) (Jarvalen-Pasasen et al. 2018; Thielmann and Böckelmann 2016). Heart rate variability is the variation in the interval between successive normal NN intervals, which has been shown to decrease as mental stress increases. Variations in heart rate (HRV) can be calculated in the time domain and in the frequency domain [as a power spectral density (PSD) analysis] as well as with non-linear analysis (Sassi et al. 2015; Sammito et al. 2015). In both time and frequency domain analyses, the time intervals between successive normal NN intervals are determined first. The NN intervals are recorded by measuring the difference between two R waves in the QRS complex. Time domain indices of HRV are more direct measures of variations in interbeat intervals (IBI) and include SDNN (standard deviation of IBI), SDANN (the standard deviation of the average IBI), RMSSD (the square root of the mean squared differences of successive IBIs), and NN50 (the number of interval differences of successive IBIs larger than 50 ms). While some specific time domain indices are thought to reflect parasympathetic control of cardiac output (with cardiac output rising in response to stress), other time domain indices cannot be assigned clearly (Schaffer et al. 2017). Time domain indices do not provide detailed information on sympathetic control; the main advantage of using time domain measures is that they are easy to calculate. Frequency domain measures perform more complex calculations on IBI (Fourier transforms), expressing variability in terms of a power density spectrum (energy in specific frequency bands). Frequency domain measures can be calculated for any frequency band, but the most common ones are LF (low frequency, 0.04–0.15 Hz) and HF (high frequency, 0.15–0.4 Hz), but also VLF (very low frequency, < 0.04 Hz) is sometimes used, as is the LF/HF ratio (ratio low frequency/high frequency). Specific frequency bands are thought to reflect sympathetic and/or parasympathetic control, and therefore give more detailed information on the effects of stress on the autonomic nervous system.

However, calculating time and frequency domain measures of HRV is not straightforward and a number of factors need to be considered before analysis. For example, artefact correction is essential. HRV analysis should always be performed on normal-to-normal beat interval data (i.e. all intervals between adjacent R waves in the QRS complexes resulting from sinus node depolarizations) (Lippman et al. 1994). Artefacts such as missed, extra or misaligned beats can significantly alter HRV parameters (Peltola et al. 2012), and analyses using sports watches without correcting the raw data, for example, deliver unreliable results (Sammito and Böckelmann 2016).

Non-linear dynamics methods indicate qualitative aspects of the series of NN intervals (Sammito et al. 2015). These methods can be used for both long-term and short-term measures and has the advantage of being less prone to artefacts.

When evaluating HR and HRV in the field of occupational medicine, several modifiable and non-modifiable factors should be taken in account, as they can affect HR and HRV, the most relevant being alcohol, breathing, fitness activities, sex, cardiovascular diseases, temperature, body weight, noise, age, psychiatric disorders, smoking, hazardous substances, shift work including night shift, metabolic disorders, stress/mental tension and circadian rhythm/time of the day (Sammito et al. 2015).

The aim of this review was to evaluate the current use of HRV measurements within the surgical setting: with what purpose are they used, how long is it measured; to assess which methods are being used for analysing HRV (time domain/frequency domain/non-linear dynamics); and to assess whether HRV was measured correctly (i.e. whether artefacts were corrected).

A total of 518 articles derived from Pubmed, EMBASE and PsycINFO were identified. 78 duplicates were removed, and thus 440 articles were screened for eligibility. 412 articles were excluded based on title and abstract. 11 articles were excluded based on full-text analysis. These articles included medical students as participants (n = 2), no HRV measurement present (n = 4), no surgical stress measurement (n = 2) or other reasons (n = 3). A total of 17 studies were included in the systematic review.

All included studies describe a surgical setting in which the surgeon's mental stress is measured by means of HRV. 8 of the 17 included studies had less than nine participants included in their studies.

This systematic review evaluated the different methods used in the studies for the analysis of HRV (time domain/frequency domain/non-linear dynamics), to assess whether HRV is being measured correctly (i.e. whether artefacts were corrected) and to evaluate the current use of HRV measurements in the surgical setting (short-term vs. long-term measurements) to identify areas for improvement in future HRV research within the surgical setting.

Almost all studies in this review used frequency domain measures, while half of the studies also included time domain measures. The fact that frequency domain measures are being used more often might be because of the fact that when analysing stationary short-term recordings, the task force recommends the use of frequency domain methods (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology 1996). The third method that can be used for calculating variations in heart rate are non-linear analyses, but these methods were not used in any of the included studies. In theory, this is a third method of cardiologists that can be used in HRV research; however because of the characterizing complex systems, successful application in the medical science fields is restricted For future research, the standards of measurement, physiological interpretation, and clinical use can be used to standardize the research into HRV as a measure of stress.

When evaluating the studies included, different goals can be identified for the use of HRV. These goals can be measuring stress during a specific operation, or assessing changes in stress levels between various surgical environments. HRV showed to be a good objective assessment method of stress induced in the workplace environment and was able to pinpoint stressors during operations. In addition, HRV was able to determine which operating techniques provided most stress for surgeons and to determine differences in stress levels between performing and assisting in the surgical procedure. Although different purposes for using HRV were found, the majority of studies had the same overall interest: measuring stress at a specific moment in time, namely during an operation. The included studies were mainly focused on the evaluation of short-term stress, instead of long-term stress and its recovery as the majority of the studies had a short duration of measurement.

Although short-duration measurements can inform us of the level of mental stress during the time frame or situation of interest, measurements of longer duration provide us with vital information on the recovery of stress. Long-term measurements (24 h or more), as opposed to short-term HRV monitoring, enable assessing stress and recovery patterns during normal working hours as well as during leisure time and during sleep (Jarvelin-Pasanen et al. 2018).

Only four of the included studies performed long-term measurements and investigated the long-term effects and recovery of mental stress. These studies found that working night shifts decreased the HRV of surgeons (Amirian et al. 2014) and that higher perceived stress in the operating room is associated with a decreased HRV at night (Rieger et al. 2014). This seems to indicate that stress increases during night shifts and that surgeons are still recovering from high stress of the operating room at night. To identify the long-term effects of stress and prevent its adverse effects on surgeons’ health, more research is needed with long-term HRV measurements, also to better understand if and how surgeons recover from mental stress during working hours.

As heart rate variability is a measure with complex underlying physiological mechanisms, it can be affected by many confounding factors, such as age, weight, physical activity, cardiac innervation, cigarette smoking, alcohol consumption, caffeine consumption, medication use, and core temperature (Jarvelin-Pasanen et al. 2018). The majority of the included studies reported on some of the above-mentioned factors and used these factors as exclusion criteria. Because of the interpersonal differences in HRV, it is recommended participants always serve as their own control.

Heart rate variability is an objective and reliable way of non-invasively monitoring stress in the clinical situation (Böhm et al. 2001; Prichard et al. 2012; Song et al. 2009). This review shows that HRV can be used successfully for different purposes to assess mental stress in the surgical setting, including the effect of operating techniques/environment on mental stress of surgeons and the change in mental stress between performing surgery and assisting. In addition, HRV shows to be a good objective assessment method of stress induced in the workplace environment, as it is able to pinpoint stressors during operations. This review also showed that the current studies are mainly focussed on the short-term measurement of mental stress. There is thus a lack of studies on the long-term effects of mental stress on surgeons, and its recovery.

To standardize HRV research, we further recommend that future research adheres to a single guideline, with using artefact correction to be the most pressing issue.