During induction of general anaesthesia a range of changes to the circulation happens in a very short span of time, with possible deleterious consequences for organ perfusion[8]: Loss of sympathetic tone leads to decreased afterload and preload, as well as negative ino- and chronotropy, the latter cardiac effects may also reflect a direct effect of the induction agents. The cumulative systemic effects are variable, but decreasing MAP and SV is frequent with modern anaesthetics, albeit with SV maintained to a higher degree than blood pressure[9, 10]. Even so, most patients will only have intermittent non-invasive monitoring of blood pressure, along with heart rate and pulse plethysmography for arterial oxygen saturation. Advanced and continuous monitoring focusing on optimizing systemic blood flow is gaining popularity [11]. However, continuous monitoring of systemic and central haemodynamics is more invasive and/or requires specialized equipment and continuous training [12]. Thus, these modalities will not be offered to all patients. A potentially large group of patients is subject to “less-than ideal-monitoring” during induction of general anaesthesia – due to lack of resources, unrecognized cardiovascular risk, and/or unanticipated susceptibility to cardiovascular compromise.
The peripheral perfusion index (PPI) is obtained non-invasively by photoplethysmography, ubiquitously present during GA. The PPI in awake patients is predictive of the complication rate in major surgery [13], is a predictor of mortality in severe acute illness [14–16], predicts hypotension during fluid withdrawal in uremic patients [17], and hypotension induced by spinal anesthesia for sectio caesarea [18, 19], as well as being an early predictor of central hypovolemia and compensatory sympathetic activation in healthy volunteers [20, 21]. In high-risk surgical patients the intra-operative PPI is a strong predictor of serious post-operative complications or death [22]. During GA the PPI tracks changes in MAP and SV induced by manipulation of preload[5]. To our knowledge only one previous study have assessed the PPI during induction of anaesthesia [6].
The plethysmographic signal consists of a pulsatile “AC” and non-pulsatile component “DC” with PPI = AC/DC. Sympathetic tone is the major determinant of PPI in the awake person. Indeed, sympathetic block demonstrated by increasing PPI has been described as determinant of successful neuraxial [23] or peripheral [24, 25] block. During general anaesthesia sympathetic outflow is reduced to a very low baseline [26–31], abolishing its dominance over the PPI. Thus, the main determinant of PI becomes SV [4, 5, 32, 33].
Clearly the PPI during the transitional period of induction of anaesthesia is subject to multiple changes from both sympatholysis and alterations of global circulatory parameters. Still, we hypothesized the possibility of
describing two different patterns of sympatholysis: One dominated by vasodilatation with preserved stroke volume and flow (≈ PPI increasing) and one with decreased cardiac stroke volume dominating (≈ PPI decreasing).
The LiDCO Rapid monitor uses pulse power analysis to calculate SV. It is an uncalibrated device, using nomograms based on the patient’s age, weight, and height to estimate SV and thus CO. The algorithm is based on the principle of conservation of mass (power), assuming a linear relationship between the net power change and the net flow in the vascular system. The algorithm tracks changes in SV/CO with good reliability [11, 34, 35]. For patients without an arterial cannula, we used the CNAP module to provide a high-resolution continuous trace of the arterial pressure curve to the LiDCO monitor. This technology tracks the intra-arterial pressure curve with good reliability, although tracking may be less precise during induction[36–38].
In this cohort of mixed surgical patients, we found a baseline PPI of 1,7[1,0–3,8]% (median and quartiles), not unlike 1,4% previously reported in healthy volunteers [3]. Mean(± SD) baseline PPI was 2,6(± 2,3) (due to the right-skewed distribution) consistent with a large recent cohort of high risk surgical patient reporting a mean PPI of 2,6(± 2,6) [22].
We did not find baseline PPI to be predictive of hemodynamic instability during induction. It would not be unreasonable to expect a very low PPI (e.g. <0,5%) to be predictive of post-induction instability as a low PPI might suggest a circulation dependent on high sympathetic tone. Most likely our cohort consisted of “too healthy” patients to show this effect, indeed only 4 patients had a baseline PPI below 0,5. Furthermore, anxiety or pain, may also lead to a low PPI, but not a higher likelihood of post-induction instability. Thus, the negative correlation between baseline MAP and PPI reflects the effect of sympathetic tone on PPI in the awake patient [4].
On average for the total cohort 2 minutes after induction PPI increased by 36%, whereas MAP, SV, and CO decreased by 35%, 26%, and 37%, respectively. This is consistent with a recently published study describing an inverse correlation between ∆MAP and ∆PPI during induction [6]. However, our study having a larger and more diverse population, as well as monitoring of cardiac stroke volume and - output, allows for a deeper insight into the mechanisms influencing the PPI. Stratifying the changes in PPI upon induction makes for a more comprehensive image of the relationship between PPI and systemic hemodynamics during induction. Even though PPI on average increases and indices of systemic circulation (MAP, SV, CO) on average decreases, a more compelling narrative appears: Stratifying PPI response in quartiles shows that with larger post-induction increase in PPI an attenuated cardiovascular compromise is seen. In the quartile of highest PPI increase especially SV is preserved, only decreasing to 86% of pre-induction values. Conversely, in the quartile of in the lowest post-induction PPI values marked cardiovascular compromise is seen, with MAP reduced to 54% of baseline values (Table 2; Fig. 1).
These changes likely reflect different degrees of reduction in cardiac function due to induction of anaesthesia. The highest increase PPI quartile thus have decreased arterial vasomotor tone but largely intact stroke volume contrasting the low PPI quartile with compromised cardiac function – due to sympathetic dependent cardiac function, reduced cardiac filling by venous dilatation, or both. As noted, the major determinant of PPI in the awake patient (i.e. pre-induction/baseline) is sympathetic tone. Upon induction, all things being equal, the PPI will increase as vascular tone decreases. However, as sympathetic tone is decreased changes in SV becomes the major determinant of PPI; with a large post-induction decrease in cardiac function leading to net decrease in PPI, whereas a cardiac function only mildly affected by induction of anaesthesia leads to a net increase in PPI [4].
A recent study using non-invasive monitoring during induction found a stable SV and decreased MAP after induction [9]. The patients differ from those in our study by being much younger (mean 36 vs 60 years) and healthier (ASA 1–2: 91% vs 70%). The pattern of preserved SV during induction thus resembles the most stable patients of out cohort; the quartile with the highest increase in PPI during induction.
To increase the clinical usefulness of our results, we dichotomized our data into PPI decreasing vs PPI stable or increasing (Fig. 2). With PPI decreasing MAP was reduced to 57% and SV to 63% of baseline; with PPI increasing MAP was 70% and SV 80% of baseline. Again, a clear pattern emerges leading to a simple recommendation to the clinician: Observe the changes in PPI during induction of anaesthesia: Should the PPI not increase during induction our data suggests increased risk of compromised cardiovascular function. Thus, when observing a decreasing trend of PPI upon induction of anaesthesia, the provider should be vigilant to intervene with vasopressor/inotrope.
In summary, in a mixed cohort of surgical patients undergoing general anaesthesia using minimally- or non-invasive monitoring of MAP, SV, and CO we demonstrated the expected effects of induction of anaesthesia with reduction of haemodynamic parameters. When stratifying the patients according to the effect of induction upon PPI it was possible to elicit a plausible physiological background to the observations and to discriminate between trivial and more severe degrees of cardiovascular compromise.