The finger plethysmographic waveform (FPW) of pulse oximeters is a qualitative indicator of blood volume changes in the fingertip. The changes of this waveform in response to vasodilatation have been evaluated as a guide to the depth of anesthesia (1,2). Airway pressure fluctuation causes a cyclical change of this waveform, which is similar to the variation in systolic blood pressure during a complete cycle of mechanical ventilation (2,3). During mechanical inspiration, venous return and afterload of the right heart decrease because of an increase in intrathoracic pressure, but squeezing of the lung causes an increase in venous return and cardiac output of the left heart. Therefore, during mechanical ventilation, blood pressure increases in inspiration and decreases during expiration. FPW has been evaluated to estimate the blood volume changes under normotensive conditions, and it has been shown that normotensive hypovolemia can be detected by measuring the relative increase in the δ-down (the increase in oximeter pulse amplitude from apneic level during mechanical lung inflation [see Appendix]) component of plethysmogram (3). The applicability of this finding is limited clinically because hypovolemia often is associated with some degrees of hypotension. The effect of the latter on FPW has not been defined clearly (4,5). Hypotension per se may be because of causes other than or in addition to hypovolemia (e.g., pharmacologic or cardiogenic). Therefore, it may be difficult to determine the presence of hypovolemia in hypotensive conditions by inspecting FPW. To evaluate the effects of pharmacologic hypotension on the FPW characteristics, we designed this study on a sample of healthy adult patients under general anesthesia with controlled hypotension.
After institutional approval and written patient consent, 33 ASA physical status I adult nonsmoking patients who were scheduled for tympanoplasty under general anesthesia with controlled hypotension were prospectively included in the study. Patients with a history of cardiovascular, respiratory, and renal diseases were excluded from the study. Lactated Ringer’s solution 10 mL/kg was infused before the induction of anesthesia. Anesthesia was induced with sodium thiopental 5 mg/kg, fentanyl 2 μg/kg, and pancuronium 0.1 mg/kg followed by controlled mask ventilation and tracheal intubation. Anesthesia was maintained with halothane 1% in 50% N2O with O2. Morphine sulfate 0.1 mg/kg was given to supplement the anesthesia. Controlled mechanical ventilation was continued throughout the procedure with a tidal volume of 10 mL/kg at an initial rate of 10 breaths/min, which then was modified for a target end-tidal CO2 range of 34–38 mm Hg (Ventilator unit of Sulla 808 anesthesia machine; Dräger, Lübeck, Germany). After reaching this end-tidal CO2 target (before the induction of controlled hypotension), the minute volume was held constant throughout the hypotensive period. IV infusion was continued using lactated Ringer’s solution at a rate of 4 mL · kg−1 · h−1.
After skin incision and complete exposure of mastoid bone (approximately 25–35 min after the induction of anesthesia), controlled hypotension was started with sodium nitroprusside 10 mg/dL at 0.5 μg · kg−1 · min−1 and then increased to reach an absolute mean arterial blood pressure of 60 mm Hg, provided a relative decrease of not more than 50 mm Hg compared with the baseline value. Monitoring of electrocardiogram and pulse oximetry was performed throughout the procedure (Datex AS/3; Datex/Division of Instrumentarium Corp, Helsinki, Finland). Automated noninvasive arterial blood pressure measurement was performed using an appropriately sized blood pressure cuff around the right arm (Datex AS/3). A circular pulse oximeter probe was attached to the index finger of the left hand. The probe and the finger were wrapped in a well-band foil to minimize probe movement and heat loss. Before the induction of hypotension, arterial blood pressure, end-tidal CO2, and components of FPW (Appendix) were recorded. FPWs were recorded from the monitor display using a fixed camcorder at a distance of 50 cm with a constant zoom to fit the whole monitor screen display. The plethysmographic gain factor of the monitor was held constant throughout the anesthesia. The output of the camcorder was fed to a computer graphic card, and the display was closely worked on using different tools of a Microsoft paint program (Microsoft, Bellevue, WA) to facilitate FPW component calculation. Ventilatory systolic variation (VSV) was measured as the vertical distance between maximum and minimum peaks of waveforms during a complete ventilatory cycle. Apneic plateau was determined as a horizontal line through plethysmographic peaks during a short apnea period. The vertical distances from the maximum and the minimum plethysmographic peaks to the apneic plateau were measured as the δ-up and δ-down components, respectively. The height and width of FPW curve together with its beginning and ending angles (in degrees) were measured. All horizontal and vertical measurement was expressed using the pixel unit of the computer display. The area under the curve was measured as the total number of pixels under a single FPW curve. Measured components of all FPW curves in three consecutive ventilatory cycles before a short apnea period were averaged to reduce beat-to-beat variability of individual curves.
The FPW components and the arterial blood pressure were measured whenever a reduction of at least 5 mm Hg in mean arterial blood pressure relative to the previous measurement occurred, and this was named an instance of measurement. For each instance of measurement, percentage changes of FPW components and blood pressure relative to their baseline values were calculated.
Pearson correlation coefficient was calculated by correlating the percentage changes of each FPW components with their corresponding percentage changes in the arterial blood pressure (systolic, diastolic, and mean). Baseline values are presented as mean ± sd. The assumption of normal distribution for all variables used in the regression analysis was evaluated with one sample Kolmogorov-Smirnov test. A P value <0.05 was regarded as statistically significant. Statistical analysis was performed on a computer using SPSS 10.0 software (SPSS, Chicago, IL).
Seventy-five instances of FPW changes with their corresponding blood pressure changes from 18 women and 15 men were included in the analysis. Each instance was in the form of percentage change from the baseline values (Table 1). The distribution of percentage changes of blood pressure and FPW components were not significantly different from normal. Except for δ-down and VSV changes, other FPW components were not significantly correlated to the changes of blood pressure. δ-down and VSV changes had a significant strong-negative correlation with changes of systolic blood pressure (Table 2, Fig. 1). The range of end-tidal CO2 changes during the hypotensive period was 3 mm Hg (1.4 ± 0.5 mm Hg), which was not correlated with the arterial blood pressure changes.
This study shows that pharmacologic hypotension without hypovolemia can cause changes of VSV and δ-down components of FPW similar to that observed during hypovolemia alone. This is evident from the significant strong-negative correlation between systolic blood pressure changes and VSV or δ-down changes. That is, the more pronounced the hypotension, the more that ventilatory variation would occur in FPW. The increased ventilatory variation is mostly because of the δ-down component, which is again similar to the effects of hypovolemia on FPW (2,6–8).
These findings are almost identical to the effects of blood loss on FPW. Therefore, one cannot determine the presence of hypovolemia accurately during coexisting hypotension.
Although this phenomenon may be partly caused by the effects of pharmacologic hypotension on cardiac preload through venous dilation (9), preload reduction may not be the sole mechanism. The relative stability of end-tidal CO2 and its independence from blood pressure changes in this study do not support the hypothesis of preload reduction. This may be because of the effect of sodium nitroprusside, which mainly reduces the systemic vascular resistance and afterload without a significant effect on cardiac output (10). Pharmacologic hypotension may exaggerate the fluctuation in left ventricular afterload in response to cyclical changes in intrathoracic pressure associated with mechanical ventilation (11). Altered elasticity of the arterial wall during pharmacologic hypotension may also cause a more pronounced ventilatory fluctuation in FPW (12). Whether these fluctuations are greater during hypovolemia or during pharmacologic hypotension is not within the scope of the present study to determine. In addition, it is not clear whether other forms of hypotension behave in a similar way to the nitroprusside-induced one. An earlier study using arterial pressure waveform (but not FPW) has shown that systolic pressure variation is greater during hemorrhage than during sodium nitroprusside-induced hypotension (7).
Finally, as Shamir et al. (3) have shown, only the δ-down and VSV components of FPW undergo changes in response to circulatory changes such as hypovolemia and hypotension. The increase in amplitude or area under the curve in response to vasodilatation, as defined in monographs by some pulse oximeter manufacturers, were not observed in the present study.
FPW is a scaleless curve and therefore cannot be used directly for clinical interpretations. In the present study, changes in FPW components were compared to their baseline values to overcome the limitation imposed by the qualitative nature of FPW.
The method of measurement of FPW components used in this study is valid because the camcorder, which was mounted at a fixed distance from the monitor, used a constant zoom to fit the monitor screen width and height.
The effect of different tidal volumes (mechanical or spontaneous) on these respiratory-induced cyclical changes in FPW and the effect caused by other important ventilatory setting variables may be useful subjects of future investigations. Although the result of this study is negative, it is particularly useful to design an improved generation of pulse oximeters, at least for research purposes, because the changes in FPW characteristics in response to cyclical variation in airway pressure, as observed on a conventional monitoring system, may not be large enough to permit a gross visual inspection for detecting subtle changes. Technically, it is possible to incorporate a hardware module into the design of contemporary pulse oximeters to automate different processes of calculation by detecting the respiratory induced changes in FPW and then to display relevant information on the screen. It may even be possible to extract secondary waveforms by plotting these data over time or as a function of blood pressure.
In conclusion, the results of this study show that pharmacologic hypotension can mimic the effects of minimal normotensive hypovolemia on FPW. Therefore, during pharmacological hypotension, one cannot judge the presence of co-existing absolute hypovolemia based on the finding of FPW.
Different measured components of pulse oximetric waveform. (A) Plethysmographic trace to measure δ-up, δ-down, and ventilatory systolic variation (VSV). A = line through maximal peaks, B = line through peaks during apnea, C = line through minimal peaks, a = VSV, b = δ-up, c = δ-down. (B) Components of a single curve: W = width at midpoint of perpendicular, h = perpendicular, LB = distance between the point of intersection of perpendicular with the curve base and beginning of curve, RB = distance between the point of intersection of perpendicular with the curve base and ending of curve, LA = beginning angle (Arc tan [h/LB]), RA = ending angle (Arc tan [h/RB]). (C) All horizontal and vertical calculations were as pixels. Area under the curve defined as the total number of pixels under the curve. Each small square is equivalent to one pixel.
1. Mattila M, Kurki T, Malinen S. Utilizing pulse oximetry in anesthesia. Monograph by Datex Corp, Helsinki, Finland, 1996.
2. Shelley KH, Murray WB, Chang D. Arterial pulse oximetry loops: a new method of monitoring vascular tone. J Clin Monit 1997; 13: 223–8.
3. Shamir M, Eidelman LA, Floman Y, et al. Pulse oximetry plethysmographic waveform during changes in blood volume. Br J Anaesth 1999; 82: 178–81.
4. Stengele E, Winkler F, Trenk D, et al. Digital pulse plethysmography as a non-invasive method for predicting drug-induced changes in left ventricular preload. Eur J Clin Pharmacol 1996; 50: 279–82.
5. Murray WB, Foster PA. The peripheral pulse wave: information overlooked. J Clin Monit 1996; 12: 349–50.
6. Perel A, Pizov R, Cotev S. Systolic pressure variation: a sensitive indicator of hypovolemia in ventilated dogs subjected to graded hemorrhage. Anesthesiology 1987; 67: 498–502.
7. Pizov R, Ya’ary Y, Perel A. Systolic pressure variation is greater during hemorrhage than during sodium nitroprusside-induced hypotension in ventilated dogs. Anesth Analg 1988; 67: 170–4.
8. Coriat P, Vrillon M, Perel A, et al. A comparison of systolic blood pressure variations and echocardiographic estimates of end-diastolic left ventricular size in patients after aortic surgery. Anesth Analg 1994; 78: 46–53.
9. Schlant RC, Sagaris TS, Robertson RJ. Studies of the acute cardiovascular effects of intravenous sodium nitroprusside. Am J Cardiol 1962; 9: 51–5.
10. Stoelting RK, Viegas O, Campbell RL. Sodium nitroprusside-produced hypotension during anesthesia and operation in the head-up position. Anesth Analg 1977; 56: 391–5.
11. Rooke GA, Schwid HA, Shapira Y. Systolic pressure variation in human during mechanical and spontaneous ventilation. Anesth Analg 1995; 80: 925–32.
12. Pizov R, Segal E, Kaplan L, et al. The use of systolic pressure variation in hemodynamic monitoring during deliberate hypotension in spine surgery. J Clin Anesth 1990; 2: 96–100.