Respiratory Variation in Pulse Oximetry: A Simple Fluid Responsiveness Parameter

Respiratory variation in arterial pulse pressure (ΔPP) has been shown to be the most accurate predictor of fluid responsiveness in mechanically ventilated patients, with a predictive cutoff value of 13%. However, this dynamic parameter is invasive and not widely available. The pulse oximetry plethysmographic (POP) signal resembles the arterial pressure waveform in both shape and amplitude variation. Thus, the respiratory variations in pulse oximetry waveform amplitude (ΔPOP) may be used as a surrogate measure of ΔPP for predicting fluid responsiveness. The aim of the study was to evaluate the relationship between ΔPP and ΔPOP by using standard monitors. Thirty-six mechanically ventilated patients were enrolled during the course of their clinical care in the medical or cardiothoracic surgical intensive care unit (ICU). The arterial pressure and plethysmographic waveform amplitude were recorded simultaneously. The ΔPP and ΔPOP were manually calculated. The relationship between ΔPP and ΔPOP was compared and analyzed for correlations. The ΔPOP is well correlated with ΔPP (r =0.78; P < 0.0001). Our study shows that a ΔPOP value above 15% accurately distinguishes patients with ΔPP above 13% and those with variation of 13% or less with a sensitivity of 93.3% and a specificity of 95.2% (positive predictive value 93.3 %). The results of our study suggest that ΔPOP can be used as a simple and noninvasive predictor of fluid responsiveness in mechanically ventilated patients. รับไว้ตีพิมพ์เมื่อวันที่ 6 ตุลาคม 2558 วารสารวัณโรค โรคทรวงอกและเวชบําาบัดวิกฤต 78 Tanakorn Anantasetagoon, Tanabute Limprukkasem INTRODUCTION Initial therapy in critically ill patients with circulatory failure is volume therapy. The aim of volume expansion is to increase cardiac output. However, if inappropriate, volume expansion is able to induce side effect such as pulmonary or tissue edema. Therefore, assessment of fluid responsiveness is an important issue in the fluid management. Fluid responsiveness assessment has been studied for many years, and it is now established that static parameters of preload, such as central venous pressure or pulmonary capillary wedge pressure, if not extremely low, fail to predict fluid responsiveness accurately1. Dynamic parameters relying on cardiopulmonary interactions in mechanically ventilated patients such as systolic pressure variation (ΔSP)2 and pulse pressure variation (ΔPP) 3 are the reliable predictors of fluid responsiveness. Among the dynamic measurements, ΔPP has consistently been shown to be the most accurate predictor of fluid responsiveness. Michard et al previously found that a ΔPP of 13% had the highest sensitivity and specificity in predicting an increase in cardiac output in response to volume therapy in a study of septic patients4. However, these dynamic indices are invasive, operator dependent, not widely available or expensive. Pulse oximeter waveform presents with variation in amplitude that are related to breathing cycles. Respiratory variation in pulse oximetry plethysmographic waveform amplitude (ΔPOP) has recently shown its potential interest in predicting fluid responsiveness in mechanically ventilated patients. We conducted a prospective study to evaluate the feasibility of the use in ordinary intensive care unit (ICU) and the relationship between ΔPP and ΔPOP. Materials and Methods Patients We prospectively screened for eligibility in only mechanically ventilated patients admitted to the medical intensive care unit or cardiothoracic surgery intensive care unit, older than 18 years. Criteria for inclusion were as follows: (1) presence of an indwelling radial or arterial catheter placed after the decision of the physician in charge of the patient (2) controlled mechanical ventilation with tidal volume of ≥ 8 mL/Kg (3) absence of arrhythmia. We excluded patients in whom pulse oximetry waveform were unreliable measurement (association with noisy waveform). The protocol used in this study was part of routine clinical practice. Although informed consent was waived, the patients or relatives were given clear information about the study. Measurement Invasive arterial blood pressure, electrocardiography, pulse oximetry, exhaled tidal volume, respiratory rate, and peak airway pressure were recorded in all patients. Patients were studied in supine position, and zero pressure was measured at the 4th intercostal space in mid-axillary line. A pulse oximeter photoplethysmographic (POP) waveform was obtained by applying the pulse oximeter probe to a finger or a toe. Arterial pressure and POP wave forms were recorded simultaneously from a bedside monitor to a central station monitor and were analyzed by an observer blinded to other hemodynamic data. Respiratory variations in pulse pressure (PP) analysis PP was defined as the difference between systolic and diastolic pressure. Maximal (PPmax) and minimal (PPmin) values were determined over the same respiratory cycle (Fig 1). To assess the respiratory changes in PP (ΔPP), the percent change in PP was calculated as described in previous study4. ปีที่ 35 ฉบับที่ 3 กรกฎาคม-กันยายน 2557 79 ΔPP (%) = 100 x (PPmax – PPmin) / [(PPmax+ PPmin)/2. Respiratory variations in POP waveform amplitude analysis POP waveform amplitude was measured as the vertical distance between peak and preceding valley trough in the wave form. Maximal POP (POPmax) and minimal POP (POPmin) were determined over the same respiratory cycle (Fig.1). ΔPOP was calculated using the following formulae: ΔPOP (%) = 100 x (POPmax – POPmin) / [(POPmax + POPmin)/2. Fluid responsiveness Patients were divided into 2 groups of fluid responders (FR) and fluid nonresponders (FN). In Figure 1. Simultaneous record of arterial pressure and plethysmographic waveform in a representative patient. accordance with previous studies4, we used the ΔPP of 13% as the cut-off value to differentiate FR from FN. 0.2). Correlation between ΔPP and ΔPOP were assessed with a Pearson Correlation. Statistical significance was defined as P <0.05. Statistical analysis was performed using Stata Statistical Software version 11.0. Results Of the 39 patients recruited from July 2014 to December 2014, 3 subjects (8%) were excluded from analysis due to poor plethysmograph waveforms. Therefore, data were analyzed from 36 subjects: 25 subjects in the cardiothoracic ICU and eleven in medical ICU (Table1). This group consisted of 16 men and 20 Respiratory parameters All patients were mechanically ventilated in a volume-controlled mode with a tidal volume of > 8 ml/ kg ideal body weight. Ventilatory variables such as respiratory rate, the inspiratory to expiratory ratio (I:E) and Positive End Expiratory Pressure (PEEP)were set according to the attending physician. Statistical Analysis We calculated that 30 subjects would be needed to detect a correlation coefficient of 0.5 at the significance level of 5% (α-value of 0.05) and power 80% (β-value of วารสารวัณโรค โรคทรวงอกและเวชบําาบัดวิกฤต 80 Tanakorn Anantasetagoon, Tanabute Limprukkasem women aged between 19 and 84 year (mean age, 56±16 years). The demographic, hemodynamic data and mechanical ventilator setting were shown in table 1. Table 1. Demographic, hemodynamic and ventilator setting data. Parameter Value (mean±SD) Range Gender Male Female 16 20 Age, years 56 ± 16 19-84