Can prolonged expiration manoeuvres improve the prediction of arterial PCO2 from end-tidal PCO2?

Size: px

Start display at page:

Download "Can prolonged expiration manoeuvres improve the prediction of arterial PCO2 from end-tidal PCO2?"

Geraldine Johnston
5 years ago
Views:

1 British Journal of Anaesthesia 1997; 78: Can prolonged expiration manoeuvres improve the prediction of arterial PCO from end-tidal PCO? B. TAVERNIER, D. REY, D. THEVENIN, J.-P. TRIBOULET AND P. SCHERREEL Summary We have studied, in 16 patients undergoing thoracoabdominal oesophagectomy, if two prolonged expiration manoeuvres improve prediction of arterial PCO ( P a CO ) from end-tidal PCO 3 ( &. 3 ( & PCO at the end of a simple prolonged expiration 3 ( & and PCO at the end of a prolonged expiration preceded by sustained hyperinflation of the lungs ( P E CO ), were measured during laparotomy, in the lateral thoracotomy position during two-lung ventilation, and after transition to one-lung ventilation. ( Pa CO P E CO ), was 1.3 (SD 0.4) kpa during laparotomy and this remained stable throughout the study. Both manoeuvres decreased the mean arterial to peak expired PCO difference, particularly during onelung ventilation. However, P E 1 CO and P E CO did not agree more closely with P a CO than P E CO at any stage of the study. We conclude that these manoeuvres did not improve estimation of P a CO from P E. (Br. J. Anaesth. 1997; 78: ). CO Key words Carbon dioxide, partial pressure. Carbon dioxide, measurement. Partial pressure, end-tidal. Partial pressure, arterial. Partial pressure, carbon dioxide. Surgery, thoracic. Surgery, abdominal. Monitoring, carbon dioxide. Continuous measurement of end-tidal carbon dioxide partial pressure ( CO ) has long been used for non-invasive estimation of arterial PCO ( P a CO ) during anaesthesia. 1 However, a major concern is that P E CO generally underestimates P a CO, with variations both between and within subjects. 3 5 The arterial to end-tidal PCO difference ( Pa C O CO ) depends on the spread of ventilation perfusion ratios. One index of an increased spread of ventilation perfusion ratios is an increased slope of the expiratory plateau of the single breath test (expired carbon dioxide plotted against expired volume). 6 Thus individuals with an increased phase III slope on the capnogram often have an increased arterial to end-tidal PCO difference. 4 During controlled ventilation in adults, the expiratory plateau is usually flat and nearly horizontal, but several factors, such as age and airways disease, and in particular smoking, are associated with an increase in the slope of the expiratory plateau. 5 6 In this situation, the CO value cannot be taken as a good estimate of alveolar PCO and hence not as an estimate of P a CO. In this case, a simple prolonged expiration (permitting expiration to continue for 5 15 s) or a forced expiration (induced by a gentle squeeze on the patient s chest) may give a more representative estimation of Pa CO. 78 Thus these manoeuvres are currently recommended to improve prediction of P a CO. from CO especially when a steeply sloping plateau is observed However, the validity and usefulness of these measurements in individual patients have not been investigated. Patients with oesophageal carcinoma are almost exclusively smokers with resultant airways disease. Their ( Pa C O CO ) and the phase III slope of their capnogram would thus be expected to be increased during anaesthesia, especially during thoracotomy We have therefore studied in patients undergoing thoracoabdominal oesophagectomy two prolonged expiration manoeuvres in order to assess if this improves prediction of P a CO from CO. Patients and methods We studied 16 successive patients (14 males), mean age 60 (range 48 71) yr, undergoing thoracoabdominal oesophagectomy for cancer. All patients were devoid of symptomatic cardiac disease. Fourteen were smokers. Lung function tests (FEV 1, FVC) were performed before operation. The study was approved by the Ethics Committee of our University Hospital and all patients gave informed consent. Patients were premedicated with hydroxyzine 100 mg orally, 90 min before induction of anaesthesia with fentanyl 5 g kg 1 and thiopentone 5 mg kg 1. Pancuronium 0.1 mg kg 1 was administered to produce neuromuscular block and facilitate tracheal intubation. Anaesthesia was maintained with fentanyl, and enflurane with 50% nitrous oxide in oxygen. Anaesthetic agents were administered as required to avoid variations in mean arterial pressure BENOIT TAVERNIER, MD, DOMINIQUE REY, MD, DIDIER THEVENIN, MD, PHILIP SCHERREEL, MD (Département d Anesthésie Réanimation Chirurgicale II); JEAN-PIERRE TRIBOULET, MD (Service de Chirurgie Générale et Endocrinienne); Hôpital Claude Huriez-Chru de Lille,, Avenue Oscar Lambret, Lille Cedex, France. Accepted for publication: January 14, 1997.

2 ( Pa ) and prolonged expiration 537 C O CO 0% of baseline. Intraoperative monitoring consisted of continuous electrocardiography, pulse oximetry, direct arterial pressure measurement via a radial artery cannula and rectal temperature. For the abdominal procedure, the lungs were ventilated mechanically (Monnal D ventilator) at a ventilatory frequency of 1 bpm with a tidal volume of 10 1 ml kg 1 and an inspiratory:expiratory (I:E) ratio of 1:. After abdominal wound closure, a double-lumen endobronchial tube with a left-sided bronchial stem (Carlens tube) was sited for right thoracotomy. Endobronchial tube position and isolation of the two lumens were checked by auscultation, spirometry and respiratory gas detection with the patient in the lateral position. During one-lung ventilation (OLV), ventilatory settings were ventilatory frequency 15 bpm, tidal volume 8 ml kg 1 and I:E ratio 1:. PROLONGED EXPIRATION MANOEUVRES To obtain a prolonged expiration, the ventilator was abruptly set, during one expiration, to spontaneous ventilation to permit expiration to continue for 10 0 s. Maximum concentration of expired carbon dioxide ( F E 1CO ) was then recorded. Values were considered valid when a well identified plateau was obtained. For the second manoeuvre, a 3-litre reservoir bag connected to the ventilator system was used. Lungs were inflated manually with fresh gases (oxygen and nitrous oxide) contained in the bag until airway pressure reached 35 cm H O. This pressure was held for 3 s, and then full expiration was permitted, with recording of maximum concentration of expired carbon dioxide ( F E CO ). Manoeuvres were performed in this non-randomized order because, in preliminary measurements, F E 1CO increased when recorded immediately after F E CO. MEASUREMENTS P a CO was measured at 37 C in duplicate (ABL 500 and 510, Radiometer), and values were corrected to the patient s temperature according to the following formula 13 : 0.019(T 37) = P a CO (37 C)--10, where T patient temperature. The differences between arterial and expired carbon dioxide partial pressures were calculated using temperaturecorrected P a CO values. End-tidal carbon dioxide fractions ( FE CO, FE 1CO and F E CO) were measured from gas sampled between the tracheal (or endobronchial) tube and the Y-piece ofthe ventilator tubing using a sidestream infrared capnometer (Capnomac Ultima, Datex) calibrated according to the manufacturer s instructions before use in each patient. 14 The capnograph traces were recorded (7041 M X-Y recorder, Hewlett-Packard) for analysis ofthe slope of the phase III of the capnograms. The alveolar plateau was arbitrarily considered normal (i.e. nearly horizontal) when the trace virtually did not increase in the final portion (last third) of the phase III of the capnogram. P E, CO P E 1CO and CO were obtained from FE CO, F E 1CO and FE CO, respectively, according to the formula: PCO = FCO --(Pb -- PHO) where Pb barometric pressure and PH O water vapour pressure at 37 C. At each stage of the study, mean arterial pressure, heart rate and rectal temperature were measured. Blood was then sampled from the radial artery cannula for immediate P a CO measurement. Syringes were rinsed with sodium heparin to prevent coagulation. Simultaneously, FE CO was recorded. Then, F E 1CO and F E CO were obtained successively by appropriate manoeuvres. Finally, arterial blood was sampled for P a CO measurement to verify stability of P a CO during the procedure. The time necessary to obtain the entire set of measurements was 10 min. Initial measurements were made during the abdominal procedure, before wound closure (Tl). A second set of measurements was performed with the patient in the lateral position after 0 min with twolung ventilation (TLV) (T). After transition to OLV with the pleura open, measurements were repeated after an additional 0 min (T3) and 50 min (T4). A final set of measurements was obtained after wound closure, after 0 min with TLV, with the patient in the lateral position (T5). STATISTICAL ANALYSIS Values ofthe differences between P a CO and each end- expiratory carbon dioxide partial pressure (i.e. ( Pa C O CO ), ( Pa C O 1 CO ) and ( Pa C O CO )) were computed at each sample point, with results expressed as mean (SD). Differences were analysed using two-way analysis of variance for repeated measurements. Multiple comparisons were made with Fisher s protected least significant difference testing. P 0.05 was considered significant. To assess the usefulness of each prolonged expiration manoeuvre to improve estimates of Pa CO from end-expiratory PCO, mean values for arterial to end-expiratory PCO differences were used as the mean error (or bias) of the method tested (i.e. P E CO, P E 1CO and P E CO ) relative to P a CO. The SD of the differences were used as the precision of the method tested. The mean range difference ( SD) defined the 95% limits of agreement of 15 each end-expiratory measure vs Pa CO. Results Lung function tests (FEV 1, FVC) were not performed in two asymptomatic non-smoking patients. Table 1 Preoperative spirometry data in 14 patients Patient No. FEV1 (% expected) FVC (% expected)

3 538 British Journal of Anaesthesia Figure 1 Typical example of P E CO (end-tidal carbon dioxide partial pressure), P E 1CO (peak expired carbon dioxide partial pressure during prolonged expiration) and P E CO (peak expired carbon dioxide partial pressure during prolonged expiration after manual hyperinflation) measurements, showing capnographic recordings during one-lung ventilation. A steep phase III slope without a good plateau was observed in the capnogram, and prolonged expiration manoeuvres clearly increased peak expired carbon dioxide partial pressure (with CO P E 1CO ). Table P a CO and arterial to peak expired partial pressure differences in 16 patients undergoing thoracoabdominal oesophagectomy (mean (SD)). 1CO peak expired carbon dioxide partial pressure during prolonged expiration; CO peak expired carbon dioxide partial pressure during prolonged expiration following manual hyperinflation. Significant differences: *P 0.01 vs ( Pa CO CO ); P 0.01 vs ( Pa CO 1 CO ); P 0.01 vs T Measurement Pa CO CO 1CO CO T1 (end of abdominal procedure) 5.3 (0.6) 1.3 (0.4) 0.8 (0.5)* 0.6 (0.5)* T (TLV for 0 min) 5.3 (0.6) 1.4 (0.5) 1.0 (0.6)* 0.4 (0.6)* T3 (OLV for 0 min) 5.5 (0.6) 1.3 (0.4) 0.4 (0.5)* 0.1 (0.5)* T4 (OLV for 50 min) 5.5 (0.6) 1.4 (0.5) 0.3 (0.6)* 0. (0.5)* T5 (TLV, after skin closure) 5.3 (0.6) 1. (0.5) 0.5 (0.7)* 0.1 (0.6)* Preoperative spirometry results were 80% of expected values in four of 14 patients (table 1). P a CO was almost unchanged during each stage of the study in all patients (the larger variation observed in one set of measurements was 0.5 kpa). For analysis, P a CO values before and after each set of measurements were averaged. Heart rate and mean arterial pressure were maintained within 0% of baseline when measurements were performed, and moderate progressive hypothermia was observed (from 37. (SD 0.3) C at T1 to 36. (0.4) C at T5). Typical examples of capnographic traces obtained with both manoeuvres are shown in figure 1. At T1, P a CO was 5.3 (0.6) kpa and P E CO was 4.0 (0.5) kpa, and both remained unchanged throughout the study. As a result, ( Pa C O P E CO ), observed at T1, was 1.3 (0.4) kpa, and did not change significantly at each stage of the study (table ). However, analysis of the capnograms showed that an almost horizontal phase III was obtained in 13 patients at T1, and in only eight patients at T. Moreover, all patients had steep phase III slopes without a good plateau at T3 and T4 (and 13 patients at T5). Both prolonged expiration manoeuvres decreased markedly the mean arterial to peak expired carbon dioxide partial pressure difference throughout the study, particularly during OLV: ( Pa C O 1 CO ) and ( Pa C O CO ) were significantly smaller than ( Pa C O P E CO ) at Tl (0.8 (0.5) kpa and 0.6 (0.5) kpa, respectively, P 0.01 vs P E CO ) and decreased further after transition to OLV (table ). In addition, ( Pa C O CO ) was significantly smaller than ( Pa C O 1 CO ) at each stage of the study (table ); T3, T4 and T5 values were not significantly different from zero. However, the manoeuvres did not reduce the variations in ( Pa C O P E CO ) observed between patients. The agreement between each endexpiratory PCO and P a CO, as represented by the SD of the differences, was not improved with P E 1CO or P E CO compared with P E CO at any stage of the study (table ). To quantify the size of the measurement error that could be observed in clinical practice, data from all stages of the study were pooled. The 95% limits of agreement of P E 1CO and P E CO compared with P a CO were 0.6 to 1.8 kpa and 1.0 to 1.4 kpa, respectively (and kpa for ( Pa C O P E CO )). This represented a range of values larger than 5% of the estimated value ( P a CO ) and thus were considered clinically unacceptable. The two manoeuvres did not reduce the variations observed within patients. To illustrate this phenomenon, differences between the largest and smallest gradients obtained throughout the study were calculated patient by patient for the three endexpiratory carbon dioxide partial pressures. The results ranged from 0.3 to 1.0 kpa (mean 0.5 (SD 0.) kpa), 0.3 to.0 kpa (1.0 (0.6) kpa) and from 0.5 to 1.8 kpa (1.0 (0.5) kpa) for ( Pa C O P E CO ), ( Pa C O 1 CO ) and ( Pa C O CO ) respectively. Discussion We undertook a study to see if two prolonged expiration manoeuvres improved estimation of P a CO from P E CO in patients undergoing oesophageal surgery, to achieve a wide range of ( Pa C O P E CO ) values with increased phase III slopes. None of the methods tested improved prediction of P a CO, even in situations which are currently recommended (i.e. a capnogram with a steeply sloping alveolar plateau ). The ( Pa C O P E CO ) value is known to be increased in several situations. 1 5 In our study, the widened ( Pa C O P E CO ) observed at the end of the abdominal procedure was characterized by a flat and nearly horizontal plateau of the capnogram in 13

4 ( Pa ) and prolonged expiration 539 C O CO patients. This may suggest predominance of spatial ventilation perfusion mismatching associated with a situation that does not influence pulmonary mechanical properties where well perfused and under perfused lung units empty simultaneously. 16 However, asynchronous emptying combined with increased ventilation perfusion spread cannot be eliminated. Analysis of the phase III slope of the capnogram, contrary to the single breath test (PCO plotted against expired volume), may not be a function of ventilation perfusion spread. A horizontal plateau on the capnogram is likely to be a result of absence of expiratory gas flow at the sampling site of the carbon dioxide analyser. 16 As a result, a moderate decrease in mean ( Pa C O P E CO ) was obtained with the first prolonged expiration manoeuvre ( P E 1CO ). A steep phase III slope was observed at further stages of the study. Few studies have examined the slope of the phase III of the capnograms during thoracotomy. Our results suggest that an increased slope may be present frequently, particularly during OLV. This may be attributed to alterations in the spread of ventilation perfusion ratios within the lungs related to prolonged anaesthesia, increased vertical height of the lungs in the lateral decubitus position 19 and also OLV. 1 0 In addition, variations in the physiological deadspace fraction of tidal volume may occur during thoracic surgery, depending on many factors (especially variations in pulmonary arterial pressure 1 ) which affect distribution of ventilation and perfusion within the lungs. Both within-units and between-units ventilation perfusion mismatching may be improved by increasing tidal volume, at least if frequency is reduced at the same time. 4 The ventilatory settings used in this study were different from those in a previous study, 0 but the values of ( Pa C O P E CO ) were similar in both studies. The steep phase III slope and the dramatic decrease in mean ( Pa C O P E CO ) obtained with the simple prolonged expiration manoeuvre during thoracotomy, and particularly during OLV, strongly suggest the existence of delayed and asynchronous emptying, resulting in dynamic lung hyperinflation. Dynamic lung hyperinflation often occurs during OLV when a double-lumen endobronchial tube is used. When exhalation is not complete, the areas with high ventilation perfusion (and thus low PCO ) are the most significant contributors to P E CO, inducing ( Pa C O P E CO ) to increase. The simple prolonged expiration manoeuvre ( P E 1CO ) allows the low ventilation perfusion, high PCO areas to leave lungs and to make a more substantial contribution to endexpiratory PCO. If such a full expiration is preceded by sustained hyperinflation of the lungs ( E CO P in this study), the extra ventilation is distributed preferentially to dependent, hypoventilated but well perfused lung regions. 3 4 The wash-out of areas with ventilation perfusion ratios close to zero resulted in smaller, and frequently negative, ( Pa C O CO ) gradients during OLV. In relatively healthy, nonpregnant adults, the incidence of negative ( Pa C O P E CO ) differences during controlled ventilation is 0 1% of patients, 5 but manual compression of the thorax (providing a forced P E CO ) induces a decrease in ( Pa C O P E CO ) with a high incidence of negative gradients in patients with or without lung disease. 8 In addition, during OLV the dependent lung is small, thus each tidal breath is relatively effective in that it has a small functional residual capacity (FRC) to wash-out. This reduced FRC may have contributed to the negative differences during OLV. 5 The manoeuvres did not reduce the variations observed with P E CO both between and within patients. The intraoperative ( Pa C O P E CO ) relationship is complex and dynamic, 5 and the reduction of one factor (i.e. time-dependent ventilation perfusion mismatching) obtained with the manoeuvres did not reduce these variations. The decrease in arterial to end-expiratory difference obtained with both manoeuvres suggests that temporal deadspace (related to the phasic variations in ventilation and perfusion during controlled ventilation, and which one would anticipate to be increased during prolonged expiration manoeuvres) was less important than spatial, between- and within-units mismatching. 16 However, interindividual differences in temporal ventilation perfusion mismatching during the manoeuvres may have contributed to their poor reliability. In summary, we have confirmed that prolonged expiration manoeuvres reduced the mean arterial to peak expired carbon dioxide partial pressure difference. However, the end-expiratory PCO obtained with each manoeuvre agreed poorly with P a CO. This suggests that these manoeuvres should no longer be recommended to improve estimation of P a CO from P uring anaesthesia. E CO References 1. Nunn JF, Hill DW. Respiratory dead space and arterial to end-tidal CO tension difference in anaesthetised man. Journal of Applied Physiology 1960; 15: Whitesell R, Asiddao C, Gollman D, Jablonski J. Relationship between arterial and peak expired carbon dioxide pressure during anesthesia and factors influencing the difference. Anesthesia and Analgesia 1981; 60: Raemer DB, Francis D, Philip JH, Gabel RA. Variation in PCO between arterial blood and peak expired gas during anesthesia. Anesthesia and Analgesia 1983; 6: Fletcher R, Jonson B. Deadspace and the single-breath test for carbon dioxide during anaesthesia and artificial ventilation. British Journal of Anaesthesia 1984; 56: Bhavani-Shankar K, Moseley H, Kumar AY, Delph Y. Capnometry and anaesthesia. Canadian Journal of Anaesthesia 199; 39: Fletcher R. On-line expiratory CO monitoring. International Journal of Clinical Monitoring and Computing 1986; 3: Tulou PP, Walsh PM. Measurement of alveolar carbon dioxide tension at maximal expiration as an estimate of arterial carbon dioxide tension in patients with airway obstruction. American Review of Respiratory Disease 1970; 10: Takki S, Aromaa U, Kauste A. The validity and usefulness of the end-tidal PCO during anaesthesia. Annals of Clinical Research 197; 4: Goldman JM, Strom JS. Respiratory monitoring. In: Kirby RR, Gravenstein NG, eds. Clinical Anesthesia Practice. Philadelphia: Saunders Company, 1994; Phillips JH, Feinstein DM, Raemer DB. Monitoring anesthetic and respiratory gases. In: Blitt CD, Mines RL, eds. Monitoring in Anesthesia and Critical Care Medicine.

5 540 British Journal of Anaesthesia New-York: Churchill Livingstone, 1995; Fletcher R. Smoking, age and the arterial end-tidal PCO difference during anaesthesia and controlled ventilation. Acta Anaesthesiologica Scandinavica 1987; 31: Fletcher R. The arterial end-tidal CO difference during cardiothoracic surgery. Journal of Cardiothoracic Anesthesia 1990; 4: Ashwood ER, Kost G, Kenny M. Temperature correction of blood-gas and ph measurements. Clinical Chemistry 1983; 9: McPeak H, Palayiwa E, Madgwick R, Sykes MK. Evaluation of a multigas anaesthetic monitor: the Datex Capnomac. Anaesthesia 1988; 43: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: Fletcher R, Jonson B, Cumming G, Brew J. The concept of deadspace with special reference to the single breath test for carbon dioxide. British Journal of Anaesthesia 1981; 53: Hedenstierna G, Lundh R, Johansson H. Alveolar stability during anaesthesia for reconstructive vascular surgery in the leg. Acta Anaesthesiologica Scandinavica 1983; 7: Pansard JL, Cholley B, Devilliers C, Clergue F, Viars P. Variation in arterial to end-tidal CO tension differences during anesthesia in the kidney rest lateral decubitus position. Anesthesia and Analgesia 199; 75: Rehder K, Knopp TJ, Sessler AD. Regional intrapulmonary gas distribution in awake and anesthetized paralyzed prone man. Journal of Applied Physiology 1978; 45: Ip Yam PC, Innes PA, Jackson M, Snowdon SL, Russel GN. Variation in the arterial to end-tidal PCO difference during one-lung thoracic anaesthesia. British Journal of Anaesthesia 1994; 7: Werner O, Malmkvist G, Beckman A, Stahle S, Nordström L. Carbon dioxide elimination from each lung during endobronchial anaesthesia. British Journal of Anaesthesia 1984; 56: Yokota K, Toriumi T, Sari A, Endou S, Mihira M. Autopositive end-expiratory pressure during one-lung ventilation using a double-lumen endobronchial tube. Anesthesia and Analgesia 1996; 8: Hatle L, Rokseth R. The arterial to end-expiratory carbon dioxide tension gradient in acute pulmonary embolism and other cardiopulmonary diseases. Chest 1974; 66: Rothen HU, Sporre B, Engberg G, Wegenius G, Hogman M, Hedenstierna G. Influence of gas composition on recurrence of atelectasis after a re-expansion maneuver during general anesthesia. Anesthesiology 1995; 8: Good M. Capnography: uses, interpretation and pitfalls. In: Barash P, eds. ASA Refresher Lectures. Pennsylvania: JB Lippincott, 1990;

Capnography. Capnography. Oxygenation. Pulmonary Physiology 4/15/2018. non invasive monitor for ventilation. Edward C. Adlesic, DMD.

Capnography. Capnography. Oxygenation. Pulmonary Physiology 4/15/2018. non invasive monitor for ventilation. Edward C. Adlesic, DMD. Capnography Edward C. Adlesic, DMD University of Pittsburgh School of Dental Medicine 2018 North Carolina Program Capnography non invasive monitor for ventilation measures end tidal CO2 early detection