Veterinary Anaesthesia and Analgesia, 2015, 42,

Transcription

1 Veterinary Anaesthesia and Analgesia, 2015, 42, doi: /vaa RESEARCH PAPER The effect of the inspired oxygen fraction on arterial blood oxygenation in spontaneously breathing, isoflurane anaesthetized horses: a retrospective study Stijn Schauvliege*, Ioannis Savvas & Frank Gasthuys* *Department of Surgery and Anaesthesia of Domestic Animals, Faculty of Veterinary Medicine, University of Gent, Merelbeke, Belgium Anaesthesiology and Intensive Care Unit, Companion Animal Clinic, Department of Clinical Sciences, School of Veterinary Medicine, Aristotle University of Thessaloniki, Thessaloniki, Greece Correspondence: Ioannis Savvas, Companion Animal Clinic, Department of Clinical Sciences, School of Veterinary Medicine, Aristotle University of Thessaloniki, St. Voutyra 11, GR Thessaloniki, Greece. isavas@vet.auth.gr Abstract Objectives To investigate the influence of two inspired oxygen fractions (FIO 2 ) on the arterial oxygenation in horses anaesthetized with isoflurane. Study Design Retrospective, case-control clinical study. Animals Two hundred equine patients undergoing non-abdominal surgery (ASA class 1 2), using a standardized anaesthetic protocol and selected from anaesthetic records of a period of three years, based on pre-defined inclusion criteria. Methods In group O (n = 100), medical oxygen acted as carrier gas, while in group M (n = 100), a medical mixture of oxygen and air (FIO ) was used. Demographic data, FIO 2, arterial oxygen tension (PaO 2 ) and routinely monitored physiologic data were recorded. The alveolar-arterial oxygen tension difference [P(A-a)O 2 ] and PaO 2 /FIO 2 ratio were calculated. The area under the curve, standardized to the anaesthetic duration, was calculated and statistically compared between groups using t-tests or Mann Whitney tests as appropriate. Categorical data were compared using Chi-square tests. Results No significant differences in age, body weight, sex, breed, surgical procedure, position, anaesthetic duration or arterial carbon dioxide tension were found. Mean FIO 2 was 0.78 in group O and 0.60 in group M. Compared to group O, significantly lower values for PaO 2 and for P(A-a)O 2 were found in group M. In contrast, the PaO 2 /FIO 2 ratio and the percentage of horses with a PaO 2 <100 mmhg (13.33 kpa) were comparable in both groups. Conclusions Although a reduction of the inspired oxygen fraction resulted in a lower PaO 2, the P(A-a) O 2 was also lower and the number of horses with PaO 2 values <100 mmhg was comparable. Clinical relevance In healthy isoflurane anaesthetized horses, the use of a mixture of oxygen and air as carrier gas seems acceptable, but further, prospective studies are needed to confirm whether it results in a lower degree of ventilation/perfusion mismatching. Keywords arterial oxygen tension, horse, inspired oxygen fraction. Introduction In anaesthetized horses, hypoxaemia is observed commonly due to ventilation-perfusion inequality 280

2 and right-to-left shunting of pulmonary blood, in turn caused by early atelectasis formation in dependent lung regions (Nyman et al. 1990). For this reason, 100% medical oxygen traditionally has been used as carrier gas during equine anaesthesia, with the aim to optimize oxygen uptake in ventilated lung regions as a compensation for atelectatic areas (Kerr & McDonell 2009). Nevertheless, the arterial oxygen tension (PaO 2 ) often remains low because the arterial oxygen content depends on the binding of oxygen to haemoglobin. In well-ventilated areas, the saturation of haemoglobin with oxygen is usually 100%, even when breathing room air. When using a higher inspired oxygen fraction, very little additional oxygen can be dissolved in plasma compared to the amount that is bound to haemoglobin. Consequently, increasing the inspired fraction of oxygen (FIO 2 ) results only in a very limited increase in arterial oxygen content in well-ventilated areas of the lung, which cannot (or only minimally) compensate for atelectatic areas when a significant degree of pulmonary shunt is present. This has been clearly illustrated in man in the so called isoshunt diagrams (Benatar et al. 1973; Kerr & McDonell 2009). It has been shown repeatedly, however, that an FIO 2 of 1.0 in the peri-operative period can result in more severe pulmonary atelectasis, so its routine use has been questioned. Atelectasis develops regularly during general anaesthesia in humans, with a reported prevalence of up to 75 90% in healthy patients during spontaneous or artificial ventilation. Collapsed lung fields may comprise 10 50% of the total lung tissue and this condition may persist for up to two days post-operatively (Hedenstierna & Rothen 2000; Magnusson & Spahn 2003; Duggan & Kavanagh 2005; Lumb 2010a; Edmark et al. 2011). Several mechanisms have been proposed to cause or contribute to the development of atelectasis, including lung compression, gas absorption, and surfactant impairment (Magnusson & Spahn 2003). The pressure exerted by abdominal organs in anaesthetized recumbent patients not only tends to compress the lungs ( compression atelectasis ), but at the same time, the functional residual capacity often drops below the closing volume of the lung, resulting in closure of small airways. Resorption of gases from alveoli distal to these collapsed airways then causes the so-called resorption atelectasis (Lumb 2010a). During anaesthesia, the elevated FIO 2 increases the rate of gas absorption from occluded alveoli considerably. Furthermore, such increased gas absorption also may occur in lung fields that are not occluded but which have a low ventilation/perfusion ratio, and as a result the lung unit becomes progressively smaller. With both mechanisms, lung unit collapse will probably occur and atelectasis will develop (Magnusson & Spahn 2003; Duggan & Kavanagh 2005). The administration of an FIO 2 of is currently suggested in human anaesthetic practice, unless the arterial blood oxygenation is compromised (Hedenstierna & Rothen 2000; Lumb 2010a). In mechanically ventilated dogs, lung aeration and gas exchange were reported to be significantly better with an FIO 2 of 0.40 compared to an FIO 2 of 1.0 (Staffieri et al. 2007). Increased atelectasis development has also been reported when using 100% oxygen in anaesthetized animals, including cats (Staffieri et al. 2010a), sheep (Staffieri et al. 2010b) and horses. In halothane anaesthetized horses, an FIO 2 >0.85 was associated with an increased alveolar/arterial oxygen tension difference, although the PaO 2 remained constant (Cuvelliez et al. 1990). During tiletamine-zolazepam anaesthesia (spontaneous breathing, left lateral recumbency), arterial blood oxygenation was higher when horses inhaled >95% oxygen, than when they inhaled air. Interestingly however, breathing air decreased intrapulmonary shunt and reduced hypoventilation (Marntell et al. 2005). In mechanically ventilated, laterally recumbent horses, the inhalation of a gas mixture with a high helium-oxygen ratio was reported better to preserve lung function than the administration of high concentrations of oxygen during isoflurane anaesthesia. Moreover, a step-wise increase of the FIO 2 causes less impairment of pulmonary gas exchange than administration of a high FIO 2 immediately after induction of anaesthesia (Staffieri et al. 2009). The ideal FIO 2 that optimizes arterial oxygenation and minimizes formation of resorption atelectasis has not been determined in horses. Some authors reported that the use of 30% (Levionnois & Kuich 2008) or 50% (Hubbell et al. 2011) oxygen did not improve arterial oxygenation or gas exchange compared to the use of >90% oxygen as carrier gas. In 2005, following the emerging evidence from medical and veterinary literature, our large animal clinic adopted a practice of reduced FIO 2 (60% oxygen) in all clinical cases of equine surgery (both elective and emergency operations). With the present retrospective study, the effect of this transition from 100% oxygen as the carrier gas (resulting most usually in 2014 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 42,

3 an FIO 2 of >80%) to a mixture of oxygen/air as a carrier gas on the arterial blood oxygenation and pulmonary function during anaesthesia was investigated in a clinical setting. Materials and methods In this retrospective, case-control study, anaesthetic records over a period of three years (about 3600 cases) were evaluated. This period included the transition point from the use of 100% oxygen to a mixture of oxygen/air as a carrier gas. Inclusion criteria for a case were: age 1 15 years, weight kg, physical status ASA 1 2, spontaneous breathing, premedication with acepromazine (20 lg kg 1 ), romifidine (80 lg kg 1 ), and morphine (0.1 mg kg 1 ), induction of anaesthesia with midazolam (0.06 mg kg 1 ) and ketamine (2.2 mg kg 1 ), maintenance of anaesthesia with isoflurane, non-abdominal surgery, and anaesthetic duration of at least 20 minutes. A final number of 200 cases were selected. In 100 of these cases, 100% oxygen was used as carrier gas (group O), in the other group an air/oxygen mixture was used (group M). Demographic data, FIO 2, PaO 2, arterial carbon dioxide tension (PaCO 2 ), respiratory rate (f R ), heart rate (HR), systolic and diastolic arterial blood pressures (SAP, DAP), end-tidal expired carbon dioxide (PE CO 2 ), and end-tidal concentration of isoflurane (FE iso) were recorded. Arterial blood samples were drawn anaerobically from the facial artery, using a pre-heparinized syringe, the analysis was performed immediately (ABL5, Radiometer, Denmark). The alveolar-arterial difference of the partial pressure of oxygen [P(A-a)O 2 = (barometric pressure water vapour pressure) 9 FIO 2 PaCO 2 /0.9 PaO 2 (Nyman et al. 1990)] and the ratio of the arterial partial pressure of oxygen to the inspired fraction of oxygen (PaO 2 /FIO 2 ) were calculated. Summary measures for serial measurements were used (Matthews et al. 1990). In particular, the area under the curve (AUC) was calculated using the trapezoidal method for each of the above variables throughout anaesthesia, for each animal. Then the AUC was standardized (divided) by the total anaesthetic duration, producing one single value (AUCst) for each variable in each horse. These values were used for the statistical processing thereafter. The percentage of horses with a PaO 2 <100 mmhg (13.33 kpa) on at least one occasion was also determined. For statistical analysis the Kolmogorov- Smirnoff test of normality was used to evaluate the normality of the distribution of the data. Depending on normality, the t-test or the Mann Whitney test were used to evaluate any significant difference of the calculated variables between the groups. Crosstabulation and a chi-square test of association were used between categorical variables. Linear regression analysis was also performed to predict P(A-a)O 2 from independent variables (FIO 2 and body weight). All statistical calculations were performed with a computer software (SPSS 19, IBM company, IL, USA). Results Demographic data of the horses included in the present study are represented in Tables 1 and 2. Data from 73 females and 127 males were collected. Four animals underwent dental, 10 diagnostic, 8 ophthalmic, 29 orthopaedic, and 149 soft tissue procedures. Sixty-six animals were placed in dorsal, 48 in left lateral, and 86 in right lateral position. One hundred and sixty-seven cases were of physical status ASA 1 while 33 horses were assessed as ASA 2. The animals were distributed homogeneously between the two groups with regard to age (Mann Whitney test, p = 0.169), body weight (t-test, p = 0.102), sex (chi-square test, p = 0.883), breed (chi-square test, p = 0.675), surgical procedure (chisquare test, p = 0.655), position (chi-square test, p = 0.254), and duration of anaesthesia (Mann Whitney test, p = 0.273). The anaesthetic duration was (mean SD) minutes in group O and minutes in group M. The AUCst PaCO 2 was 56 6 mmhg ( kpa) in group O and 55 5 mmhg Table 1 Distribution of breeds of 200 animals in this retrospective study Breed Warmblood horses 143 Standardbreds 19 Thoroughbreds/Arabians/Anglo-Arabians 12 Andalusians/Lusitano s 4 Draught Horses 4 Irish Tinker horses 4 Ponies 4 Friesians 3 Haflingers 3 Quarter horses 3 Tennessee walking horse 1 Total 200 Frequency Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 42,

4 Table 2 Descriptive statistics of age and weight of the animals in 2 groups of isoflurane anaesthetized horses. In group O 100% oxygen was used as the carrier gas (average FIO 2 of 0.78), in group M a mixture of oxygen and air (average FIO ) Group O Weight (kg) Mean SD Minimum Maximum Age (years) Mean SD Minimum 1 1 Maximum ( kpa) in group M, the difference being statistically non-significant (t test, p = 0.275). The AUCstf R was breaths minute 1 in group O and in group M, the difference being statistically non-significant (Mann Whitney U test, p = 0.621). The AUCstFIO 2 was and in groups O and M respectively (Mann Whitney U test, p < ). The mean AUCstPaO 2 was mmhg ( kpa) in group O and mmhg ( kpa) in group M, the difference being statistically significant (Mann Whitney U test, p = 0.01). In contrast, no significant difference was found in the percentage of horses where a PaO 2 <100 mmhg (13.33 kpa) was detected on at least one occasion (group O 8%, group M 11%) (p = 0.234). The mean AUCstP(A-a)O 2 was significantly higher (t test, p < ) in group O ( mmhg ( kpa)) than in group M ( mmhg ( kpa)). The mean AUCstPaO 2 /FIO 2 was mmhg ( kpa) in group O and mmhg ( kpa) in group M, however the difference was statistically non-significant (Mann Whitney U test, p = 0.337). The AUCstFIO 2 correlated significantly, although weakly, with AUCstP(A-a)O 2 (p < , r 2 = 0.276, effect size = 4.183, standard error = 0.482, constant = ). In other words, for an increase of 1% of FIO 2 an increase of mmhg (95% confidence interval ) in the alveolar-arterial difference would be expected. However, AUCstFIO 2 did not correlate significantly with M AUCstPaO 2 /FIO 2 (p = 0.867). There was also a significant but very weak correlation between AUCstP(A-a)O 2 and body weight (p < , r 2 = 0.111, effect size = 0.331, standard error = 0.067, constant = 84.42). Discussion The PaO 2 was more than 100 mmhg (13.33 kpa) throughout anaesthesia in about 90% of the horses in both groups. Nevertheless, the values for PaO 2 were quite low when compared to theoretically expected values, while the P(A-a)O 2 was high and the PaO 2 /FIO 2 ratio low. These observations clearly confirm the presence of ventilation/perfusion mismatching, in the form of increased scatter, of true shunt or of both. However, at the relatively high FIO 2 s used in the present study, the influence of scatter on the P(A-a)O 2 and PaO 2 /FIO 2 is usually limited. More likely, a certain degree of true shunt, caused by pulmonary atelectasis, was present in both groups. Such atelectasis formation does not only cause arterial hypoxaemia, but may also result in decreased lung compliance, increased pulmonary vascular resistance, and eventually lung injury (Magnusson & Spahn 2003; Duggan & Kavanagh 2005). Comparison of the P(A-a)O 2 and PaO 2 /FIO 2 values between the two groups showed that P(A-a)O 2 was higher and PaO 2 /FIO 2 ratio lower in group O than in group M, but only the difference in P(A-a)O 2 was significant. Additionally, a significant correlation was found between the FIO 2 and the P(A-a)O 2. The P (A-a)O 2 is an index of venous admixture (caused by pulmonary shunting or ventilation/perfusion scatter), but can be affected by several factors, such as the alveolar oxygen tension (PAO 2 ), the cardiac output, the body temperature, ph and base excess of the blood, the haemoglobin concentration, and the alveolar ventilation (Lumb 2010b). Moreover, the sigmoid shape of the oxyhaemoglobin dissociation curve does not allow for a linear relation of the alveolar/arterial oxygen difference with venous admixture (Cuvelliez et al. 1990; Lumb 2010b). Additionally, the body position, body weight and thoracic conformation can also affect the oxygenation of anaesthetized horses (Moens et al. 1995; Mansel & Clutton 2008). In the present study, a weak but significant correlation was found also between the body weight and P(A-a)O 2. However, the anaesthetic protocol was the same in all horses and no differences in age, sex, bodyweight, breed, 2014 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 42,

5 procedure, position, duration of anaesthesia and PaCO 2 were found. It can therefore be concluded that, most likely, the higher value for P(A-a)O 2 in group O was mainly, if not completely, attributable to the increased FIO 2 and represented a higher degree of venous admixture in this group. The PaO 2 /FIO 2 ratio, a second index of venous admixture, was somewhat lower in group O but the difference between the two groups was not significant, which seems to contradict the findings for P(A-a)O 2. It is well known that an increase in the PAO 2 (or FIO 2 ) causes a larger increase of the P(A-a)O 2 than of the PaO 2 /FIO 2 ratio. It could therefore be hypothesized that the higher P(A-a)O 2 values in group O were not caused by a larger degree of venous admixture but were simply a result of increasing the FIO 2. However, it has been shown both theoretically and experimentally, by comparing model simulations with measured values of the PaO 2 /FIO 2 ratio in human patients, that not only the P(A-a)O 2 but also the PaO 2 /FIO 2 ratio depends on the FIO 2, as well as on the SaO 2 levels (Karbing et al. 2007). Furthermore, the influence of the FIO 2 on the P(A-a)O 2 becomes smaller when the FIO 2 is above 0.60 (Harris et al. 1974). Since the FIO 2 was above this value in both groups of the present study, it seems likely that the difference in P (A-a)O 2 was at least partly attributable to a higher degree of venous admixture. It can be stated cautiously that the use of a mixture of oxygen and air may result in less atelectasis formation in anaesthetized horses under clinical circumstances. Further studies are however needed to confirm this observation and would preferably include calculation of the degree of venous admixture using the shunt equation and/or calculation of the F-shunt. One of the limitations of this retrospective study is that such calculations could not be made, because packed cell volume was not measured consistently for each sample and no mixed venous blood gas values were available. A second and major limitation of the present study is that only spontaneously breathing horses were included. Possibly, several horses that were hypoxaemic were excluded because mechanical ventilation was initiated at a certain point during the course of the anaesthesia, in an attempt to improve oxygenation. This may have falsely elevated the mean PaO 2 values in both groups. Regardless of the degree of mismatching, it must be noted that in group M, the mean values for PaO 2 were lower than in group O, but still well above the range to fully saturate haemoglobin. The difference in arterial oxygen content between the two groups should therefore be minimal, suggesting that the use of an FIO 2 of 0.60 does not automatically result in a clinically relevant reduction in oxygen delivery to the tissues. However, a reduction of FIO 2 cannot be advocated when the respiratory system is compromised. It also is interesting to note that in humans, an increased FIO 2 intra- and post-operatively reduced the incidence of postoperative nausea, vomiting and wound infection (Magnusson & Spahn 2003; Duggan & Kavanagh 2005). In conclusion, the results of the present study show that a reduction of the FIO 2 from an average of 0.78 to an average of 0.60 resulted in a significant decrease of the alveolar-arterial oxygen difference in anesthetized horses in a clinical setting. Although the mean PaO 2 was also lower in the horses receiving a high percentage of oxygen, it was still well above the range to fully saturate haemoglobin. These findings are encouraging for future research in this area. References Benatar SR, Hewlett AM, Nunn JF (1973) The use of isoshunt lines for control of oxygen therapy. Br J Anaesth 45, Cuvelliez SG, Eicker SW, McLauchlan C et al. (1990) Cardiovascular and respiratory effects of inspired oxygen fraction in halothane-anesthetized horses. Am J Vet Res 51, Duggan M, Kavanagh BP (2005) Pulmonary atelectasis: a pathogenic perioperative entity. Anesthesiology 102, Edmark L, Auner U, Enlund M et al. (2011) Oxygen concentration and characteristics of progressive atelectasis formation during anaesthesia. Acta Anaesthesiol Scand 55, Harris EA, Kenyon AM, Nisbet HD et al. (1974) The normal alveolar-arterial oxygen-tension gradient in man. Clin Sci Mol Med 46, Hedenstierna G, Rothen HU (2000) Atelectasis formation during anesthesia: causes and measures to prevent it. J Clin Monit Comput 16, Hubbell JAE, Aarnes TK, Bednarski RM et al. (2011) Effect of 50% and maximal inspired oxygen concentrations on respiratory variables in isoflurane-anesthetized horses. BMC Vet Res 7, 23. Karbing DS, Kjærgaard S, Smith BW et al. (2007) Variation in the PaO2/FiO2 ratio with FiO2: mathematical and experimental description, and clinical relevance. Crit Care 11, R Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 42,

6 Kerr CL, McDonell WN (2009) Oxygen Supplementation and Ventilatory Support. In: Equine Anesthesia Monitoring and Emergency Therapy. Muir WW, Hubbell JAE (eds). Saunders Elsevier, St. Louis, MO. pp Levionnois OL, Kuich N (2008) Lower intra-operative oxygen inspiration fraction do not improve arterial oxygenation after isoflurane anaesthesia with mechanical ventilation in clinical horses. In: AVA Autumn Meeting. Barcelona, pp Lumb AB (2010a) Anaesthesia. In: Nunn s Applied Respiratory Physiology (7th edn). Lumb AB (ed). Churchill Livingston Elsevier, Edinburgh. pp Lumb AB (2010b) Oxygen. In: Nunn s Applied Respiratory Physiology (7th edn). Lumb AB (ed). Churchill Livingston Elsevier, Edinburgh. pp Magnusson L, Spahn DR (2003) New concepts of atelectasis during general anaesthesia. Br J Anaesth 91, Mansel JC, Clutton RE (2008) The influence of body mass and thoracic dimensions on arterial oxygenation in anaesthetized horses and ponies. Vet Anaesth Analg 35, Marntell S, Nyman G, Hedenstierna G (2005) High inspired oxygen concentrations increase intrapulmonary shunt in anaesthetized horses. Vet Anaesth Analg 32, Matthews JN, Altman DG, Campbell MJ et al. (1990) Analysis of serial measurements in medical research. BMJ 300, Moens Y, Lagerweij E, Gootjes P et al. (1995) Distribution of inspired gas to each lung in the anaesthetised horse and influence of body shape. Equine Vet J 27, Nyman G, Funkquist B, Kvart C et al. (1990) Atelectasis causes gas exchange impairment in the anaesthetised horse. Equine Vet J 22, Staffieri F, Franchini D, Carella GL et al. (2007) Computed tomographic analysis of the effects of two inspired oxygen concentrations on pulmonary aeration in anesthetized and mechanically ventilated dogs. Am J Vet Res 68, Staffieri F, Bauquier SH, Moate PJ et al. (2009) Pulmonary gas exchange in anaesthetised horses mechanically ventilated with oxygen or a helium/oxygen mixture. Equine Vet J 41, Staffieri F, De Monte V, De Marzo C et al. (2010a) Effects of two fractions of inspired oxygen on lung aeration and gas exchange in cats under inhalant anaesthesia. Vet Anaesth Analg 37, Staffieri F, Driessen B, Monte VD et al. (2010b) Effects of positive end-expiratory pressure on anesthesia-induced atelectasis and gas exchange in anesthetized and mechanically ventilated sheep. Am J Vet Res 71, Received 26 November 2013; accepted 12 March Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 42,