Comparison of cardiorespiratory variables in dorsally recumbent horses anesthetized with guaifenesin-ketamine-xylazine spontaneously breathing 50% or maximal oxygen concentrations

Abstract

This study compared cardiorespiratory variables in dorsally recumbent horses anesthetized with guaifenesin-ketamine-xylazine and spontaneously breathing 50% or maximal (> 90%) oxygen (O₂) concentrations. Twelve healthy mares were randomly assigned to breathe 50% or maximal O₂ concentrations. Horses were sedated with xylazine, induced to recumbency with ketamine-diazepam, and anesthesia was maintained with guaifenesin-ketamine-xylazine to effect. Heart rate, arterial blood pressures, respiratory rate, lithium dilution cardiac output (CO), inspired and expired O₂ and carbon dioxide partial pressures, and tidal volume were measured. Arterial and mixed-venous blood samples were collected prior to sedation (baseline), during 30 minutes of anesthesia, 10 minutes after disconnection from O₂, and 30 minutes after standing. Shunt fraction, O₂ delivery, and alveolar-arterial O₂ partial pressures difference [P(A-a)O₂] were calculated. Recovery times were recorded. There were no significant differences between groups in cardiorespiratory parameters or in P(A-a)O₂ at baseline or 30 minutes after standing. Oxygen partial pressure difference in the 50% group was significantly less than in the maximal O₂ group during anesthesia.

Introduction

Arterial oxygen partial pressures (PaO₂) less than those seen in standing horses are commonly encountered in horses during inhalant anesthesia despite augmentation of the fraction of inspired oxygen (FiO₂) (1–3). Low PaO₂ (< 100 mmHg) develops within the first 30 to 90 min of anesthesia, persists throughout the anesthetic event, and resolves within 60 min of standing (2). Potential causes of low or subnormal PaO₂ include: decreased FiO₂, diffusion impairment, vascular shunt, hypoventilation, and ventilation-perfusion mismatching (1,2). Low FiO₂ is unlikely when oxygen (O₂) is supplemented and the alveolar–capillary membrane is intact. Anatomic vascular shunts do not develop in exercising horses and thus are unlikely to be present in anesthetized horses. Hypoventilation (PaCO₂ > 45 mmHg) occurs commonly due to the effects of anesthetic drugs, particularly inhalants, and can be corrected via mechanical ventilation (4). The early use of mechanical ventilation in conjunction with increased FiO₂ decreases the incidence of low PaO₂ but is inconsistent in resolving low PaO₂ when it exists (3,5,6).

Ventilation-perfusion (V/Q) mismatches are major contributors to decreased oxygenation in the anesthetized horse. Functional residual capacity (FRC) and residual volume are reduced during inhalant anesthesia (1) and alternations in diaphragmatic function and mechanical ventilation shift the distribution of ventilation to nondependent lung fields (7–11). Atelectasis is associated with “pseudo” or physiologic shunt, characterized by collapsed alveoli that are perfused but not ventilated. There are 2 primary causes of atelectasis: compression and absorption. Compression atelectasis occurs when the weight of abdominal viscera is positioned over the thorax when horses are in dorsal or lateral recumbency and is exacerbated by the oblique nature of the equine diaphragm (12). Mechanical ventilation, if implemented early, may delay or partially offset compression, but does not always achieve expected PaO₂ values (10,13,14). Absorption atelectasis causes alveolar collapse due to absorption of alveolar gas (15).

Maximal FiO₂ (> 95%) exacerbates absorption atelectasis due to the increased absorption of O₂ compared to less soluble gases present in ambient air, primarily nitrogen. In humans, studies have shown that maximal FiO₂ contributes to alveolar damage, reduction in cardiac index, and increases in peripheral vascular resistance (15,16). The use of 50% FiO₂ has the potential to reduce the incidence of absorption atelectasis compared with maximal FiO₂ while providing greater FiO₂ than ambient air (6,17). If absorption atelectasis is a significant component of V/Q mismatch in the horse, oxygenation and delivery of oxygen (DO₂) could be improved using a 50% FiO₂.

The purpose of this study was to compare the cardiorespiratory effects of 50% FiO₂ to maximal O₂ concentrations in spontaneously breathing horses during anesthesia due to injected agents. We hypothesized that 50% FiO₂ would result in less V/Q mismatch, increased PaO₂, and improved DO₂.

Materials and methods

Animals

The study population consisted of 12 healthy mares (8 Quarterhorses, 2 Thoroughbreds, and 2 Standardbreds). Mean age was 13.5 y (range: 5 to 25 y) and mean weight was 527 kg [standard deviation (SD): 46.5 kg]. Physical examination, complete blood cell count, and serum biochemistry were performed prior to inclusion in the study. This study was approved by The Ohio State University Animal Care and Use Committee.

Study design

Food, but not water, was withheld from horses for approximately 6 h prior to anesthesia and surgery. Each horse was anesthetized once for incisional biopsies of subcutaneous adipose tissue. All surgeries were performed by the same experienced surgeons. On the day of the experiment, hair over the jugular veins and a transverse facial artery was clipped and the skin was aseptically prepared for placement of IV and intra-arterial catheters. Lidocaine 2% (Lidocaine; Vedco, St. Joseph, Missouri, USA), 1 mL/site, was injected SC over the jugular veins. A 14-gauge 8.3 cm catheter (Angiocath; Becton Dickinson, Sandy, Utah, USA) was inserted in the left jugular vein for administration of anesthetic drugs. An 8 F catheter introducer (Catheter introducer CL-07811; Arrow International, Reading, Pennsylvania, USA) was placed in the right jugular vein to facilitate the placement of a 110-cm polyethylene 240 catheter (Intramedic PE-240 tubing; Becton Dickinson and Co, Sparks, Maryland, USA), which was positioned such that its distal tip lay in the right atrium for collection of mixed-venous blood. A 20-gauge catheter (Surflo catheter; Terumo Medical Corp, Elkton, Maryland, USA) was percutaneously inserted in a transverse facial artery for measurement of systolic arterial blood pressure (SABP), diastolic ABP (DABP), and mean ABP (MABP), collection of heparinized arterial blood samples for analysis of pH and blood gases, and for blood withdrawal for cardiac output monitoring by lithium dilution (18). Positioning of the right atrial and arterial catheters was confirmed by the visualization of characteristic cardiac chamber or vessel pressure waveforms by attaching the catheter to an appropriately calibrated pressure transducer (TruWave Disposable Pressure Transducer; Edwards Lifesciences, Irvine, California, USA). The scapulohumeral joint and the sternum were used as zero-pressure reference points when the horses were standing and in dorsal recumbency, respectively.

Horses were left undisturbed for ≥ 10 min after catheter placement. Baseline values of heart rate (HR) and respiratory rate (RR) were then measured. A mixed-venous blood sample was obtained by withdrawal of blood from the right atrial catheter, and an arterial blood sample was collected via withdrawal from the transverse facial artery. Arterial and mixed-venous blood samples were obtained for determination of baseline values for the following variables: pH, hemoglobin, blood gases, lactate, and oxygen saturation (SO₂) (ABL 725; Radiometer America, Westlake, Ohio, USA). Five minutes after baseline values were obtained, xylazine (Xylazine; Vedco), 1.0 mg/kg body weight (BW), IV, was administered for sedation and horses were moved to a padded induction stall and positioned behind a restraining door for induction of anesthesia. Five minutes after xylazine administration, ketamine (Ketaset; Fort Dodge Animal Health, Fort Dodge, Iowa, USA) 2.2 mg/kg BW and diazepam (Diazepam; Hospira, Lake Forest, Illinois, USA), 0.1 mg/kg BW, were administered as an IV bolus. Horses were orotracheally intubated with a cuffed endotracheal tube (internal diameter, 26 mm) and positioned in dorsal recumbency on a padded large animal surgery table. Horses were moved into the surgical suite and the endotracheal tube was connected to a standard large animal circle anesthetic machine (Mallard Medical, Redding, California, USA) that had been primed for 5 min with the fresh gas mixture to be tested at a total flow rate of 6 L/min. One of two O₂:air fresh gas mixtures was randomly administered: Group 1 — 100% O₂, no air; Group 2 — a mixture of O₂ and air to ensure an FiO₂ of 50%. Horses breathed spontaneously for the duration of the experiment. Anesthesia was maintained using a mixture of guaifenesin (Vedco) 50 g/L with ketamine (1 g/L of guaifenesin), and xylazine (500 mg/L of guaifenesin) at a rate sufficient to ensure a surgical plane of anesthesia. A base-apex ECG (Datascope Passport Model EL; Datascope, Paramus, New Jersey, USA) was used to determine heart rate and rhythm. Respiratory rate was determined by observing chest excursions for 1 min. Tidal volume was measured by attaching a digital respirometer (Respirometer Model 00-295; Anesthesia Associates, San Marcos, California, USA). A sidestream airway gas analyzer (Poet IQ2 Model 8500Q; Criticare Systems, Waukesha, Wisconsin, USA) was utilized for determination of inspired and expired percentage of O₂ and partial pressure of carbon dioxide (CO₂)

Horses were continuously observed for any response to positioning, manipulation, or surgical stimulation during the adipose tissue biopsies. Ketamine (0.2 to 0.4 mg/kg BW, IV) was administered if a horse moved its head or limbs during the procedures. No horses received positive inotropes. Fifteen and 30 min after induction of anesthesia the following measurements were recorded: HR, SABP, MABP, and DABP, cardiac output by lithium dilution, respiratory rate, tidal volume, inspired and expired O₂ concentration, and inspired and peak expired CO₂ concentration. Heparinized samples of arterial and mixed-venous blood were anaerobically collected for determination of pH, blood gases, hemoglobin, and O₂ saturation. Samples were placed on ice after collection and analyzed within 10 min.

Horses were transported 3 m and hoisted from the table to a padded recovery stall and positioned in right lateral recumbency. Ambient light and noise were minimized. Xylazine (0.2 mg/kg BW, IV) was administered at the time the infusion was discontinued to provide sedation during recovery. Arterial and mixed-venous blood samples were collected 10 min after the horses were moved to recovery and 30 min after the horses stood. The endotracheal tube was removed when horses resumed spontaneous swallowing. Time to removal of the endotracheal tube and time to standing were recorded from the time the horse was disconnected from the anesthetic circle. The number of attempts to stand was recorded.

Derived variables calculated from measured data include: cardiac index (CI; L/kg BW/min), shunt fraction (Qs/Qt; %), arterial and mixed-venous O₂ content (CaO₂ and CvO₂, respectively; vol%), O₂ delivery/kg BW (DO₂; mL O₂/kg BW/min), arterial-mixed venous O₂ content difference [C(a-v)O₂; vol%], O₂ extraction ratio (O₂ER; %), alveolar dead space (Vd/Vt; %) venous O₂ consumption (VO₂; L/min), alveolar arterial O₂ partial pressures difference (P(A-a)O₂; mmHg), and arterial partial pressure of O₂/inspired partial pressure of O₂ ratio (PaO₂-IO₂ ratio; mmHg). Values were calculated according to the following equations:

$$\begin{array}{l}\text{CI\hspace{0.17em}}(\text{L}/\text{kg\hspace{0.17em}BW}/\text{min}):\text{CO}/\text{BW}\\ {\text{CaO}}_2\hspace{0.17em}(\text{vol}\%):\hspace{0.17em}(1.39*\text{Hb}*{\text{SaO}}_2+0.003{\text{PaO}}_2)\\ {\text{CvO}}_2\hspace{0.17em}(\text{mL}/\text{Lvol}\%):(1.39*\text{Hb}*{\text{SvO}}_2+0.003{\text{PvO}}_2)\\ {\text{O}}_2\text{ER\hspace{0.17em}}(\%):{\text{VO}}_2*100/{\text{DO}}_2\\ {\text{VO}}_2\hspace{0.17em}(\text{mL}/\text{kg\hspace{0.17em}BW}/\text{min}):({\text{CaO}}_2-{\text{CvO}}_2)*\text{CI}/1000\\ (\text{C}(\text{a-v}){\text{O}}_2)\hspace{0.17em}(\text{vol}\%):{\text{CaO}}_2-{\text{CvO}}_2\\ {\text{PaO}}_2:{\text{InspO}}_2\hspace{0.17em}(\text{mmHg}):{\text{PaO}}_2/{\text{InspO}}_2\\ {\text{PAO}}_2\hspace{0.17em}(\text{mmHg}):({\text{FIO}}_2*(\text{PB}-{\text{PH}}_2\text{O})-(1.2*{\text{PaCO}}_2)\\ \text{P}(\text{A-a}){\text{O}}_2\hspace{0.17em}(\text{mmHg}):{\text{PAO}}_2-{\text{PaO}}_2\\ {\text{DO}}_2\hspace{0.17em}(\text{mL\hspace{0.17em}}{\text{O}}_2/\text{kg\hspace{0.17em}BW}/\text{min}):{\text{CaO}}_2*\text{CI}/1000\\ \text{Qs}/\text{Qt\hspace{0.17em}}(\%):({\text{CcO}}_2-{\text{CaO}}_2)/({\text{CcO}}_2-{\text{CvO}}_2)\\ \text{Vd}/\text{Vt\hspace{0.17em}}(\%,\hspace{0.17em}\text{alveolar\hspace{0.17em}dead\hspace{0.17em}space}):[({\text{PaCO}}_2-{\text{PETCO}}_2)/{\text{PaCO}}_2]*100\end{array}$$

Analysis of data

Cardiovascular and blood data were determined to be normally distributed (Kolmogorov-Smirnov normality test). A 2-way analysis of variance (ANOVA) and Tukey-Kramer post-test were used to evaluate effects of time and treatment on the variables of interest. Values of P < 0.05 were considered significant. Hemodynamic and pulmonary data are reported as mean ± SD. Recovery times were normally distributed and analyzed using an unpaired t-test. Attempts to stand were analyzed using a Mann-Whitney rank sum test.

Results

The hemodynamic and patient variables are shown in Table 1, the blood variables are shown in Table 2, and the oxygen variables are presented in Table 3. There was no significant difference in HR, RR, CO, CI, or arterial blood pressures between groups. There was no significant difference in HR and RR between groups prior to anesthesia, during the 30-minute anesthetic period, or during or after recovery. There was no significant difference in TV or end-tidal CO₂ over time during the anesthetic period or between groups. There was no significant difference between groups or over time for any blood variables.

Table 1.

Hemodynamic and patient variables from TIVA-anesthetized horses breathing 50% or maximal (> 90%) oxygen concentrations

Parameter	FiO2 (%)	−15 min	15 min	30 min	+10 min	Stand + 30 min
HR (beats/min)	50	40 ± 8	31 ± 7	35 ± 6	32 ± 3	42 ± 9
	> 90	35 ± 6	29 ± 6	33 ± 9	37 ± 6	41 ± 11
RR (breaths/min)	50	21 ± 11	11 ± 5	8 ± 2	14 ± 4	15 ± 2
	> 90	20 ± 14	11 ± 3	11 ± 3	18 ± 3	16 ± 7
TV (L/breath)	50	—	4.91 ± 1.37	4.90 ± 0.63	4.08 ± 1.62	—
	> 90	—	5.02 ± 1.09	5.21 ± 0.50	4.32 ± 0.97	—
EtCO2 (mmHg)	50	—	45 ± 6	46 ± 4	—	—
	> 90	—	43 ± 4	44 ± 5	—	—
SABP (mmHg)	50	—	124 ± 13	119 ± 11	133 ± 23	147 ± 23
	> 90	—	139 ± 20	129 ± 18	140 ± 12	138 ± 6
MABP (mmHg)	50	—	98 ± 9	94 ± 7	108 ± 12	114 ± 12
	> 90	—	107 ± 18	100 ± 15	112 ± 9	111 ± 11
DABP (mmHg)	50	—	82 ± 10	82 ± 5	92 ± 8	95 ± 8
	> 90	—	93 ± 16	86 ± 14	96 ± 9	92 ± 14
CO (L/min)	50	—	43.35 ± 14.01	59.13 ± 9.79	—	—
	> 90	—	47.66 ± 13.58	46.83 ± 14.14	—	—
CI (mL/kg BW/min)	50	—	83.5 ± 28.8	116.2 ± 20.6	—	—
	> 90	—	90.0 ± 25.0	89.9 ± 25.2	—	—

TIVA — total intravenous anesthesia. Data are presented as mean ± standard deviation. HR — heart rate; RR — respiratory rate; TV — tidal volume; EtCO₂ — end-tidal CO₂; SABP — systolic arterial blood pressure; MABP — mean arterial blood pressure; DABP — diastolic arterial blood pressure; CO — cardiac output; CI — cardiac index.

Table 2.

Blood variables from TIVA-anesthetized horses breathing 50% or maximal (> 90%) oxygen concentrations

Parameter	FiO2 (%)	−15 min	15 min	30 min	+10 min	Stand + 30 min
pHa	50	7.43 ± 0.03	7.35 ± 0.04	7.35 ± 0.03	7.43 ± 0.03	7.44 ± 0.02
	> 90	7.42 ± 0.04	7.36 ± 0.02	7.36 ± 0.02	7.46 ± 0.03	7.45 ± 0.02
PaO2 (mmHg)	50	101 ± 8	123 ± 52	117 ± 42	57 ± 8	99 ± 14
	> 90	95 ± 4	167 ± 51	164 ± 48	52 ± 4	91 ± 11
Arterial Hb (g/dL)	50	13.9 ± 2.4	11.0 ± 2.0	10.4 ± 1.3	11.4 ± 2.9	12.0 ± 1.7
	> 90	14.4 ± 1.8	11.8 ± 1.1	11.3 ± 1.4	12.0 ± 1.2	11.1 ± 2.3
PaCO2 (mmHg)	50	39 ± 3	49 ± 6	50 ± 4	42 ± 5	43 ± 4
	> 90	39 ± 3	47 ± 4	49 ± 4	38 ± 6	42 ± 5
PaHCO3 (mmol/L)	50	25.8 ± 0.9	25.4 ± 0.5	25.4 ± 0.3	27.0 ± 0.9	28.4 ± 1.0
	> 90	25.3 ± 2.6	25.1 ± 2.4	25.7 ± 1.7	26.9 ± 2.0	28.7 ± 2.0
SaO2 (%)	50	96.6 ± 0.5	96.1 ± 1.4	96.4 ± 1.1	91.0 ± 2.8	96.5 ± 1.2
	> 90	96.3 ± 0.2	97.2 ± 0.5	97.2 ± 0.5	89.8 ± 1.5	96.2 ± 0.8
pHv	50	7.40 ± 0.02	7.34 ± 0.03	7.33 ± 0.02	7.39 ± 0.02	7.40 ± 0.02
	> 90	7.40 ± 0.04	7.33 ± 0.02	7.34 ± 0.02	7.41 ± 0.02	7.41 ± 0.02
PvO2 (mmHg)	50	36.0 ± 7.7	33.0 ± 3.0	33.9 ± 2.8	30.3 ± 2.0	26.1 ± 2.8
	> 90	35.6 ± 4.7	33.3 ± 2.4	34.2 ± 3.5	28.8 ± 2.1	25.5 ± 4.1
VLactate (mmol/L)	50	1.0 ± 0.8	1.0 ± 0.9	1.2 ± 1.1	1.4 ± 1.2	1.9 ± 1.6
	> 90	0.6 ± 0.5	1.0 ± 1.4	0.7 ± 0.6	0.9 ± 0.7	1.2 ± 1.1

TIVA — total intravenous anesthesia. Data are presented as mean ± standard deviation. pHa — arterial pH; PaO₂ — arterial partial pressure of oxygen; Arterial Hb — arterial hemoglobin; PaCO₂ — arterial partial pressure of carbon dioxide; PaHCO₃ — arterial bicarbonate; SaO₂ — arterial oxygen saturation; pHv — venous pH; PvO₂ — venous partial pressure of oxygen; vLactate — venous lactate.

Table 3.

Oxygen variables from TIVA-anesthetized horses breathing 50% or maximal (> 90%) oxygen concentrations

Parameter	FiO2 (%)	−15 min	15 min	30 min	+10 min	Stand + 30 min
CaO2 (mL/L)	50	187 ± 33	147 ± 28	140 ± 18	144 ± 36	161 ± 24
	> 90	193 ± 25	160 ± 15	153 ± 19	150 ± 13	149 ± 30
CvO2 (mL/L)	50	136 ± 43	93 ± 16	92 ± 10	87 ± 6	78 ± 20
	> 90	136 ± 27	108 ± 25	102 ± 25	92 ± 19	80 ± 23
O2ER (%)	50	—	34.8 ± 7.9	33.9 ± 4.7	—	—
	> 90	—	32.8 ± 13.6	32.7 ± 8.4	—	—
VO2 (mL/kg/min)	50	—	4.0 ± 0.6	5.6 ± 1.9	—	—
	> 90	—	4.4 ± 1.7	4.3 ± 0.9	—	—
C(a-v)O2 (L/dL)	50	51 ± 21	54 ± 17	47 ± 10	57 ± 35	82 ± 22
	> 90	57 ± 18	52 ± 23	51 ± 10	57 ± 17	69 ± 24
PaO2-IO2 ratio	50	0.68 ± 0.06	0.34 ± 0.15a	0.32 ± 0.12a	0.38 ± 0.05a	0.66 ± 0.09
	> 90	0.64 ± 0.03	0.26 ± 0.08a	0.26 ± 0.08a	0.35 ± 0.03a	0.61 ± 0.07
P(A-a)O2 (mmHg)	50	1.5 ± 6.1	174.6 ± 51.4a,b	184.9 ± 43.8a,b	42.3 ± 12.5a	−0.4 ± 12.6
	> 90	7.4 ± 2.6	414.7 ± 61.3a,b	421.8 ± 70.0a,b	52.9 ± 10.7a	8.9 ± 6.6
DO2 (mL/kg/min)	50	—	11.8 ±1.3	16.3 ±4.1	—	—
	> 90	—	14.5 ±4.5	14.2 ±5.9	—	—
Qs/Qt (%)	50	12.8 ± 4.8	10.8 ± 2.8	11.0 ± 3.2	21.6 ± 7.8	6.7 ± 2.1
	> 90	12.3 ± 3.5	11.8 ± 6.3	10.4 ± 3.8	23.9 ± 6.6	8.0 ± 1.1
Vd/Vt (%)	50	—	8.6 ± 6.2	8.6 ± 6.9	—	—
	> 90	—	7.6 ± 4.5	10.0 ± 6.2	—	—

TIVA — total intravenous anesthesia. Data are presented as mean ± standard deviation. CaO₂ — arterial oxygen content; CvO₂ — venous oxygen content; O₂ER — oxygen extraction ratio; VO₂ — oxygen consumption; C(a-v)O₂ — oxygen content difference; PaO₂-IO₂ ratio — arterial oxygen tension to inspired oxygen tension ratio; P(A-a)O₂ — alveolar arterial oxygen tension difference; DO₂ — oxygen delivery; Qs/Qt — shunt fraction; Vd/Vt — alveolar dead space.

^aWithin a treatment, value is significantly different from the −15 minutes value for this variable.

^bWithin this time point, value is significantly different from the other treatment.

There was no difference in any calculated O₂ variable between groups or over time, with the exception of P(A-a)O₂. There was no difference in baseline P(A-a)O₂ between groups (1.5 ± 6.1 and 7.4 ± 2.6 mmHg for 50% and maximal O₂, respectively). In the 50% group, P(A-a)O₂ ranged from 174.6 ± 51.4 mmHg to 184.9 ± 43.8 mmHg during anesthesia, and was different from the maximal O₂ group (ranging from 414.7 ± 61.3 to 421.8 ± 70.0 mmHg). There was no difference in P(A-a)O₂ between groups 30 min after standing.

Two horses in the maximal O₂ group received a ketamine bolus of 400 or 200 mg. Two horses in the 50% O₂ group received a bolus of 200 mg.

The mean ± SD time to extubation was 23.8 ± 9.5 min for horses breathing 100% oxygen, and the mean ± SD was 22.3 ± 8.8 min for horses breathing 50%. The mean ± SD time to standing was 41.4 ± 10.8 min for horses breathing 100% oxygen, and the mean ± SD time to standing was 42.5 ± 10.8 min for horses breathing 50% oxygen. Horses in the 100% group stood on the first or second attempt. Five of the 6 horses in 50% group stood on the first or second attempt; 1 horse stood on the third attempt. There were no differences between groups in any recovery variables.

Discussion

An FiO₂ of 50% versus 95% did not reduce shunt fraction or improve CaO₂ or DO₂ during anesthesia or the subsequent recovery period. The lack of effect on shunt fraction supports the contention that compression rather than absorption atelectasis is responsible for V/Q mismatching in the anesthetized horse.

Nitrogen comprises approximately 80% of alveolar gas when ambient air is breathed. Nitrogen is relatively insoluble so its presence tends to maintain alveolar volume, an effect that has been termed alveolar splinting (19). Absorption atelectasis occurs when nitrogen is replaced with more soluble gases such as O₂. The more soluble gases are absorbed into blood, the splinting effect is reduced, and alveoli tend to collapse. The use of less than maximal FiO₂ would reduce the fraction of alveolar gas that is O₂, decrease this potential reduction in alveolar volume, and thus reduce shunt fraction by maintaining more functional alveoli. The results of this study support the hypothesis that atelectasis in the anesthetized horse is not the result of absorption atelectasis but is more likely due to compression of the lungs by overlying abdominal contents, as indicated by an increase in P(A-a)O₂ over time. Our results agree with a study finding no difference in shunt fraction with alteration in FiO₂in horses breathing helium mixtures (20). Our results also agree with previous studies by this laboratory and others using isoflurane anesthesia and mechanical ventilation (6,21). However, our results differ from studies finding increased Qs/Qt in horses breathing maximal FiO₂ versus air under intravenous anesthesia and increased Vd/Vt in horses under inhalant anesthesia breathing 35% versus maximal FiO₂ (12,17). Differences between our study and previous studies include an FiO₂ of 50% compared to 0.2 or 0.35, and the placement of horses in dorsal recumbency (5,8,22–24). As expected, P(A-a)O₂ increased in both groups after induction of anesthesia but there was no difference between groups. This study showed no benefit in time to extubation or standing, or in attempts to stand by altering FiO₂.

This study was limited by a small sample size, short anesthetic period, lack of investigator blinding, and the lack of a crossover design. A longer anesthetic duration could have demonstrated an effect of FiO₂ reduction on shunt fraction but low FiO₂ values are seen early during the course of most anesthetic episodes and another study from this laboratory did not show beneficial effects over 3 h of anesthesia. The investigators were not blinded to treatment group, and anesthetic depth may have been different between groups. Different anesthetic depths may have confounded respiratory variables. A crossover design could not be used because each horse was anesthetized on only once.

In conclusion, the reduction of FiO₂ from > 95% to 50% did not improve oxygenation or DO₂ in anesthetized horses in dorsal recumbency.