Introduction

Gastro-esophageal reflux (GER) is one of the major complications of general anesthesia and may result in pulmonary aspiration of gastric contents (GC), a potentially devastating complication. Indeed, one of the commonest causes of death related directly to anesthesia in humans is pulmonary aspiration of GC.¹ In dogs and cats, there has been no published report of pulmonary aspiration of GC at induction of anesthesia, but intraoperative GER has been considered to be one of the main causes of esophagitis, which may lead to stricture formation.^2–10

Traditional risk criteria for identifying human patients at increased risk for perioperative pulmonary aspiration include high gastric fluid volume and acidity (volume over 0.4 ml kg⁻¹ and pH less than 2.5). Fasting before general anesthesia is considered essential to patient safety in order to reduce the volume and acidity of stomach contents, thus decreasing the risk of regurgitation/aspiration. Strict preoperative fasting rules to ensure an empty stomach at induction has been a major concern for the anesthesiologists, which has evolved to become the accepted nil-by-mouth from midnight for those on a morning surgical list and an early light breakfast for those on an afternoon list.¹¹

However, over the past two decades, several authors have questioned the scientific basis of these rules.^12,13 It has been shown that the occurrence of GER during induction of anesthesia does not correlate with the GC volume.¹⁴ Numerous studies have demonstrated that fasting neither diminishes GC volume nor decreases gastric acidity, and the risk of pulmonary aspiration is not increased by the preoperative intake of clear liquids.^11,13,15,16 Moreover, a long starvation may lead to thirst, general discomfort, dehydration, and possible hypoglycemia.¹⁷

On reviewing the evidence base behind modern fasting guidelines in humans, it has been concluded that current guidelines and research state that a 2-h fast from fluids and a 6-h fast from solids is safe for healthy adults.¹¹ Indeed, based on the available data, various professional groups have produced several practice guidelines over the last 10 y recommending a more liberal approach to preoperative fasting guidelines in healthy patients undergoing elective procedures. The newer recommendations allow the consumption of clear liquids up to 2 h before elective surgery, a light breakfast 6 h before the procedure, and a heavier meal 8 h beforehand.^13,18–23

Moreover, recently, two large systematic reviews with meta-analysis have been published in the Cochrane Collaboration. In the first one, the authors concluded that there was no evidence to suggest that a shortened fluid fast results in an increased risk of aspiration, regurgitation, or related morbidity compared with the standard “nil by mouth from midnight” fasting policy. Permitting patients to drink water preoperatively resulted in significantly lower gastric volumes.²⁴ In the second review, concerning children who are considered to be at normal risk of aspiration/regurgitation during anesthesia, it was also concluded that there is no evidence that children who are denied oral fluids for more than 6 h preoperatively benefit in terms of intraoperative gastric volume and pH compared with children permitted unlimited fluids up to 2 h preoperatively. Children permitted fluids had a more comfortable preoperative experience in terms of thirst and hunger.²⁵

In veterinary medicine, in dogs in particular, there is considerable variation in the fasting guidelines.^26–30 However, there is evidence that increasing the duration of preoperative fasting is associated with an increased incidence of reflux in dogs.³¹ Moreover, recently, the effect of various types of food given at two different preoperative fasting times (3 or 10 h before anesthesia) on the GC volume and acidity has been reported in dogs.³² The authors showed that canned food at a half daily rate administered 3 h before anesthesia did not increase significantly the GC volume, while the GC acidity was reduced to probably clinically significant levels.

The aim of this study was to investigate the effect of two different preoperative fasting times on the incidence of GER during general anesthesia in dogs. Our hypothesis was that a 3-h fasting might reduce the incidence of GER.

Materials and Methods

One hundred and twenty dogs, status 1–2 on the American Society of Anesthesiologists scale, submitted to our clinic for elective surgery, were included in the study. This study was approved by the Institution Ethics Committee, and in all cases informed consent was obtained from the owners. Animals with symptoms, history, or any indication of gastrointestinal disease, or being treated with substances, which might affect gastrointestinal motility, were excluded. Emergency cases, as well as pregnant, obese, or elderly animals or animals scheduled for open cavity surgery (thoracotomy, celiotomy) or surgery requiring operating table tilt were also excluded.

The dogs were randomly assigned to one of the following two groups: administration of canned food 3 h before premedication (group C3, n = 60) and administration of canned food 10 h before premedication (group C10, n = 60). In group C3, the food was given at 06:00 h on the day of surgery, and in group C10, it was given at 23:00 h on the day before surgery. After having a normal meal, in all dogs of both groups a 12-h fasting period was applied before the administration of the test meal. The quantity of the latter was calculated as the half daily energy requirements, according to the formula E = 523 × BW^0.75 kJ (E = 125 × BW^0.75 kcal), in which BW is the body weight in kilograms.³³ The canned food given contained the following ingredients: water 80–83%, proteins 8%, fat 6–7.5%, carbohydrates 3.7–5.4%, inorganic salts 2–2.5%, fibers 0.3%, with 100 kcal/100 g. All dogs had free access to water for up to 3 h before premedication.

A few minutes before premedication, an indwelling intravenous catheter was introduced in the cephalic vein, and an infusion of Lactated Ringer's solution at 5 ml kg⁻¹ h⁻¹ began. Propionylpromazine^a at 0.2 mg kg⁻¹ was administered intravenously for premedication; 5 min later, thiopental sodium^b 2.5% solution at 5–10 mg kg⁻¹ was administered for induction of anesthesia, followed by tracheal intubation. In order to ensure a smooth transition to surgical plane anesthesia and avoid coughing or straining of the animal against the tube, from the time of intubation until anesthesia could be maintained with the inhalation anesthetic alone, there was usually a need for one or two further small doses of thiopental. These doses, added to the initial induction dose, constituted the total dose. Anesthesia was maintained with halothane^c in pure oxygen via a semi-closed circle (rebreathing) circuit, with the vaporizer out of circuit. Initially, the vaporizer was set at 3% until a surgical plane of anesthesia was achieved, and then the setting was reduced so as to maintain an end-tidal halothane concentration at 1.4 MAC. A standard monitoring protocol^d,e including electrocardiogram, non-invasive blood pressure, body temperature, capnography, spirometry, and anesthetic gas analysis was applied. Because no analgesic regiment was administered in premedication, in any elevation of the physiological variables more than 10%, which could not be reduced with an increase of the inspired halothane fraction (up to 1.6 MAC), rescue analgesia with fentanyl (1–3 μg kg⁻¹, intravenously) was given, and the animal was excluded from the study. Postoperatively and prior to anesthetic recovery, all the animals received analgesia with carprofen^f and an opioid (morphine^g 0.1–0.2 mg kg⁻¹ or pethidine^h 2–3 mg kg⁻¹, intramuscularly, depending on their condition and surgery performed).

Following intubation of the trachea, a pH electrode connected to a recording deviceⁱ was introduced into the esophagus, with its tip placed about 5 cm above the lower esophageal sphincter (LES). The position of the LES was estimated by measuring the length from lower jaw incisor tooth to the anterior border of the head of the tenth rib.³⁴ Esophageal pH was constantly monitored and recorded every 5 min. Esophageal pH of less than 4 or greater than 7.5 was taken as an indication of GER.³⁵ In case of GER, a second pH probe was introduced, with its tip at the upper esophageal sphincter (at the level of the larynx), in an attempt to detect the spread of the refluxate as an indicator of the potential for pulmonary aspiration. The pH recording was discontinued just prior to extubation. Throughout the whole procedure, all possible precautions were taken in order to prevent the increase of the intra-abdominal pressure from the manipulations of the animals during handling and surgical operation. No transportation of the animals to another operation room was required.

The Mann-Whitney test was used for continues variables and χ² test for categorical variables. Exact significance using the Monte Carlo simulation method was calculated. Probability (P) values of equal or less than .05 were considered to be statistically significant. A statistical analysis software application^j was used for the statistical calculations.

Results

A total of 149 dogs were recruited. Rescue analgesia was given to 29 animals, in which the pH recording was discontinued, and they were eliminated from the study, leaving a valid 120 animals. The sex distribution among the dogs was 69 males and 51 females. They were 3.7 (1.8) (mean [standard deviation]) y old (range 1–8 y) and weighed 21.9 (8.7) kg. The animals were homogenously distributed between the two groups regarding age (P = .235), weight (P = .112), breed (P = .095), and sex (P = 1.000). The operations performed were 42 orthopedic, 8 dental, and 70 soft tissue surgeries. There was no association between the type of operation and the allocation group (P = .109). Descriptive statistics of weight, age, and duration of anesthesia for the two groups are given in Table 1.

Table 1 Descriptive Statistics of Weight, Age, and Duration of Anesthesia

View Fullscreen

The time between induction of anesthesia and first pH recording had a median of 3 and a range of 2 to 5 min. In 108/120 (90%) cases, the first recording was obtained within 3 min after induction. The mean duration of pH recording was 74.8 (31.1) min. Fifty-five dogs were placed in left lateral, 40 in right lateral, 6 in sternal, and 11 in dorsal recumbency, whereas in 8 animals the placement was changed from left to right lateral (or vice versa). There was no association between the recumbency and the allocation group (P = .146).

In three of the 60 (5%) dogs of group C3, GER was recorded, whereas as many as 12/60 (20%) dogs of group C10 experienced GER. On all occasions, the refluxate was acidic (pH <4). In group C10, in three of the 12 cases, the refluxate reached the upper esophageal sphincter (as this was detected by the second pH electrode), and regurgitation and flow of GC from the mouth was observed in one of these dogs. Statistical analysis revealed that group C10 was significantly associated with increased incidence of GER (P = .025). Cross-tabulation of sex, type of operation, and recumbency by the fasting groups are shown in Table 2.

Table 2 Cross-Tabulation of Sex, Type of Operation, and Recumbency by the Fasting Groups

View Fullscreen

In one case of group C10, GER had already occurred when the electrode was placed in the esophagus (3 min after induction). In 12 of the 15 dogs of both groups that had a reflux episode, GER occurred within 30 min after induction of anesthesia. In the remaining 3 cases, GER occurred 47, 56, (group C3), and 99 (group C10) minutes after induction. The mean duration of GER (the time from the detection of the drop of pH below 4 to the end of anesthesia, during which time the pH remained below 4) was 37.15 (18.01) min.

Discussion

In the present study, 5% (3/60) of the dogs fasted for 3 h experienced a reflux episode during anesthesia, whereas in the group of the dogs fasted for 10 h the rate of reflux was 20% (12/60), the difference between the two groups being statistically significant. This is in agreement with previous findings in dogs according to which increasing the duration of preoperative fasting was associated with an increased incidence of reflux. In particular, none (0/30) of the dogs fasted for 2–4 h refluxed, whereas four (4/27, 14.8%) of the dogs fasted for 12 to 18 h and eight (8/26, 26.9%) of those fasted for at least 24 h had a reflux episode during anesthesia.³¹ The maximum risk estimation for reflux in the group of 2–4 h of fasting in the latter study was 9.5%, whereas the 95% confidence intervals of group C3 in the present study is virtually the same (0–10.5%). Therefore, in practice there is no difference in the results of the two studies. It is noteworthy that the two studies are similar with respect to the anesthetic protocol used, and the comparison of the results may be straightforward. The comparability of the design and the results of the two studies was one reason for omitting any analgesic drug in premedication.

Dogs fasted for 10 h have been reported to have greater GC acidity (lower pH) than those fasted for 3 h.³² Also, it has been found that a longer period of fasting was associated with increased gastric acidity in human beings as well as in dogs.^31,36 It has been shown that the administration of an alkali or an antacid can increase LES pressure.³⁷ This may be due to the neutralization of the acidic GC, followed by release of gastrin from the pylorus.³⁸ These authors have also studied the effect of liquid nutrients with varying pH on LES. Acidic nutrients produced more transient LES relaxations than nutrients with high pH. The authors believed that a mechanism in gastric mucosa may be activated in low pH values and cause transient LES relaxations and, hence, GER. These two studies may suggest that a mechanism of promoting (at low pH values) or preventing (at high pH values) GER may exist. Therefore, the increased incidence of GER in the dogs of group C10 may have been due to increased gastric acidity in this group. Furthermore, it may be argued that the administration of a “light meal” 3 h before anesthesia may reduce the incidence of GER by increasing GC pH.

The volume of GC is another factor that may affect the incidence of GER. It is generally believed that an increase in GC volume increases the risk of GER, and preoperative fasting is used to decrease that volume. However, the concept that prolonged fasting produces an “empty stomach” has been shown to be incorrect.^{1,12,22,39,40} The stomach can never be completely empty even after a midnight fast, because it continues to secret gastric juices.^1,41–43 In this context, it was recently shown that canned food at a half daily rate administered 3 h before anesthesia did not significantly increase the GC volume, and gastric residual volume was lower than the aforementioned volume to overcome the sphincter in humans as well as in anesthetized cats.³² Then, it could be argued that, within the range of the usual duration of preoperative fasting, including a 2 to 4 h fast, GC volume does not play an important role in affecting the occurrence of GER. In contrast, it seems that, in dogs, low GC pH is a more important risk criterion for identifying patients at increased risk for intraoperative GER.

The overall incidence of GER in this study was 12.5% (15/120), 5% in the dogs fasted for 3 h and 20% in those fasted for 10 h prior to the induction of anesthesia. Higher incidences of GER during anesthesia in dogs have been reported by other investigators.^44–46 In particular, in one study, an incidence of 27% (8/30) was reported, in contrast to an overall incidence of 13.3% (12/90) reported in another study in dogs similarly anesthetized except for propionyl-promazine and halothane, which were used instead of acepromazine and isoflurane, respectively.^44,47 These investigators speculated that the reasons for this difference might have been due to the differences in the anesthetic protocol. However, although there was evidence that the risk of GER may be higher when animals are anesthetized with isoflurane than with halothane, in another study, the same authors found that there was no difference in respect to the incidence of GER among the groups of dogs that were administered isoflurane, halothane, or sevoflurane.^45,48–50 Moreover, acepromazine and propionyl-promazine belong to the same group of drugs, that of phenothiazine derivatives. It follows that the cause of the difference in the incidence of GER between the two groups of study may be the different fasting protocols. In one study, the mean duration of the preoperative fast was 17.8 ± 4.1 h (range, 10-30 hours); in another study, none of the 30 dogs fasted for 2–4 hours refluxed, whereas 13.3% (4/30) of the dogs fasted for 12–18 h and 26.7% (8/30) of those fasted for at least 24 h had a reflux episode during anesthesia.^44,47

As mentioned above, increased gastric acidity decreases LES pressure, and prolongation of preoperative fasting increases gastric acidity and GER in dogs. In this respect, it is of interest to note that the preoperative administration of omeprazole, a proton pump inhibitor leading to inhibition of gastric acid production, resulted in a significant reduction of the incidence of GER in dogs.⁴⁶ It should be mentioned, however, that in another study within the range of times included, there was no relationship between duration of food withholding and the incidence of vomiting or GER.⁴⁴

Even higher incidences of GER (46.7% to 63.3%) have been reported.^44–46 Factors that may have contributed to this increased incidence of GER include prolonged preoperative fasting and, more importantly, the fact that an opioid was included in the premedication—morphine or methadone.^44–46 In the latter study, morphine was also administered, either alone or in combination with lidocaine, epidurally. Morphine decreases LES pressure and increases the probability of reflux in rhesus monkeys and human beings.⁵¹ The administration of morphine to healthy dogs prior to anesthesia was related to a significant increase in the incidence of GER during the subsequent anesthetic episode.⁴⁴ Interestingly, a recent systematic review suggested that intravenous crystalloids reduce the incidence of several, although not all, postoperative nausea and vomiting (PONV) outcomes and proposed that the effect of supplemental crystalloids on the incidence of PONV may be mediated by antidiuretic hormone (arginine vasopressin [AVP]).⁵² AVP is strongly associated with nausea and vomiting; intravenous AVP infusions induce nausea, retching, and vomiting in dogs and humans.⁵³ Plasma AVP levels are increased at the start of surgery and are significantly higher in patients who experience PONV than in those who do not.⁵⁴ Although the effect of AVP on intraoperative GER is not known, it is of interest to note that the administration of morphine, which is a strong predictor of PONV, raises plasma AVP levels in ferrets and humans.^55–57 The effect of opioids on the LES is the second reason why no analgesia was given during premedication in this study.

Another factor that may have contributed to the high incidence of GER reported in a prior report is that anesthesia in that study was induced with propofol, which is associated with a much higher incidence of GER than is the incidence of GER after induction with thiopental, probably due to the greater decrease of LES pressure induced by propofol than by thiopental in dogs.^46,58

In the present study, regurgitation (passive discharge of GC from the mouth or nose of a dog during anesthesia) was observed in only one case (1/120, 0.83%), although on another two occasions the refluxate reached the pharynx, as detected by the pH probe, with its tip at the upper esophageal sphincter (at the level of the larynx). This is in accordance with other studies, in which visible regurgitation occurred in 3 out of 510 (0.59%) dogs intraoperatively or in 41 out of 4271 (0.96%) perioperatively, namely either during anesthesia or before return to normal consciousness immediately after general anesthesia.^31,47 The latter percentage was reduced to 0.63% when the animals in which the pre-existing disease was considered a predisposing factor to regurgitation were excluded.⁵⁹

In the study reported here, the refluxate in all 15 cases with GER was acidic. In all reported studies alkaline reflux was also either not found or its incidence was rare.^{31,44–47,58}

In conclusion, under the above anesthetic protocol, feeding the dog 3 h before anesthesia at a half daily rate significantly reduces the incidence of GER during anesthesia, compared to the administration of the same type of food 10 h beforehand. If, however, GER eventually occurs, its consequences (esophagitis and/or esophageal stricture) may be mild, as long as GC pH is high. On that basis, the administration of a half daily dose of an ordinary canine diet may be addressed, and although the above findings have not been used in a sufficiently large number of clinical cases, it seems that the traditional “nil per os after midnight” or nil per os for 6–12 h prior to anesthesia in adult healthy dogs undergoing elective procedures should be reconsidered. Moreover, that current anesthetic practice of including premedication with an opioid drug and this fasting recommendation should be further evaluated under a more relevant anesthetic protocol.