We examined the impact of drinking water on gastric volume using a randomized single blinded crossover design. This is the first study to quantify gastric volume using ultrasound in healthy volunteers who were asked to follow the ESA perioperative fasting guidelines (nil-by-mouth two hours before induction of anaesthesia) and the more liberal Scandinavian guidelines (150 ml water with medication up to 1 h before induction). We observed that gastric volumes increased following ingestion of water compared to the same individuals in a fasted state. Water left the stomach rapidly following ingestion in most individuals, and typically within 30 minutes. A reduction in thirst was observed in the group allowed to drink water, whereas hunger and anxiety scores were unaffected.
Our findings are consistent with results from other studies which demonstrate complete gastric emptying of a drink of water within 20–40 minutes in most people.17,18 Our observation that two individuals had a gastric volume greater than 1.5 ml kg-1 (participants 11 & 12) despite undergoing a period of fasting is consistent with previous studies. These studies in large groups of patients reported 5–6.2% of fasted surgical adult patients present with solid gastric contents, residual gastric volume > 1.5 ml kg-1 or fluid/solid contents in the antrum in both the supine and lateral decubitus positions.19,20
Using standard bedside ultrasound imaging, we were able to easily and rapidly (scanning time < 5 minutes) distinguish those who had followed the ESA guidelines from those who had followed the alternative liberal fluid regime. While the majority of participants in our study showed rapid gastric emptying, one participant (Fig. 3: participant 6) had a residual gastric volume < 1.5 ml kg-1 after following ESA guidelines but > 1.5 ml kg-1 after a more liberal fluid regimen. This participant described eating a substantial rice, egg and spinach salad 4.5 hours prior to the start of scanning for their H2O protocol. This was compliant with the ESA guidelines, but it was observed that not only was the participant’s baseline gastric volume increased compared to their control protocol, it was also possible to identify rice sonographically in the stomach, appearing to ‘swell’ further after each drink of water. This observation suggests a potential limitation of the ESA guidelines, as what is eaten may be as important as when it is eaten.
Our study had several strengths. First, participants were randomized to the order in which they underwent the control or intervention protocol, to minimise influence on their oral intake prior to attending the study. Second, the measurement of gastric volume was performed by trained sonographers who were blinded to treatment protocol. Previous literature has suggested that 33 supervised scans are required for an anaesthetist to achieve competence in accurate performance of gastric ultrasound.21 Using qualified sonographers mitigates the learning curve required and was of benefit to this study but is less applicable to point of care ultrasound in clinical practice. Third, a validated method was used for the measurement of gastric volume. The ultrasound assessment of gastric volume as described by Van de Putte was successfully reproduced by qualified sonographers in this study. Agreement between gastric volumes measured 5 minutes apart by each sonographer in the fasted state was good to excellent, indicating that measurement of gastric volume via this method was reliable and repeatable, in line with previous studies.10
Our study has several limitations. First, we studied a small sample of healthy volunteers who were not due to undergo surgery. Our aim was to recruit individuals who were likely to have normal gastric physiology and thus all participants had a normal BMI, were without significant comorbidity or taking medication known to alter gastric emptying. Furthermore, participants were not exposed to the stress and anxiety of waiting for an operation. The elective surgical population commonly have comorbid illness, such as diabetes or obesity, which may directly or indirectly (i.e. through medications) reduce the rate of gastric emptying, thus these results cannot be simply extrapolated to the typical surgical population. Second, these findings may not apply to other clear fluids owing to the fact that properties such as energy content have been shown to alter rate of gastric emptying.22 Third, it is possible that the volume of drink consumed may also alter the rate of emptying. In this study we administered 150 ml as it is the volume of water allowed with medication up to 1 hour before general anaesthesia in the Scandinavian guidelines, with no apparent increased incidence of adverse events.23 Fourth, although agreement of gastric volumes between the baseline scans taken in the fasted state in both protocols were good, there was some variability. This most likely represents normal variability in gastric physiology. However, it could be due to measurement error or differences in scanning techniques between sonographers including probe pressure, the part of the gastric antrum assessed and timing of the scan in the respiratory cycle. The sonographers in this study were experienced at performing abdominal ultrasound, however, had not previously been required to assess gastric contents or volume. Providing detailed guidance on all aspects of the scanning technique may help reduce inter-observer variability. Finally, we observed that participant 12 in the control group had a gastric volume at the end of the scanning period that was greater than their baseline gastric volume. This participant had a single additional scan that was unchanged from the previous 90 minutes (1.9 ml kg-1) and should therefore have stayed for additional scans until the gastric volume returned to baseline. However, we were unable to complete these scans as the participant declined further scans.
The rapidity of water leaving the stomach as demonstrated in this study adds support to guidelines and studies that suggest reducing fasting times for clear fluids prior to induction of general anaesthesia is safe. Therefore, current guidance on pre-operative fluid restriction should be reviewed. The value of routine pre-operative gastric ultrasound is less clear. If gastric ultrasound was used to assess gastric volume in all patients prior to anaesthesia then our results suggest that some operations would be delayed or anaesthetic techniques altered. Fasted patients with high residual gastric volumes currently go undetected and there is no evidence that they suffer any adverse consequences.23
Further study is required, using this established ultrasound protocol, in patients presenting for surgery (with associated increased anxiety) and should include individuals with comorbidities. It would be beneficial to determine whether administering 150 ml water 1 hour before induction of anaesthesia leads to an increased incidence of high-risk ultrasound features. These typically occur in 5–6.2% of patients undergoing elective operations with conventional fasting regimes and features include: presence of solid gastric contents, residual gastric volume > 1.5 ml kg-1 or fluid/solid contents in the antrum (in both the supine and lateral decubitus positions).19,20 Future work should take into account the many factors that are known to delay gastric emptying and aim to stratify the risk of adopting a more liberal fasting policy in the presence of such risk factors.
Current guidance on pre-operative fasting to reduce aspiration risk may result in unnecessary dehydration. We conducted a trial in healthy volunteers of fluid restriction (ESA guideline)1 compared to more liberal fluid intake (Scandinavian guideline)23 to explore the impact on gastric volume. We found that after 30 minutes following water consumption, most participants had a gastric volume that corresponded with a low risk of aspiration. Future studies should explore the impact of more liberal fluid regimes in surgical patients.