Impact of preoperative fasting times on blood glucose concentration, ketone bodies and acid–base balance in children younger than 36 months

A prospective observational study

Abstract

BACKGROUND 

In contrast to preoperative fasting guidelines in paediatric anaesthesia, actual fasting times are often too long.

OBJECTIVE 

The objective of this study was to evaluate the effect of preoperative fasting on glucose concentration, ketone bodies and acid–base balance in children.

DESIGN 

A prospective, noninterventional, clinical observational study.

SETTING 

A single-centre trial, study period from June 2014 to November 2014.

PATIENTS 

One hundred children aged 0 to 36 months scheduled for elective paediatric surgery.

MAIN OUTCOME MEASURES 

Patient demographics, fasting times, haemodynamic data, glucose and ketone body concentrations, and acid–base parameters after induction of anaesthesia were documented using a standardised case report form.

RESULTS 

Mean fasting period was 7.8 ± 4.5 (3.5 to 20) h, and deviation from guideline (ΔGL) was 3.3 ± 3.2 (-2 to 14) h. Linear regression showed a significant correlation between fasting times and ketone bodies, anion gap, base excess, osmolality as well as bicarbonate (for each, P < 0.05), but not glucose or lactate. In children with ΔGL more than 2 h (54%), ketone bodies, osmolality and anion gap were significantly higher and base excess significantly lower than children with ΔGL less than 2 h (for each, P < 0.05).

CONCLUSION 

After prolonged preoperative fasting, children younger than 36 months can present with ketoacidosis and (low) normal blood glucose concentrations. Actual fasting times should be optimised according to existing guidelines. In small infants, deviations from fasting guidelines should be as short as possible and not longer than 2 h.

Introduction

Preoperative fasting is important to reduce the risk of regurgitation and pulmonary aspiration of gastric contents in children undergoing anaesthesia. For children, the preoperative fasting guidelines published by the American Society of Anesthesiologists (ASA) and European Society of Anaesthesiology (ESA) allow clear fluids up to 2 h, breast milk up to 4 h and other milk and solids up to 6 h before induction of deep sedation or general anaesthesia.^1,21,2 However, actual fasting times in children are often longer than recommended and many children suffer from a considerable amount of preoperative discomfort because of excessive fasting times.^3,43,4 Whereas hunger and thirst are mostly obvious, and blood glucose concentrations are often measured during anaesthesia, the concentration of ketone bodies is normally not monitored in routine cases. Therefore, this study investigated the relationship between the duration of preoperative fasting and blood glucose concentrations, ketone bodies and acid–base balance in children younger than 36 months. In addition, we hypothesised that children younger than 1 year or those with prolonged fasting times (≥2 h more than the guideline recommendations) would display significantly different concentrations of ketone bodies and acid–base values from other children.

Materials and methods

Following the local ethics committee's approval (Ethical Committee Hanover Medical School, Germany, Chairperson Prof. Dr H. D. Troeger, No. 2310-2014 on 26 June 2014), 100 children aged 0–36 months scheduled for elective paediatric surgery were included in this prospective, noninterventional, observational study, all of them in a stable haemodynamic condition. Children who received intravenous fluids during the fasting period, were admitted to critical care, experienced nausea and vomiting, or had a metabolic disorder were excluded. The study was conducted from June 2014 to November 2014 in the Clinic for Anaesthesiology and Intensive Care Medicine, Hanover Medical School, Germany.

As a routine, the ward staff were given written individual instructions for fasting times before the estimated time of induction of anaesthesia according to the institutional guideline: 6 h for solid foods, 4 h for breast milk or infant formula, and 2 h for clear fluids (apple juice or sugared tea). On arrival in the operating theatre, the accompanying nurse and parents were asked when the child last ate or drank and the reported time was documented.

After induction of anaesthesia, a venous blood sample was collected routinely from all children. In each sample, pH, pO₂, pCO₂, base excess, bicarbonate, sodium, potassium, chloride, lactate, haemoglobin, haematocrit, oxygen saturation (Rapidlab 860; Bayer Diagnostics, Fernwald, Germany) and ketone bodies (FreeStyle Precision Neo H, Abbott, UK; Fig. 1) were measured. The anion gap was calculated as the sum of sodium and potassium minus the sum of bicarbonate and chloride concentrations. Ketone bodies less than 0.6 mmol l⁻¹ were accepted as normal and hypoglycaemia was defined as glucose concentration less than 2.8 mmol l⁻¹.

Fig. 1: Bedside ketone meter (FreeStyle Precision Neo H, Abbott, UK).

Monitoring for all children included electrocardiogram, noninvasive arterial blood pressure, pulse oximetry, oesophageal temperature and end-tidal carbon dioxide. A warm-air blanket was used to maintain normothermia during the procedure. Arterial blood pressure was documented before and after induction of anaesthesia. All recorded data were analysed using MS Excel (Excel 2010; Microsoft, Seattle, Washington, USA) and Medcalc Statistical Software (Medcalc 2014; Medcalc Software bvba, Ostend, Belgium) and presented as mean values ± standard deviation (SD) (range) or frequency. Wilcoxon test, Mann–Whitney test and linear regression analysis were performed for metric data with a predefined significance level of α equal to 0.05. Sample size calculation (Medcalc 2014; Medcalc Software bvbv, Ostend, Belgium) indicated that a sample size of 100 children would allow the detection of a 20% difference in ketone body concentration between children with prolonged [difference from guideline (ΔGL) >2 h] and normal (ΔGL <2 h) fasting times.

Results

A total of 100 children were included. Patient characteristics are summarised in Table 1. Mean fasting time for milk or solids was 7.8 ± 4.5 (3.5 to 20) h and mean deviation from guideline (ΔGL) was 3.3 ± 3.2 (-2 to 14) h. In 54% of the children, ΔGL was more than 2 h, 84% of the children received milk or solids after midnight and 20% clear fluids up to 2 h before anaesthesia. No incidents of regurgitation or pulmonary aspiration were reported.

Table 1: Patient characteristics (n = 100)

Linear regression analysis showed significant correlation between fasting time and ketone bodies, anion gap, base excess, osmolality as well as bicarbonate (for each, P < 0.05), but not glucose or lactate (Table 2). Systolic arterial blood pressure (SAP) and mean arterial blood pressure (MAP) decreased significantly after induction of anaesthesia. The fasting times did not correlate with the percentage change in SAP (ΔSAP%) or MAP (ΔMAP%). Three children presented with hypoglycaemia (glucose <2.8 mmol l⁻¹), two of them with ΔGL more than 2 h. In 23 children, ketone body concentration was 0.6 to 1.5 mmol l⁻¹, 20 of them with ΔGL more than 2 h. Seven children had ketone body concentrations more than 1.5 mmol l⁻¹, all of them with ΔGL more than 2 h.

Table 2: Correlation between fasting times and percentage changes of systolic (ΔSAP%) and mean (ΔMAP%) arterial blood pressure, and the absolute changes in blood glucose, lactate, acid–base-balance, osmolality and ketone bodies (n = 100)

Infants younger than 12 months had shorter fasting times [mean 6.8 ± 2.6 (3.5 to 15.3) h] and displayed higher bicarbonate and base excess values as well as lower anion gaps and ketone bodies. The ΔMAP% during induction of anaesthesia was significantly greater than in the older children (for each, P < 0.05).

Children with ΔGL more than 2 h (54%) displayed significantly higher levels of ketone bodies, osmolality and anion gap as well as lower levels of base excess and bicarbonate than children with ΔGL less than 2 h (for each, P < 0.05). Fifty percent of the children with ΔGL more than 2 h had increased ketone bodies and 90% of the children with increased ketone bodies had ΔGL more than 2 h (Fig. 2). Furthermore, ΔSAP%, ΔMAP%, pH, glucose and lactate were comparable in both groups (Table 3, Fig. 3).

Fig. 2: Association between the number of hours deviation from fasting guideline (ΔGL) (x-axis) and ketone bodies (y-axis) (n = 100). The grey shaded area includes cases with increased ketone bodies and ΔGL >2 h. The number in the corner of each quadrant is the percentage of the total number of children in that quadrant.

Table 3: Comparison of the percentage changes in systolic (ΔSAP%) and mean (ΔMAP%) arterial blood pressure, and the absolute changes in blood glucose, lactate, acid–base balance, osmolality and ketone bodies in fasting children with deviation from guideline (ΔGL) <2 h (n = 46) vs. ΔGL >2 h (n = 54)

Fig. 3: Glucose, ketone bodies and negative base excess in fasting children after deviation from guideline <2 h (n = 46) and >2 h (n = 54). ^* P < 0.05.

Discussion

In line with our hypothesis, this study demonstrates that prolonged fasting times can result in ketoacidosis with (low) normal glucose concentrations in children younger than 36 months; this is more pronounced if the fasting time exceeds the guideline by more than 2 h.

Preoperative fasting in paediatric anaesthesia is a routine strategy to reduce the risk of regurgitation and pulmonary aspiration. Research findings document that pulmonary aspiration is a rare complication in children (incidence 4 to 10/10 000), and the reported mortality is extremely low.^{5–85–85–85–8} Most institutions use a fasting regimen according to the ASA or ESA guidelines (6 h for solids or infant formula, 4 h for breast milk and 2 h for clear fluids),^{1–31–31–3} but some authors use shorter fasting times (e.g. 4 h for a light meal or infant formula, 3 h for breast milk and 2 h for clear fluids) without reporting a higher incidence of pulmonary aspiration.^9,109,10 A recent study showed that a 1-h clear fluid fast did not alter gastric pH or residual volume significantly as compared with 2-h fasting.¹¹ MRI did not reveal a difference between residual gastric content volumes at 4 and 6 h after a light meal.^12,1312,13

The prolonged fasting times in the current study were possibly caused by the following reasons: First, children may have been fed earlier in the evening and have then slept through the whole night before surgery; second, unscheduled patients requiring emergency surgical procedures took priority in the operating schedule; third, unexpected cancellations leading to changes in case order; a child's late arrival on the ward or delays on the ward could result in patients missing their scheduled operation time; fourth, the operation schedule may change for other reasons after the fasting orders had been written. In infants and children, extended fasting can be associated with significant discomfort and distress that reduce the quality of their inpatient stays and contribute to parent and patient dissatisfaction with their hospital experience.^3,43,4 Clear liquids appear to entail no additional risk of regurgitation or pulmonary aspiration of gastric content in normal healthy children and may provide some psychological benefit, as demonstrated by a decrease in irritability before induction of anaesthesia.¹⁴ Unfortunately, neonates and young infants, mostly used to milk, will often not drink clear fluids. We found shorter real fasting times in infants younger than 1 year, more in line with the guidelines, resulting in a more stable metabolic state and acid–base balance. Nevertheless, this age group had a higher percentage fall in MAP during induction of anaesthesia, possibly emphasising the importance of keeping real fasting times as short as possible in this age group to avoid further dehydration.¹⁵ It would seem that improvements in organisation and communication are urgently needed in order to minimise actual fasting times so that they come closer to guideline-recommended times. Prescription of milk and clear fluids at a specific time and the implementation of maximum fasting periods may improve the control of preoperative fasting times in children.

In agreement with other authors, this study has shown that blood glucose concentrations remain within a (low) normal range in most children even after long fasting periods,^{16–1816–1816–18} whereas the concentration of ketone bodies increases with the duration of fasting. During the early stages of fasting, hepatic glycogenolysis is the primary glucose source. As fasting continues and glycogen stores are depleted, hepatic gluconeogenesis, lipolysis and subsequent fatty acid beta-oxidation and ketogenesis become the major energy sources.¹⁹ In children with high concentrations of ketone bodies, uptake and oxidation of ketone bodies in the central nervous system are possible. This physiological mechanism protects the central nervous system from energy deficiency in the presence of hypoglycaemia. The accumulation of negatively charged ketone bodies leads to a decrease in bicarbonate and base excess and to an increase in the anion gap. The algorithm used by the blood gas analyser did not include ketone bodies for the calculation of osmolality; therefore, the increase in osmolality was more likely related to dehydration caused by prolonged fasting.

Optimal fasting times are an integral part of modern fast track concepts to enhance postoperative recovery.²⁰ Attention to detail is important in paediatric anaesthesia and keeping the fasting time as short as possible is an important detail in improving perioperative homoeostasis in infants and children. Performing the surgical procedure on all patients no later than 2 h after the recommended fasting period would have avoided or mitigated 90% of the observed cases with ketoacidosis (Fig. 2). If a child presents with metabolic acidosis after prolonged fasting, the anaesthetist should consider ketoacidosis as a possible cause. Ketone bodies can be easily measured by a small bedside ketone meter (Fig. 1).²¹ In children with high ketone body concentrations, the blood glucose levels should be increased by infusion of a glucose-containing isotonic balanced electrolyte solution.^22,2322,23

In conclusion, after prolonged preoperative fasting, children younger than 36 months can often present with ketoacidosis and a (low) normal blood glucose concentration. Actual fasting times should be optimised according to existing guidelines. In small infants, deviations from the fasting guidelines should be as short as possible and not longer than 2 h.

Acknowledgements relating to this article

Assistance with the study: none.

Financial support and sponsorship: none.

Conflicts of interest: none.

Presentation: none.