Fasting Before Anesthesia: An Unsettled Dilemma

Sekar S. Bhavani, MD; Basem B. Abdelmalak, MD


Anesth Analg. 2017;125(2):369-371. 

The exact frequency of perioperative aspiration is not known.[1,2] This is largely because the signs and symptoms may be minimal and often nonspecific such as tachycardia, tachypnea, coughing, laryngospasm, or desaturation. In general, current aspiration risk estimates indicate that it is in the neighborhood of 1 in 2000 to 3000 anesthetics in adults[3–5] with a slightly higher risk in children (1 in 1200 to 1 in 2600).[3] Therefore, preoperative fasting for surgical patients became a standard practice in an attempt to minimize the risk of aspiration.

The anesthesiologists' concern over the aspiration risk is more so for colonoscopy procedures because the frequency of aspiration is higher than for other elective procedures, approximately 1 in 1000 (0.10%) when performed under sedation provided by the gastroenterologist. It is even higher in colonoscopy with anesthesia services (0.14%) as shown in a large Medicare population enrolled in the Surveillance, Epidemiology, and End Results [SEER] program.[6] Although the absolute risk is relatively small, the number of patients at risk can be magnified due to the large total number of patients undergoing this procedure with anesthesia services. The increasing number of colonoscopy procedures with anesthesia services is largely because of the latest recommendation of the American College of Gastroenterology, advising a colonoscopy every 10 years, beginning at age 50 years, as an important cancer prevention screening test, and the increasing utilization of deep sedation, often with propofol under the guidance of an anesthesiologist for routine colonoscopies. Nationwide, it has been estimated that almost one third of all colonoscopies are conducted under the guidance of an anesthesiologist versus a gastroenterologist providing procedural sedation.[7] The use of anesthesia services varies depending on the geographic region within the United States (53.4% in the Northeast as compared to 7.8% in the West in 2010)[7] and the severity of the underlying comorbidities.

Adequate bowel preparation plays a very important role in assuring the efficacy, ease of examination, accuracy, and quality of the colonoscopy procedure.[8] Nonetheless, establishing NPO (nil per os) guidelines for colonoscopy can be frustrating.

The American Society of Anesthesiologists (ASA) NPO guidelines[9] recommend a minimum of 2 and 6 hours of fasting following intake of clear liquids and a light meal, respectively, before anesthesia.[9] These recommendations apply to normal healthy patients scheduled for elective procedures. Emptying of the stomach following intake of liquids in a healthy individual is related to a pressure gradient between the stomach and the duodenum,[10] which is affected by the volume ingested.[10] Increased osmolality, acidity, and higher fat and protein content also retard the rate of emptying fluids from the stomach.[10]

This issue is further muddied by local policies, national guidelines, and the paucity of data regarding the safety and efficacy of different recommendations.[8] Moreover, there is still controversy regarding the efficacy of the different types of bowel preparation and the volume of the solute to be ingested.[11]

The use of a preparation solution that has a lot of particulate matter is more of a concern compared to a clear liquid preparation solution. The commonly used bowel preparation protocols include high-volume (4 L) or low-volume (2 L) polyethylene glycol (PEG) as a split (between the evening and the day of the procedure) or single dose. PEG has additives like bisacodyl or sodium ascorbate and ascorbic acid or the sports drink Gatorade (PepsiCo, Purchase, NY), sodium phosphate, magnesium citrate, combination of sodium, potassium, and magnesium sulfate.[12,13] Moreover, PEG solutions are associated with a higher frequency of nausea, vomiting, and bloating. In contrast, sodium phosphate bowel preparation solution has a lower volume; however, it is associated with a higher risk of dizziness and electrolyte abnormalities.[13]

The key factor that determines the quality of bowel preparation appears to be the timing of solution ingestion. There is a growing evidence that the use of a split preparation with a minimal time lapse between the conclusion of the preparation solution consumption and the colonoscopy examination (between 2 and 6 hours)[8,14–20] improves the efficacy of the procedure. Thus, as one might expect, there has been resistance (historically and anecdotally) from anesthesiologists for any shortening in the NPO duration before colonoscopy as recommended by the gastroenterologists' national organizations, considering that some bowel preparation solutions have particulate matter as well as the large volume (up to 4 L).

In this issue, Tandon et al[21] provided some more data to help with this challenge. They reported on the findings of their observational trial regarding the gastric residual volume (GRV) and its pH in 2 groups of patients who had their bowel preparation in the evening versus in the morning of the upper and lower gastrointestinal endoscopy. They reported that the GRV was not significantly different between the 2 groups. The same held true for gastric fluid acidity; however, the quality of the bowel preparation was better in the same-day group.

This study is helpful in terms of confirming the gastroenterologists' notion of the superiority of the quality of the bowel preparation/examination when the solution is administered on the same day. At the same time, they demonstrated that administration of the PEG solution resulted in very small GRV after as little as 3 hours from the end of consumption and not different from that resulting from its administration on the day before the procedure.

One strength of this study is that they used direct measurement of GRV by aspiration of the gastric contents through upper gastrointestinal endoscopy by the gastroenterologist compared to the other traditional methods such as ultrasound-guided or magnetic resonance image-guided estimation. Another strength is the direct measurement of the GRV pH, another important item when discussing aspiration risk.

While this is reassuring regarding the concern over the GRV and its pH after ingestion of a large volume of fluid close to the induction of deep sedation, we are still left with uncertainty regarding its impact on the aspiration risk.

The authors understandably did not use aspiration as the study outcome because of its very low frequency (that would require a formidable large number of patients to study) and instead used surrogate outcomes such as GRV and pH. While many believe that there is a relationship between GRV and its pH and aspiration risk per se, there are other important factors in determining the outcome, should aspiration occur, such as the presence of particulate matter and the infective load.[3,4] The GRV and pH at which this risk becomes significant has not been clearly defined.[2,22] The literature seems to suggest that a pH of <2.5 and a volume of >0.3 mL/kg are required to produce aspiration pneumonitis.[5] Raidoo et al[23] using a primate model showed that as the gastric volume increased from 0.4 to 0.6 mL/kg to 0.8 to 1.0 mL/kg at pH of 1.0, the risk of developing a significant respiratory compromise and death increased by almost 50%. If we extrapolate these data to adults, it would mean that the critical volume in adult humans would be approximately 50 mL[24] or conservatively estimated to be >25 mL as per the ASA NPO guidelines.[9]

Aspiration risk assessment is a complex matter. In addition to inadequate fasting, aspiration has also been associated with patient-, surgery-, and patient positioning-related factors.[25–27] Patient factors include advanced age, male sex, emergency surgery, gastroparesis, prior upper gastrointestinal surgery, diabetes, chronic opioid use, recent trauma, pain, pregnancy, morbid obesity, decreased laryngeal reflexes, increased intracranial pressure, reflux disease, and gastrointestinal obstruction. Surgery-related factors[25,26] include upper abdominal surgery or cholecystectomy, increased manipulation of the gut, and increased intra-abdominal pressure. Patient positioning factors include steep head-down position and lithotomy position.

Furthermore, both very deep and very light planes of anesthesia[25,26] may be associated with a risk of aspiration. Lighter planes of anesthesia may be associated with coughing, bucking, gagging, or recurrent swallowing that may result in distension of the stomach and regurgitation and vomiting. A deeper plane may be associated with loss of airway protective reflexes and an increased risk of aspiration. Propofol, which is commonly used for moderate, deep sedation and/or general anesthesia in colonoscopy, has been shown to be associated with an increased risk of aspiration due to a failure of the normal protective reflexes.[28] During anesthesia, the barrier to reflux is affected not only by a decrease in the lower esophageal sphincter relaxation, but also by a decrease in the upper esophageal sphincter occlusion pressure.[26] This risk is increased by the administration of benzodiazepines, opioids, and anticholinergics as part of preanesthetic medications.[26,29] While this study addressed some of these factors, it did not address them all.

Moreover, regurgitation and aspiration can occur perioperatively in the presence of an empty stomach as a result of retrograde transmission from the small intestines.[25] In colonoscopy, this is more likely to occur when a large amount of air is insufflated into the bowel and excessive external manipulation is required to complete a colonoscopy through navigating the splenic flexure. Another difference between gastrointestinal endoscopy procedures and other elective surgeries that impact the generalizability of the findings of this study to other elective procedures is that the prone, semiprone, or varying degrees of lateral positions are utilized frequently, which help to direct any regurgitated gastric contents out of the mouth (gravity driven) and prevents pooling in the mouth or aspiration into the trachea.

In conclusion, this observational trial by Tandon et al[21] provides some new insights regarding GRV and pH following administration of a large amount of bowel preparation solution. Such data, together with the emerging new evidence from other investigators, can be very helpful in understanding this complex issue. For example, a recent study indicated that adding milk to coffee does not impair gastric emptying as previously thought, evidenced by similar GRV measured by magnetic resonance imaging in healthy volunteers who ingested coffee with and without added milk in a randomized crossover design.[30] As we learn more, we might end up seeing some new directions when the ASA NPO guidelines are revisited. However, because aspiration pneumonitis is such a grave complication, we are still in need of more definitive data from a larger number of subjects done in a randomized fashion to reliably modify our practice and/or the current ASA NPO guidelines. Whether guidelines are changed or not, considering the complex nature and all the risk factors for aspiration and its outcome as discussed previously, 1 size will probably not fit all. Thus, the individualized care and medical judgment on a case-by-case basis by anesthesiologists cannot be overemphasized.