Traditional fasting guidelines before elective surgery have required the patient to fast starting at midnight before their procedure. The use of general anesthesia can alter gag, cough, and swallow reflexes that protect the lungs from stomach contents. Fasting is intended to reduce stomach content acidity as well as volume in an attempt to decrease morbidity and mortality rates from anesthesia-related aspiration. However, clinical studies and practice recommendations do not always support this tradition and encourage the use of more liberal fasting guidelines. These fasting guidelines generally allow patients to consume clear liquids (e.g., water, apple juice, black tea, black coffee) up to 2-3 hours before their scheduled surgery or procedure requiring anesthesia. Although being "nil per os" (NPO [nothing by mouth]) at midnight was often considered a safe practice with minimal patient impact, new research indicates that preoperative fasting may increase the risk of complications. In addition, the use of a carbohydrate-rich clear liquid beverage prior to surgery may actually improve patient outcomes. This article reviews the science behind fasting, current guidelines and practices, and studies evaluating the use of preoperative carbohydrate-rich clear liquid beverages.

Literature Review

It is estimated that 1 of every 2000-3000 elective surgeries results in aspiration (Janda, Scheeren, & Noldge-Schomburg, 2006). However, a review of 83,844 surgical cases receiving anesthesia reported only five episodes of aspiration (Fasting & Gisvold, 2002). Complications of aspiration include acid-associated aspiration pneumonitis, bacterial infection, and particle-associated aspiration. Aspiration pneumonia may result in increasing length of stay up to 15 days and additional costs of $22,000 (Kozlow, Berenholtz, Garrett, Dorman, & Pronovost, 2003).

It is difficult to determine at what point NPO at midnight before elective surgery became routine. Maltby (2006) provides a historical perspective on preoperative fasting guidelines and suggests that in the 1960s, NPO at midnight became standard practice. These initial recommendations to fast before surgery were made to decrease the discomfort of nausea, not to protect the patient from an adverse medical complication. An unpublished study of Rhesus monkeys that were injected with an acidic fluid into their lungs prompted a recommendation in 1974 by Roberts and Shirley (1974) to limit gastric contents to 25 ml to reduce aspiration risk. An additional evaluation of acid fluid administered into the trachea of monkeys suggested that gastric contents should be limited to 0.8 ml/kg (Raidoo, Rocke, Brock-Utne, Marszalek, & Engelbrecht, 1990).

Alternately, multiple studies in the 1980s reported that survival rates in rats after hemorrhagic shock were lower in fasting rats than in fed rats (Alibegovic & Ljungqvist, 1993; Ljungqvist, Jansson, & Ware, 1987; Nettelbladt, Alibegovic, & Ljungqvist, 1996). Studies investigating measured gastric volume in humans after various fasting periods were not undertaken until the late 1970s (Ong, Palahniuk, & Cumming, 1978). In 1986, Maltby, Sutherland, Sale, and Shaffer (1986) published one of the first studies investigating gastric residual volumes after fasting or consuming water. The results showed that residual gastric volumes (RGVs) were statistically lower in the group that drank water than in those that had been fasting (17.6 +/- 14.5 ml vs. 26.7 +/- 18.9 ml; p < .02) (Maltby et al., 1986). However, in a follow-up study, there was no difference in RGVs in those who consumed coffee/tea or orange juice or completed an overnight fast (Hutchinson, Maltby, & Reed, 1988).

Consumption of Clear Liquids Prior to Anesthesia

Although the theory of fasting before surgery to prevent or reduce the risk of aspiration seems intuitive, the physiology of gastric emptying does not support the need for fasting 12 hours prior to anesthesia. Gastric emptying is dependent on many factors including hormonal stimulation, calorie and macronutrient content, volume consumed, size of food particles, and osmolality (Calbet & MacLean, 1997; Hellstrom, Gryback, & Jacobsson, 2006). Osmolality of liquids is determined by the concentration of the particles in the solution and higher osmolality liquids may decrease gastric emptying (Vist & Maughan, 1995). Typically 90%-95% of iso-osmolar (300 mOsm/kg) fluids pass through the pylorus within 1 hour after consumption (Hunt, 1956; Jolliffe, 2009). Because many factors are involved in gastric emptying, the patient should be instructed on the difference between clear liquids and liquids containing additional components such as fat or fiber.

Multiple studies have evaluated if allowing clear liquids up to 2 hours before surgery increases the risk of increased RGVs. McGrady and MacDonald (1988) provided 100-ml water before surgery or required the patient to be NPO at midnight. Median RGV was lower for the water group (16.5 ml) than for the NPO group (25 ml), but this was not statistically different. Maltby et al. studied 199 elective surgery patients and provided unrestricted clear liquids until 3 hours before surgery and compared them with patients requested to be NPO at midnight (Maltby, Lewis, Martin, & Sutherland, 1991). Again, there was no statistical difference in RGVs between the groups.

Shevde and Trivedi (1991) measured RGV every 30 minutes after 240 ml of water, coffee, or pulp-free orange juice was consumed. All of the healthy volunteers had RGV of less than 25 ml within 2 hours of consumption and there was no statistically significant difference between the groups. Several additional studies in Figure 1 highlight the lack of consistency in RGVs after fasting or consuming liquids prior to surgery.

FIGURE 1. Residual gastric volumes after fasting at midnight, drinking water or a placebo 2-3 hours before surgery, or consuming a carbohydrate-containing beverage 2-3 hours before surgery.

The majority of studies compared RGVs between experimental and control groups without determining the associated risk for aspiration. The North American Summit on Aspiration in Critically Ill Patients: Consensus Statement reported that RGV correlates poorly with gastric emptying and is not a reliable indicator of aspiration risk (McClave et al., 2002). Research in the critical care setting provides further support that RGV is not a valid marker for risk of aspiration (McClave et al., 2005). The mean RGV for all aspiration events in a study of 1,118 gastric volume measurements was 30.6 ml (range, 0-700 ml) and the frequency of aspiration did not correlate with an increase in RGV.

There are additional effects of NPO status beyond aspiration risk. A pediatric study evaluated irritability and dehydration rates in patients undergoing orthopedic surgeries that were either made NPO at midnight or provided with 250-ml apple juice about 2.5 hours before surgery (Castillo-Zamora, Castillo-Peralta, & Nava-Ocampo, 2005). Patients in the NPO group fasted for an average of 13.2 +/- 3.3 hours and had increased complaints of irritability (odds ratio [OR] = 4.5; 95% confidence interval [CI] [1.9, 10.3]) and dehydration (OR 21.6; 95% CI [5.9, 79.0]). Another study evaluated hunger (p < .05) and thirst (p < .05) and found decreased rates in those who received water prior to surgery compared who those who were NPO (Agarwal, Chari, & Singh, 1989).

Avoiding prolonged preoperative fasting can also decrease risk of hospital-acquired malnutrition. A survey of practices at the University of Louisville Hospital, a large U.S. tertiary and academic care center, reported that 22% of patients admitted to the hospital were NPO or received a clear liquid diet for 3 or more days (Franklin et al., 2011). Prolonged postoperative avoidance of oral nutrition would clearly be complicated by an extended preoperative fasting and may contribute to overall malnutrition.

Malnutrition in the hospital setting has been estimated at 35% with an additional 30%-35% of patients who are at risk for developing malnutrition. Extending preoperative fasting times beyond those absolutely essential creates an unnecessary delay in providing nutrition to the patient. Malnutrition has been associated with increased infection rates, impaired wound healing, and increased hospital length of stay (Barker, Gout, & Crowe, 2011). A prospective observational study in a surgical intensive care unit determined that negative energy balance from inadequate nutrition correlated with statistically significant increased length of stay (p < .001), infectious complications (p < .0042), and days on mechanical ventilation (p < .0002) (Villet et al., 2005).

The American Society for Enteral and Parenteral Nutrition, in conjunction with the Society for Critical Care Medicine, published guidelines for the nutrition care of critically ill adult patients in 2009 (McClave et al., 2009). These guidelines state that efforts should be made to minimize the time a patient is NPO before, during, and after any test or procedure in an effort to prevent inadequate provision of nutrition.

There has been a trend in the past several decades to adopt liberalized preoperative fasting guidelines. A Cochrane review published in 2003 concluded that decreasing preoperative fasting did not increase aspiration events after receiving anesthesia (Brady, Kinn, Stuart, & Ness, 2003). Updated American Society of Anesthesiologists (2011) guidelines from 2011 are outlined in Table 1. The American College of Gastroenterology guidelines for colorectal cancer screening published in 2009 support consuming clear liquids until 2 hours before receiving sedation (Rex et al., 2009). The European Society for Parenteral and Enteral Nutrition recommendations state that it is unnecessary for most patients to fast at midnight before surgery (Braga et al., 2009). Its recommendations further state that preoperative oral carbohydrate loading is recommended for most patients and intravenous carbohydrate administration can be considered for those unable to safely take oral nutrition.

TABLE 1. Preoperative Fasting Recommendations of the American Society of Anesthesiologists for Healthy Patients Undergoing Elective Surgery

Survey of Fasting Practices

A diverse and changing fasting practice has been reported in the literature. In 1996, a survey determined that 49% of responding chairpersons in university anesthesiology programs and medical directors of ambulatory surgery centers in the United States require patients to fast at midnight whereas 24% allow clear liquids up to 4 hours before surgery (64.6% response rate; Green, Pandit, & Schork, 1996). Four years later, a similar survey was sent to members of the Society of Ambulatory Anesthesia in the United States. Of the 59.6% of participants who responded, 62% reported having policies that allow clear liquids 2-3 hours before surgery (Pandit, Loberg, & Pandit, 2000). More recently, Shime et al. evaluated the fasting practices of chief anesthesiologists in anesthesia-teaching hospitals in Japan. Fifty-seven percent of the surveys were returned with a median time for abstaining from liquids of 6-9 hours reported by 90% of the respondents (Shime, Ono, Chihaba, & Tanaka, 2005).

Actual Fasting Practices

Actual fasting time ranged from 3.75 to 29 hours with a mean of 11 hours (SD = 4 hours) in a survey by Chapman (1996). In 2002, a study of 155 patients undergoing elective surgery showed that the majority abstained from liquids for an average of 11.9 +/- 3 hours with 97% of the patients abstaining from liquids for more than 6 hours (Crenshaw & Winslow, 2002). Nearly all patients (91%) were ordered to fast starting at midnight including patients scheduled for the afternoon surgery (79%).

Great efforts have been undertaken to liberalize the antiquated extended preoperative fasting time. Crenshaw and Winslow (2008) completed an aggressive campaign to decrease fasting times in a hospital. This quality improvement initiative included providing physician and nurse education, updating policies and procedures, revising standing order forms, and disseminating current research and practice guidelines for preoperative fasting. A follow-up survey published in 2008 showed that patients abstained from clear liquids an average of 11 +/- 3 hours, which is statistically less than the previously measured 11.9 +/- 3 hours (p < .005), but unfortunately not clinically relevant (Crenshaw & Winslow, 2008). There was no statistical improvement in the percentage of patients instructed to remain NPO after midnight throughout the initiative (p = .19). The instructed time for abstaining from liquids was 9 +/- 3 hours, significantly more than the recommended 2 hours. The authors reported understandable frustration with their results after a strong and comprehensive effort to improve fasting practices. This study highlights difficulties clinicians may experience when trying to liberalize institutional fasting guidelines.

Metabolic Effects of Fasting and Use of Carbohydrate-Rich Clear Liquids

During an overnight fast, breakdown of glycogen to maintain serum blood glucose levels through glycogenolysis can deplete hepatic glycogen stores (Ling & McCowen, 2007). Gluconeogenesis (using muscle protein) is then required for glucose production. Postoperative glucose metabolism is further altered by insulin resistance caused by inhibition of nonoxidative glucose disposal (Soop, Nygren, Myrenfors, Thorell, & Ljungqvist, 2001). Surgery may also result in an increased production of counterregulatory hormones such as catecholamines as well as an increase in inflammatory cytokines leading to a state of hypermetabolism.

Hypermetabolism results in a catabolic environment characterized by breakdown of glycogen, fat, and protein (Smiley & Umpierrez, 2006). Increased levels of insulin and blood glucose further lead to an increased rate of gluconeogenesis. Glucose control is beneficial during conditions of metabolic stress to decrease hospital length of stay (Thorell et al., 1999), decrease postoperative infection rates (Pomposelli et al., 1998), and reduce mortality rates (Finney, Zekveld, Elia, & Evans, 2003). Preoperative administration of carbohydrate is, therefore, expected to increase glucose oxidation rates, which decrease peripheral tissue glucose disposal. In addition, the rate of protein breakdown may be decreased with glucose administration. Improved outcomes from preoperative consumption of carbohydrate-rich clear liquid beverages (carbohydrate loading) appear to be related to both a decrease in insulin resistance and promotion of an anabolic state (Ljungqvist, 2009). These factors appear to ameliorate the surgical stress response.

Use of an intravenous glucose infusion during surgery was found to decrease insulin resistance (Ljungqvist, Thorell, Gutniak, Ha[spacing macron]ggmark, & Efendic, 1994; Nygren, Soop, et al., 1998; Nygren, Thorell, et al., 1998). Unfortunately, peripheral vein administration of concentrated dextrose produces phlebitis (Ljungqvist, 2009). Thus, using a specialized oral carbohydrate-rich clear liquid beverage to stimulate a state similar to that after a meal (fed state) reduces insulin resistance while limiting risks of adverse effects of intravenous dextrose administration. Initial studies used a 12.5% carbohydrate iso-osmolar (300 mOsm/L) beverage that was composed mainly of maltodextrin (a rapidly absorbed polysaccharide).

As previously stated, osmolality may be an important component of gastric emptying and beverages with an osmolality of more than 300 mOsm/L may decrease gastric emptying rate. A study of healthy volunteers determined that the average gastric emptying rate for 400 ml of a preoperative iso-osmolar beverage is about 90 minutes (Nygren, Thorell, Jacobsson, Larsson, Schnell, Hylen, & Ljungqvist, 1995). Residual gastric volumes were evaluated between groups that consumed a carbohydrate-rich drink and those that fasted. Gastric volumes were not statistically different between the groups (p = .61) (Yagci et al., 2008). When soy peptides were added to a carbohydrate-rich clear liquid beverage, this appeared to have no significant impact in gastric emptying rates as well (Henriksen et al., 2003). In a randomized, blinded, three-way crossover study of healthy volunteers, 10 people were provided a carbohydrate-rich clear liquid beverage, an iso-caloric equivalent containing carbohydrate and glutamine, or carbohydrate and lipid containing beverage. The 90% mean gastric emptying time was lowest in the group that received the carbohydrate and lipid-containing beverage (p = .017) (Awad et al., 2011).

In addition to improved gastric emptying, patient satisfaction may improve with administration of carbohydrate-rich clear liquid beverages. Postoperative discomfort may be decreased when patients consume a preoperative carbohydrate-rich clear liquid beverage. Hausel et al. reported decreased nausea and vomiting in laparoscopic cholecystectomy patients who consumed a carbohydrate-rich clear liquid beverage (p < .001) with the highest incidence of nausea and vomiting occurring 12-24 hours after surgery in the fasting group (p < .039) (Hausel, Nygren, Thorell, Lagerkranser, & Ljungqvist, 2005). In an earlier study of 252 elective abdominal surgery patients by Hausel et al. (2001), the carbohydrate-rich clear liquid beverage group experienced decreased rates of hunger (p < .05) and anxiety (p < .001). No statistically significant difference in postoperative appetite (p = .392), pain (p = .228), nausea (p = 1.000), vomiting (p = .336), total minutes of sleep (p = .830), or the number of night movement arousals (p = .846) was observed when a carbohydrate-rich clear liquid beverage was consumed compared to a placebo (Bisgaard et al., 2004).

Postoperative clinical benefits of providing preoperative oral carbohydrate-rich clear liquid beverage also include increased muscle strength (Henriksen et al., 2003; Yuill, Richardson, Davidson, Garden, & Parks, 2005), improved insulin sensitivity (Nygren, Soop, et al., 1998; Nygren, Thorell, et al., 1998; Perrone et al., 2011; Soop et al., 2001; Svanfeldt et al., 2005), and improved cellular immune function (Melis et al., 2006). Only a 5% loss of muscle mass was reported 1 month postsurgery in an elective bowel surgery group that received a carbohydrate-rich clear liquid beverage, compared with a 13% decrease seen in the group that consumed only water (p < .05) (Henriksen et al., 2003). In a study of 15 patients undergoing total hip replacement, the carbohydrate-rich clear liquid beverage group had an 18% decrease in whole-body insulin sensitivity compared with a 43% decrease in the placebo group (p < .05) (Soop et al., 2001).

A decrease in nitrogen losses and a 50% reduction in insulin resistance have been reported in the literature (Ljungqvist, 2009). Table 2 provides additional studies investigating insulin resistance after consuming a preoperative carbohydrate-rich clear liquid beverage. Melis et al. (2006) evaluated carbohydrate beverages in 30 orthopedic surgery patients. The fasting group experienced a significant decrease in human leukocyte antigen expression that correlates to risk an increase in postsurgical infections. Not all studies have shown improved outcomes in the carbohydrate-rich clear liquid group (Bisgaard et al., 2004). However, a summary of literature by Ljungqvist (2009) reported that more than 2000 patients have been studied and more than 2 million patients have received a preoperative carbohydrate-rich clear liquid drink in clinical practice with no apparent adverse effects.

TABLE 2. Literature Reports of Blood Glucose and Insulin Levels After Carbohydrate-Containing Preoperative Beverage

Following major operations, including colorectal surgery, multimodal early recovery programs have been implemented to decrease morbidity and mortality rates (Teewen et al., 2010). These programs recommend avoiding bowel preparation, limiting the use of nasogastric decompression tubes, and utilizing liberal fasting guidelines to ultimately decrease surgical stress and reduce postoperative complication rates (Lassen et al., 2009). Studies evaluating the impact of multimodal early recovery programs using preoperative carbohydrate loading have shown both a decreased length of stay and a decreased 30-day morbidity rate (Walter, Collin, Dumville, Drew, & Monson, 2009).

Barriers to Change

Clinicians may believe that a prolonged fast before surgery improves safety (Crenshaw & Winslow, 2008). One survey of 100 elective surgery participants found that 82% knew fasting practices were related to anesthesia and patient safety. Aspiration and risk for complications were cited as the rationale for fasting by 73% of the nurses surveyed (Baril & Portman, 2007). It may also be easier for clinicians to offer one set of instructions rather than individualizing preoperative recommendations (Crenshaw & Winslow, 2008).

There is a concern that patients would have a difficult time determining what liquids are considered clear liquids and that "NPO at midnight" directions may lead to greater compliance. However, a study of elective colorectal resection patients instructed to consume a carbohydrate-rich clear liquid beverage and complete mechanical bowel preparation showed that 74% of patients complied with preoperative orders (Hendry et al., 2008). Fourteen of the 124 subjects (11%) did not consume the carbohydrate-rich beverage because the medical staff failed to provide the product. Failure to tolerate the oral carbohydrate-rich clear liquid beverage, vomiting, and incomplete records are responsible for nine subjects who did not complete the protocol. Only 2 of 124 subjects refused to consume the carbohydrate-rich drink. Although the instructions for consuming preoperative carbohydrate-rich clear liquids beverages may initially appear more confusing, they are likely not more difficult to understand than other pre- and postoperative nutrition, medication, and lifestyle instructions.

Another concern expressed by surgical facilities is that changing fasting policies would limit flexibility in surgery schedules, but research has failed to substantiate this finding (Murphy, Ault, Wong, & Szokol, 2009). In a survey of ambulatory anesthesiologists in the United States, 65% of anesthesiologists would not delay surgery if toast and tea were consumed 6 or more hours before surgery and another 32% would delay the surgery to later in the day (response rate of 59.6%) (Pandit et al., 2000).

There are certain patient populations that may be need to be excluded from more liberal fasting guidelines. Obese patients have been identified as a population at risk for delayed gastric emptying, but Horowitz et al. reported that there is no difference in gastric emptying of liquids in the obese patient (Horowitz, Collins, Cook, Harding, & Shearman, 1983). Matlby et al. specifically included obese (body mass index > 30 kg/m²) patients undergoing elective surgery in a randomized study to evaluate effects of clear liquid intake 2 hours before the scheduled surgery (Maltby, Pytka, Watson, Cowan, & Fisk, 2004). No difference in gastric volumes was noted (p = .46).

There are also concerns that diabetic patients have delayed gastric emptying; however, a study of 14 diabetic patients who received 400 ml of a 12.5% carbohydrate beverage (preOp) reported that gastric emptying was similar to that of healthy individuals (Nygren et al., 2004). Those with gastrointestinal disorders, delayed gastric emptying, ileus, or intestinal obstruction may not be candidates for liberalized fasting guidelines and use of preoperative carbohydrate-rich clear liquid beverage (Doswell, Jones, & O'Donnell, 2002; Ljungqvist & Soreide, 2003; Crenshaw & Winslow, 2008; Winslow, Crenshaw, & Warner, 2002).

Pregnant women have been considered a population at higher risk for aspiration during anesthesia and may also not be appropriate for preoperative carbohydrate-rich clear liquid beverages (Soreide et al., 2005). Elderly patients and those with dysphagia and inability to safely consume thin liquids should be evaluated for their ability to safely consume any liquid prior to anesthesia.

Recommendations

Based on the available evidence, it would appear that the use of preoperative carbohydrate-rich clear liquid beverage should be a standard of care. However, the most significant limitation to utilizing carbohydrate-rich clear liquids in the United States is the lack of a commercially available, specifically designed, preoperative carbohydrate loading beverage. Most of the studies to date provided 800 ml of a 12.5% carbohydrate iso-osmolar beverage the night before surgery and then another 400 ml 2-3 hours before surgery.

The beverage most frequently used is preOp, which is available in Europe, but is not available in the United States. This drink provides 50 g of carbohydrate per 400 ml. Only one study used Resource Breeze, which is available in the United States (Perrone et al., 2011). To receive 50 g of carbohydrate using apple juice would require the patient to consume 430 ml. However, apple juice is different from preOp in that it is hyperosmolar (690 vs. 240 mOsm/L), which may delay gastric emptying.

Vermeulen et al. (2011) evaluated gastric emptying comparing preOp to a fruit-based lemonade that was hyperosmolar (805 mOsm/kg). Gastric emptying was delayed in the lemonade group; however, this did not reach statistical significance (p = .600). Sports drinks are typically iso-osmolar; however, the patient would need to consume 860 ml to obtain 50 g of carbohydrate. Table 3 reviews the osmolarity and carbohydrate content of various clear liquid beverages. Because preOp has been available in Europe, Ljungqvist (2009) reported that more than 2 million patients had received preOp. One U.S.-based company has a Web site advertising a carbohydrate-rich beverage specifically designed to be used during the preoperative period. However, at the time of this publication, no ordering information, availability, or cost was available on the Web site.

TABLE 3. Carbohydrate (CHO)-Containing Clear Liquid Beverages

Summary

Although NPO at midnight has been standard practice before elective surgery for decades, this practice has little to no scientific support. Liberal fasting guidelines allowing clear liquids 2-3 hours before surgery have been published and supported by professional organizations in the United States since 1999. Surveys of practice policies and actual fasting times indicate that current practice is not consistent with these guidelines.

New evidence has highlighted the importance of oral carbohydrate loading. Providing a 12.5% carbohydrate-rich clear liquid beverage during the preoperative period may reduce insulin resistance by 50% and decrease nitrogen losses (Ljungqvist, 2009; Ljungqvist et al., 1994). Additional benefits include improved patient comfort from decreased rates of postoperative nausea and vomiting along with decreased levels of anxiety, hunger, and thirst. When choosing a preoperative beverage for carbohydrate loading, the clinician should look for the characteristics of the beverage most frequently studied: clear liquid beverage, 50-g carbohydrate per 400 ml, iso-osmolar (<300 mOsm/L), and fiber free.

Most of the studies provided 800 ml of the carbohydrate-containing drink the evening before surgery and then an additional 400 ml 2-3 hours before induction of anesthesia. Certain patients may not be candidates for liberalized fasting guidelines and should receive a presurgical evaluation. These patients include those who are obese, have a gastrointestinal disorder, are elderly, pregnant, exhibit, delayed gastric emptying, have an ileus or intestinal obstruction, or exhibit dysphagia with inability to safely consume thin liquids.

Although clinicians may have assumed that longer fasting would save the patient from surgical complications, starving them and keeping them in the fasted state may be the real danger. Revised fasting guidelines and carbohydrate-rich clear liquid beverages should be utilized by clinicians for healthy individuals undergoing elective surgery.

REFERENCES