Fasting is the act of refraining from food or drinks for a specific purpose. This could be due to religious or personal reasons, but could also be due to requirements for certain activities such as undergoing elective surgery. According to Mathews et al (2013), preoperative fasting, originated with Lister (1883), who suggested that it was desirable for no solid food to be in the stomach when chloroform was given and that it was beneficial to give a cup of tea about 2 hours prior to surgery. Mendelson (1946) further discovered that vomiting during induction of anaesthesia brought about respiratory distress, resulting from pulmonary aspiration of gastric hydrochloric acid. This discovery led to the common practice of ‘nil by mouth’ from midnight before surgery (Mathews et al, 2013).
Today, this is a common practice observed in the clinical environment: if procedures are scheduled for the morning, patients are required to fast from 12 midnight the night before. This preoperative fasting policy was perceived to be easy to administer and enable alterations in the theatre list (Brady et al, 2003).
Patients routinely fast for elective surgery to help reduce gastric volume and acidity (Borges et al, 2011). This subsequently helps prevent any acute respiratory tract obstruction such as aspiration pneumonia (Xu et al, 2017). According to the World Health Organization (WHO) (2009), aspiration is a significant risk to all patients undergoing sedation or anaesthesia for elective surgery. Preoperative fasting is an essential practice, which is adhered to in clinical practice; however, research has suggested that prolonged preoperative fasting in patients undergoing elective surgery is not safe (Maltby, 2006).
Guidelines produced by the American Society of Anesthesiologists (ASA) Committee (ASA, 2017) recommends that all healthy adults undergoing elective surgery should be allowed to consume solid meals 8 hours before surgery, a light meal such as toast and tea 6 hours before surgery, and unlimited clear fluids (but not alcohol) up to 2 hours before surgery. The minimum fasting periods recommended by the ASA (2017) are shown in Table 1. The Royal College of Nursing (RCN) (2013) and the European Society of Anaesthesiology (ESA) (Smith et al, 2011) both recommend intake of water up to 2 hours before induction of anaesthesia because this helps to improve patient wellbeing. Both the RCN and ESA also recommend a minimum preoperative fasting time of 6 hours for solid foods and milk (Smith et al, 2011; RCN, 2013). However, Gunawardhana (2012) has reported that most hospitals inappropriately adopt the traditional practice of requiring patients to fast from midnight on the night before surgery.
| Ingested material | Minimum fasting period |
|---|---|
| Clear liquids | 2 hours |
| Non-human milk | 6 hours |
| Light meal | 6 hours |
| Fried foods, fatty foods or meat | Additional fasting time of 8 hours or more |
Research shows that most patients fast for prolonged periods of time before undergoing elective surgery. The causes for this have been attributed to delays in surgical theatres, maximisation of fasting by patients who believe it is healthy, and changes in daily operation schedules (de Aguilar-Nascimento and Dock-Nascimento, 2010). Jodlowski and Dobosz (2014) suggested that most patients fast for longer than recommended or prescribed, with times ranging upwards of 12 hours.
Extended preoperative fasting is thought to have adverse clinical outcomes after surgery (Bopp et al, 2011). Evidence suggests that fasting for prolonged periods leads to dehydration, distress, anxiety, hunger and thirst during the preoperative waiting period (Hamid, 2014). There are assumptions that prolonged fasting compromises the metabolic pathways of patients (Longo and Mattson, 2014), thereby triggering biochemical reactions that may initiate gluconeogenesis, lipolysis and proteolysis (Ljungqvist, 2009). The RCN (2013) suggests that higher risk patients should follow the same preoperative fasting regimen as healthy adults by adhering to the ASA and ESA guidelines, which are considered to be the international gold standards, unless contraindicated. The ESA guideline recommends that high-risk patients with gastro-oesophageal reflux should also follow the same preoperative fasting regimen. However, overall anaesthetic management may vary due to potential delays in gastric emptying in these patients (Smith et al, 2011). As reported by Smith et al (2011), studies on preoperative fasting do not include high-risk patients, hence there is a limitation in definitive evidence for this practice in such patient groups.
The volume of administered clear fluids up to a couple of hours prior to surgery is considered to not have an impact on a patient's residual gastric volume and gastric pH (RCN, 2013). Emergency surgical patients, on the other hand, have a higher risk of postoperative morbidity and mortality than those having elective procedures, hence it is important to identify the risk factors and reduce their impact (Lake, 2015). The RCN and ESA guidelines both recommend that emergency patients are treated similarly to patients undergoing elective procedures.
Due to current recommendations on preoperative fasting, the findings of the review reported here could be useful for gaining knowledge about preoperative fasting practices, and why these are adopted. Staff, including nurses, educate patients about fasting requirements in the preassessment and preoperative assessment phases, hence fasting practices by patients may reflect the level of patient education and effectiveness of the information given by health professionals. Consequently, this review also sought to address health professionals' knowledge and preferences regarding preoperative fasting.
Search strategy
A systematic literature search was carried out using the following key words:
The databases searched included CINAHL Plus, PubMed, Medline, the Cochrane Central Register of Controlled Trials, Science Direct, Sage Journals and Embase. The time span for the search was originally limited to a 10-year period (2007–2017) to enable retrieval of more current literature material. Seminal literature was accessed due to scarcity of recent UK literature and the limited number of research studies on the subject. Although this information was read to gain knowledge and depth on the subject being explored, it was not included in this review.
Due to the dearth of UK literature the search was also extended to incorporate international literature to help gain a wider perspective, irrespective of differing cultures and healthcare systems. The search yielded little international diversity; it identified studies from Brazil, Egypt and Turkey. The search inclusion and exclusion criteria are presented in Table 2. Figure 1 presents the number of studies yielded at each stage of the retrieval process.
| Search inclusion criteria | Search exclusion criteria |
|---|---|
| All adult populations ranging from 18 years and above | Individuals participating in the studies were aged under 18 years |
| All studies written and transcribed in English | Studies written in a language other than English |
| Evidence-based (EB) research studies relevant to the topic, including randomised controlled trials and peer-reviewed studies | Non-evidence-based research or EB literary materials such as literature reviews |
| All studies involving human participants | All studies involving animal subjects |
| Documents and articles published on relevant databases, including the Royal College of Nursing, World Health Organization, European Society of Anaesthesiology, National Institute for Health and Care Excellence and the Department of Health and Social Care | Documents and articles published on non-academic platforms, including Wikipedia and social media sites |
Findings
The studies remaining after the exclusion criteria were applied are summarised in Table 3.
| Study | Study design | Sampling method | Data collection method | Key findings | Methodological strength | Study limitations |
|---|---|---|---|---|---|---|
| de Aguilar-Nascimento et al, 2014 Brazil | Correlational study | A probability consecutive sampling method was used to study patients (n=3715) | Data were collected by local co-ordinators in each hospital using a study protocol designed by a team of anaesthetists and surgeons | Most hospitals used traditional rather than modern fasting guidelines. Actual preoperative fasting time exceeded prescribed fasting times | Large sample size used, which enhances the generalisability of the findings. Statistically significant results (P<0.001) suggested validity of findings | Data collection method reported lacked clarity on specific data collection method used and could be confusing to the reader |
| Salman et al, 2013 Egypt | Questionnaire survey design | A simple random sampling method was used to study anaesthetists (n=101) | Data were collected using questionnaires designed by Breuer et al (2010) | Most anaesthetists were aware of new international preoperative fasting guidelines. However, they still strictly followed the traditional ‘nil by mouth’ practice from midnight on the day of surgery | Use of simple random sampling allows an equal chance of selection in sample population. Ethical considerations were addressed, ensuring anonymity | Low participant response rate, which could be the result of having to complete lengthy, complex questionnaires; these may also introduce non-response bias |
| Tosun et al, 2015 Turkey | A descriptive, cross-sectional survey | A simple random sampling method was used to study patients (n=99) | Data were collected using the visual analogue scale (VAS) developed by Hayes and Patterson (1921) and the State-Trait Anxiety Inventory self-assessment tool developed by Spielberger (1970) | Fasting after midnight preoperatively is a routine practice that continues to persist, resulting in discomfort for patients. Mean fasting time was found to be 14 hours. Patients fasting longer than 12 hours had higher pain scores for thirst, hunger, nausea and higher scores than those fasting 12 hours or less | Value of P=0.03 made findings statistically significant, hence validity and generalisability of results. Study design was quick, easy and cheap to use | Sampling method used was not clearly defined and potentially misleading. Use of a VAS for data collection can result in bias because reliability of the scale is low |
Data collection
The three studies examined used a variety of data collection methods; these included the use of a developed study protocol (de Aguilar-Nascimento et al, 2014), questionnaires (Salman et al, 2013) and a visual analogue scale (VAS) Tosun et al (2015).
Fasting times
de Aguilar-Nascimento et al (2014) found that the traditional practice of nil by mouth from midnight on the day of surgery predominated in practice, despite the recommendations of the ASA, ESA and RCN that this is not best practice. Actual fasting times also exceeded prescribed fasting times.
In contrast, Salman et al (2013) demonstrated an awareness of the current ASA, ESA and RCN fasting guidelines among anaesthetists. However, despite this knowledge, they still practised the traditional nil by mouth from midnight.
The research by Tosun et al (2015) found that the traditional nil by mouth approach from midnight was a continuous/repeated practice. They also found extended fasting to be associated with great patient discomfort.
Participant numbers
The study populations in the three studies ranged from 99 (Tosun et al, 2015) and 101 (Salman et al, 2013) to 3715 (de Aguilar-Nascimento et al, 2014).
Duration of fasting, prescribed v actual fasting times
de Aguilar-Nascimento et al (2014) conducted a correlational study to investigate the gap between prescribed and actual preoperative fasting times in hospitals in Brazil and the factors associated with this gap. A probability, consecutive sampling method was used to study the preoperative practices applied in 25 hospitals; the study population consisted of adults aged 18 to 94 years, with a median age of 49 years (n=3715). Of these, women comprised 58% (n=2157) of total participants.
Data were collected by local co-ordinators in each hospital using a study protocol designed by a team of anaesthesiologists and surgeons at the Julio Muller University Hospital, in Cuiabá, the state capital of Mato Grosso. The median preoperative fasting time was found to be 12 hours. The findings also revealed that in some cases (1.4% n=52) actual fasting time exceeded, or was equal to, 24 hours for reasons that included surgery being postponed until the following day. The investigators concluded that actual preoperative fasting time was significantly longer than prescribed fasting time in Brazilian hospitals; the P value was found to be <0.001, which is a methodological strength indicating that the findings are statistically significant, and hence valid and reliable. The large sample size was a strength of the study; according to Gravetter and Forzano (2009) the larger the sample size, the more likely it is that values obtained from the sample will be similar to the actual values for the population. A limitation of this study was the lack of clarity on the data collection methods used.
Similarly, Tosun et al (2015) carried out a descriptive, cross-sectional survey in Turkey to assess the effects of preoperative fasting and liquid limitation. A simple random sampling method was used to study 99 patients, of whom 59.60% (n=59) were women. Data were collected using the VAS developed by Hayes and Patterson (1921) and the State-Trait Anxiety Inventory (STAI), a self-assessment tool developed by Spielberger (1970).
The main findings showed that mean preoperative fasting time was 14.70+3.14 hours, and mean preoperative liquid limitation time was 11.25+3.74 hours. In patients fasting for longer than 12 hours, mean VAS scores for thirst, hunger, nausea and pain were higher in the preoperative period than the mean scores of patients fasting for less than 12 hours. The difference between the two groups was statistically significant (P=0.03). The investigators concluded that receiving nothing by mouth after midnight preoperatively is part of routine ongoing practice and results in patient discomfort. A methodological strength reflects the significance value of (P=0.03), demonstrating that the findings are statistically significant, thereby increasing the reliability and validity of this study.
The use of cross-sectional studies is also useful, which Sedgwick (2014) has suggested is relatively quick, easy and cheap to perform without any loss to follow-up because participants are interviewed once only. A limitation of the Tosun et al (2015) study is a lack of clarity on the sampling method used. In addition, the reliability of the VAS is low, owing to participants' need to report their pain score a number of times (Sharma, 2014), requiring them to recall their initial score against which to gauge their subsequent pain levels—and this is not straightforward.
Awareness and practice of preoperative fasting by health professionals
Salman et al (2013) carried out a study in Egypt designed to examine preoperative fasting routines advised by anaesthetists for adults undergoing elective surgery. A simple random sampling method enrolled 101 anaesthetists. Data were collected using the questionnaire devised by Breuer et al (2010). The study found that 18.9% of anaesthetists usually recommended a preoperative fasting period for solid food of 6 hours and a period of 2 hours for drinking clear fluids. The main findings also showed that 72% of anaesthetists still practised traditional nil by mouth from midnight, despite most of them being aware of current ASA and ESA preoperative fasting guidelines.
The use of simple random sampling allows ease in selecting the study's sample, and ensures that everyone in a population has the same probability of being selected. This allows for an unbiased representation of the population (Baran and Jones, 2016:111). In the Salman et al (2013) study, however, the investigators addressed ethical issues and concerns by informing participants that they would remain anonymous. Nieswiadomy (2007) explained that, if a researcher is unable to guarantee anonymity, then it is necessary to address confidentiality relating to protecting participants' identity.
Out of a total of 147 anaesthetists, 101 correctly completed the questionnaire and were included in the study, a response rate of 69%. This response rate could have been due to the length and complexity of the questionnaire, which may have been offputting to potential participants (Iglesias and Torgenson, 2000). The investigators' acknowledged a further failing in that the questionnaire lacked clear definitions about certain key aspects, such as the definition of aspiration. They also acknowledged that a response rate of 69% could introduce a non-response bias that cannot be excluded for this study. The potential implication of this is that it would not be possible to generalise the findings, particularly in healthcare studies. Cheung et al (2017) suggest that people with poorer health and understanding tend to avoid participating in health surveys.
Discussion
The evidence shows that preoperative overnight fasting is associated with the beliefs of health professionals rather than scientific evidence (Pimenta and de Aguilar-Nascimento, 2014). Remaining ‘nil by mouth’ from midnight for surgery was introduced at a time when anaesthetic techniques were undeveloped. The aim was to ensure complete gastric emptying and prevent complications during surgery such as aspiration pneumonia, which can lead to death (Warner, 2000; Maltby, 2006).
More recent recommendations have advised reduced fasting, which is believed to have benefits such as a reduction in morbidity rates, thereby decreasing length of hospital stay following surgery (Varadhan et al, 2010). However, traditional preoperative fasting practices are still considered essential by many anaesthetists globally simply based on old norms (de Aguilar-Nascimento and Dock-Nascimento, 2010). The study by de Aguilar-Nascimento et al (2014) found median preoperative fasting time to be 12 hours. The findings also revealed that, in some cases, actual fasting time was equal to or exceeded 24 hours. This was found to be due to surgery being postponed to the next day for reasons such as operations on theatre lists over-running. Patients accept current preoperative fasting practices without question and without an understanding of the purported underlying rationale. It is therefore essential to provide patient education during the preoperative phase.
There is also evidence showing that fasting for extended periods has its own consequences, with patients experiencing thirst, hunger, nausea, pain, and a potential increase in the length of hospital stay (Tosun et al, 2015). This patient discomfort indicates that extended fasting should be avoided. Both the ESA guidelines (Smith et al, 2011) and the RCN (2013) guidance cites the benefits of drinking clear fluids up to 2 hours prior to surgery, which needs to be encouraged in clinical practice, especially during preoperative assessments.
A study by Falconer et al (2014) in the UK used survey questionnaires to examine compliance with current best practice guidelines to identify areas for improvement in adult patients undergoing elective general orthopaedic, gynaecology and vascular surgery at the Royal Infirmary of Edinburgh. The study's conclusions are keeping with the findings of this review. The researchers found that for elective patients median fasting from solids was 13.5 hours and median fasting from fluids was 9.36 hours. They concluded that elective patients fasted for longer than recommended, confirming that clinical practice is slow to change. The use of universal fasting instructions and patient choice are factors that unnecessarily prolong preoperative fasting (Falconer et al, 2014). Falconer et al's (2014) findings bear similarities to those from the other studies discussed in this article. This shows that preoperative practice in the UK is similar to that of other countries and allows for generalisations, for example that patients across the globe tend to fast longer than prescribed or than is recommended. Health professionals' awareness and practice of preoperative fasting is vital in the preoperative environment. As already discussed, there are prescribed and actual fasting times, which vary from practitioner to practitioner.
A patient and staff survey in a hospital in Germany that assessed the implementation of new fasting recommendations discussed in the review, including the ASA recommendations, found that 76% of staff expressed concern about increased patient risk related to reduced fasting (Breuer et al, 2010). Anaesthetists were found to possess significantly more knowledge about reduced preoperative fasting than other health professionals. This identifies possible discrepancies in knowledge about preoperative fasting between health professionals. The findings of this review show that a majority of anaesthetists still practised traditional ‘nil by mouth’ fasting from midnight, irrespective of their awareness of modern pre-operative fasting guidance and recommendations.
Future studies should seek to investigate the reasons for the aforementioned phenomenon, and the reasons why anaesthetists are hesitant to adopting modern fasting recommendations. Nurses working in roles that involve preoperative assessment work as patient advocates in the preoperative phase and need to liaise with anaesthetists about what is best practice for patients. It is also important for future research to focus on nurses' knowledge and awareness about preoperative assessment, particularly relating to fasting requirements before elective surgery.
Limitations
The dearth of UK literature limits comparisons between the UK healthcare system and that of other countries, since the studies included in this review were conducted in Brazil, Egypt and Turkey.
Conclusion
More studies are needed to explore nurses' knowledge of guidance on reduced preoperative fasting and modern guidelines.
Contrary to the findings of the studies included in this review, a study by Sada et al (2014) explored the oral administration of carbohydrate-rich liquid drinks during the fasting period (Table 4), which they considered to be clear fluids that patients could take up to 2 hours before surgery. In the postoperative phase, their study showed improvement in the health and wellbeing of participants given these drinks. However, the study found no significant change in length of hospital stay. A randomised controlled trial by Dock-Nascimento et al (2012) found that the preoperative intake of carbohydrate beverages improved insulin resistance and antioxidant defences, and increased the inflammatory response following surgery. These findings provide some evidence of the benefits of administering a carbohydrate-rich beverage in the fasting period. Carbohydrate drinks contain the nutrients necessary to maintain the body's energy intake, so patients consuming these drinks up to 2 hours before surgery will maintain their intake during the preoperative wait. Although this was not the main focus of this review, it is an interesting area for future studies to explore and for clinicians to take into account when considering preoperative practices for elective surgery.
| Study | Study design | Sample size | Key findings |
|---|---|---|---|
|
Sada et al, 2014
|
Prospective, double-blind, randomised controlled trial | 142 | Oral administration of carbohydrate rich liquid drinks improves the wellbeing of patients undergoing cholecystectomy, but the effect is less evident in patients undergoing colorectal surgery. |
|
Dock-Nascimento et al, 2012
|
Randomised controlled trial | 48 | Preoperative intake of a GLN enriched CHO beverage appears to improve insulin resistance and antioxidant defences and decreases the inflammatory response after laparoscopic cholecystectomy |
Future studies should include high-risk individuals, such as those with gastro-oesophageal reflux and emergency admissions. Most studies have been focused mainly on surgery involving colorectal and laparoscopic or open cholecystectomy. It is vital that future research focuses on other types of elective surgery to identify and explore potential risks. Men are under-represented in most studies, which can introduce a gender bias. It is recommended that further studies on the subject area strive to recruit equal numbers of male and female participants to strengthen representation across populations.
This review has identified gaps in UK literature on preoperative fasting, and it is vital that more research is carried out in the UK in this area.
Nurses and other health professionals working in the surgical environment should be educated on the latest international fasting guidelines so they can relay this information effectively to patients preoperatively. The ASA (2017) considers that drinking clear fluids up to 2 hours before surgery is considered safe, so nurses and other clinicians may consider giving patients small amounts of water or clear fluids in the preoperative waiting period when there are foreseen delays and the wait is likely to exceed 2 hours.
Most patients wait for lengthy periods during the preoperative wait, with surgeries sometimes being cancelled as observed in clinical practice. As a result, consideration could be given to preoperative hydration measures. This could be in the form of offering preoperative patients some sips of water on an hourly or 2-hourly basis while they wait. This can be incorporated in clinical practice through effective communication among teams. The risks and benefits of this practice is another area that future studies could address.
The review revealed that patients were required to fast for longer than prescribed, despite the fact that drinking clear fluids 2 hours before surgery is safe and does not put patients at risk of aspiration. It was also found that anaesthetists preferred to follow the traditional nil by mouth routine from midnight, rather than follow modern recommendations, which is reflected in current clinical practices on preoperative fasting. There was, however, a paucity of studies reporting on health professionals' knowledge and attitudes about preoperative fasting.