Esophagogastroduodenoscopy (EGD) is one of the most commonly performed endoscopic procedures^1,2 and plays a pivotal role in the diagnosis and treatment of upper gastrointestinal disease. Prior to EGD, an adequate fasting period is required to ensure safety and mucosal visibility. Clinical guidelines recommend a 6-hour fasting period for solids and a 2-hour or 4-hour fasting period for liquids.^3,4 However, in practice, subjects whose EGDs are scheduled in the morning are usually designated as nil per os after midnight, and the real fasting duration may be much longer than recommended, especially for those undergoing EGD in the late morning. A long duration of fasting may cause discomfort, including hunger, thirst, weakness, and anxiety,^5-7 or even adverse events such as fluid and glycemic disorders.^8,9

In a previous randomized controlled trial, ingestion of 200 mL caloric clear liquid prior to EGD alleviated general discomfort, while a considerable proportion of subjects still complained of hunger (44%) and thirst (46%).¹⁰ Compared to clear fluid, semifluid could further improve satiety and energy supply without greatly prolonging gastric emptying.¹¹ In this study, we proposed a modified protocol of a 4-hour fasting period for semifluids and a 2-hour fasting period for water (“4+2” protocol) for gastric preparation prior to EGD in late morning, and conducted a randomized controlled trial to investigate participants’ comfort and procedural safety and quality under this protocol, compared with the conventional protocol.

1. Participants

This parallel-group, endoscopist-blinded, randomized controlled trial was conducted at the endoscopy centers of two tertiary referral hospitals from October 2021 to December 2021 (registered at clinicaltrials.gov, identifier: NCT05219136). The inclusion criteria were as follows: (1) between the ages of 18 and 70; (2) underwent unsedated diagnostic EGD; and (3) appointment times from 10:30 AM to 12:00 PM. Exclusion criteria were as follows: (1) a history of upper gastrointestinal surgery; (2) a history of upper gastrointestinal tumor; (3) emergency or therapeutic EGD; (4) pregnant; (5) taking medications that affect gastrointestinal motility; (6) signs and symptoms of gastrointestinal obstruction; and (7) diabetes. This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethical Committee of Changhai Hospital (approval number: CHEC2021-199) and the Affiliated Traditional Chinese Medicine Hospital of Southwest Medical University. Written informed consents were obtained.

2. Randomization and fasting protocols

Eligible participants were randomized 1:1 to either the intervention or control group using a computer-generated random allocation table in the EGD registry. Paper-based instructions and videos regarding the fasting protocols were provided to the participants. In both groups, the participants were asked to maintain their usual diet containing food and liquids prior to the fasting protocol. In the intervention group, the participants adopted the “4+2” protocol. Nil per os was required after midnight. Four hours before the EGD appointment time (approximately 6:30 AM), participants were asked to ingest 300 g of rice porridge (one sale unit of a commercial product, with a volume of 250 mL and total energy of 315 kJ, the nutrient content is listed in Supplementary Table 1) after heating or at room temperature as they wished. The semifluid meal contains only rice and water, and no lipids or coarse fibers were present. The participants were allowed to drink <300 mL (one sale unit of a commercial product) clear water until 2 hours before EGD. In the control group, nil per os was required after midnight, and the participants were allowed to drink clear water (<300 mL) after getting up until 2 hours before EGD. Participants were asked to take photos of all food and water they ingest in the morning to assure compliance.

3. EGD procedures

Endoscopists participating in this study completed over 1,000 EGDs and were blinded to the fasting protocols of the participants. All procedures were performed using GIF-H260 or GIF-H290 endoscopes (Olympus Optical Co. Ltd., Tokyo, Japan). All participants received oral dyclonine mucilage and protease prior to EGD. Dyclonine mucilage is a compound preparation containing dyclonine hydrochloride for local anesthesia and polyoxyethylene polyoxypropylene pentaerythritol ether for anti-foam. Antispasmodic agents were not routinely administered. After endoscope insertion into the stomach, endoscopists assessed the mucosal visibility of the four gastric domains: the antrum, lower gastric body, upper gastric body, and fundus. After lumen cleaning, photos were taken according to a systematic screening protocol for the stomach.¹² If suspicious lesions were found, narrow-band imaging, magnifying endoscopy, or chromoendoscopy was used for further observation, and biopsy was performed as indicated.

4. Data collection

Data on age, sex, height, weight, medical history, and indications for EGD were retrieved from the registration system. Prior to EGD, participants responded to a questionnaire about their comfort during the fasting period. Within 30 minutes after EGD, participants responded to a questionnaire about their comfort during the EGD and whether they would be willing to adopt the same fasting protocol for another EGD if necessary. Comfort during the fasting period and the EGD procedure was evaluated separately using a visual analog scale (VAS) from 0 to 10 points, with 0 indicating the least and 10 indicating the most discomfort. VAS scores of 1–3, 4–6, and 7–10 were defined as mild, moderate, and severe discomfort, respectively. Participant satisfaction was defined as a VAS score of ≤3. The endoscopists responded to a structured interview about mucosal visibility immediately after completing each procedure. The visibility score of each gastric domain ranged from 1 to 4 and was defined as follows: score 1, no adherent mucus on the gastric mucosa; score 2, little mucus on the gastric mucosa, but not obscuring vision; score 3, large amounts of mucus on the gastric mucosa requiring less than 50 mL of water to clear; and score 4, large amounts of mucus on the gastric mucosa requiring more than 50 mL of water to clear. The sum of visibility scores of the four gastric domains was defined as the total visibility score (TVS) for each participant.^13-15 The EGD examination time was retrieved from the endoscopic database. Adverse events including aspiration, bleeding, infection, perforation, etc., were documented during the study period.

5. Outcomes

The primary outcome was participant’s comfort during the fasting period prior to EGD. The secondary outcomes included participants’ comfort during the EGD procedure, TVS, and adverse events.

6. Statistical analysis

The sample size calculation was based on the assumption that the proportion of participant satisfaction increased from 58% to 82% in the intervention group compared with the control group,¹⁰ under an α of 0.05, and a power of 0.8. At least 60 participants were required for each research group. We performed the Shapiro-Wilk test for normality. Continuous variables were described as mean±standard deviation and compared using the Student t-test if normally distributed, and otherwise described as median (quartiles) and compared using the Mann-Whitney U test. Categorical variables were compared using the chi-square test or the Fisher exact test. A p-value <0.05 was considered statistically significant. We performed both intention-to-treat (ITT) and per-protocol (PP) analyses and primarily reported results from the ITT analysis. We further prespecified subgroup analyses of participant age, sex, and prior experience of EGD for satisfaction during the fasting period. All analyses were performed using SPSS 26.0 (SPSS Inc., Chicago, IL, USA).

1. Baseline characteristics

A total of 246 participants from the two centers were assessed for eligibility, and 32 were excluded (Fig. 1). Six endoscopists participated in the study. Of the 214 participants who underwent randomization, 106 were allocated to the intervention group (“4+2” protocol) and 108 were allocated to the control group (conventional protocol). In the intervention group, seven participants did not ingest semifluids and ingested only water, two did not respond to the interview on comfort after EGD, and one had incomplete EGD due to intolerance. In the control group, three participants did not respond to the interview after EGD and two had incomplete EGD (one with gastric food retention and one due to intolerance). A total of 214 and 199 participants were included in the ITT and PP analyses, respectively. The baseline characteristics of the participants in the intervention and control groups were evenly distributed in both the ITT analysis (Table 1) and PP analysis (Supplementary Table 2).

Table 1 . Characteristics of the Subjects.

Characteristics	Intervention group (n=106)	Control group (n=108)	p-value
Age, yr	50.0 (40.0–59.3)	49.0 (40.0–58.0)	0.570
Sex			0.652
Male	51 (48.1)	59 (54.6)
Female	55 (51.9)	49 (45.4)
Education level			0.609
Less than 12 yr of education	6 (5.7)	9 (8.3)
High school graduate	54 (50.9)	49 (45.4)
College and above	46 (43.4)	50 (46.3)
Prior experience of EGD			0.242
Yes	35 (33.0)	44 (40.7)
No	71 (67.0)	64 (59.3)
Body mass index, kg/m2	23.3±2.8	23.2±2.8	0.647
Medical history			0.494
Hypertension	11 (10.4)	14 (13.0)
Hyperthyroidism	3 (2.8)	5 (4.6)
Others	9 (8.5)	7 (6.5)
Indication of EGD			0.466
Screening	28 (26.4)	18 (16.6)
Diagnosis	55 (51.9)	60 (55.6)
Surveillance	23 (21.7)	30 (27.8)
Endoscopist			0.312
I	10 (9.4)	18 (16.7)
II	10 (9.4)	9 (8.3)
III	13 (12.3)	10 (9.3)
IV	15 (14.2)	11 (10.2)
V	19 (17.9)	17 (15.7)
VI	39 (36.8)	43 (39.8)
Center			0.649
I	67 (63.2)	65 (60.2)
II	39 (36.8)	43 (39.8)

Data are presented median (interquartile range), number (%), or mean±SD..

EGD, esophagogastroduodenoscopy..

Figure 1. Flow diagram of the study. The “4+2” protocol refers to the 4-hour fast for semifluid and 2-hour fast for water protocol. The conventional protocol refers to the 6-hour fast for solid and 2-hour fast for water protocol.
PPI, proton pump inhibitor; GI, gastrointestinal; EGD, esophagogastroduodenoscopy.

2. Comfort

The VAS scores during fasting were significantly lower in the intervention group than in the control group (1.0 [1.0–2.0] vs 3.0 [1.0–4.0], p<0.001). The proportion of satisfaction was significantly higher in the intervention group than in the control group (86.8% vs 63.9%, p=0.002). The VAS scores during the procedure were significantly lower in the intervention group than in the control group (3.0 [1.25–5.0] vs 4.0 [2.0–6.25], p=0.002). The proportion of satisfaction was significantly higher in the intervention group than in the control group (59.4% vs 45.4%, p=0.039) (Table 2). A significantly higher proportion of participants in the intervention group (84.9% vs 75.0%, p=0.037) claimed to be willing to adopt the same fasting protocol for another EGD, if necessary. Similar results were found in the PP analysis (Supplementary Table 3). Subgroup analyses showed that no differences in improvement of satisfaction during the fasting period were found among participants with different age, sex, and prior EGD experience (Supplementary Table 4).

Table 2 . Comfort of the Subjects During the Fasting Period and Esophagogastroduodenoscopy Examination.

Variable	VAS score	Fasting period, No. (%)				During examination, No. (%)
		Intervention group (n=106)	Control group (n=108)	Total (n=214)		Intervention group (n=106)	Control group (n=108)	Total (n=214)
Satisfied	0	21 (19.8)	12 (11.1)	33 (15.4)		5 (4.7)	3 (2.8)	8 (3.7)
	1–3	71 (67.0)	57 (52.8)	128 (59.8)		58 (54.7)	46 (42.6)	104 (48.6)
Dissatisfied	4–6	11 (10.4)	34 (31.5)	45 (21.0)		37 (34.9)	32 (29.6)	69 (32.2)
	7–10	3 (2.8)	5 (4.6)	8 (3.7)		6 (5.7)	27 (25.0)	33 (15.4)

VAS, visual analog scale..

3. Visibility and safety

The TVS of the intervention and control groups were similar (5.0 [4.0–5.0] vs 4.0 [4.0–5.0], p=0.266). The visibility scores of all four gastric domains did not differ significantly (Table 3, Supplementary Table 5). Similar results were found in the PP analysis (Supplementary Tables 6 and 7). The examination time did not differ significantly between the intervention and control groups (308 seconds [293–317 seconds] vs 311 seconds [288–333 seconds], p=0.522). No adverse events, including aspiration, bleeding, infection, or perforation, were documented during the study period.

Table 3 . Mucosal Visibility Scores in Different Gastric Domains and Total Mucosal Visibility Scores of the Intervention and Control Groups.

Group	Intervention group	Control group	p-value
Antrum	1.0 (1.0–1.0)	1.0 (1.0–1.0)	0.647
Lower gastric body	1.0 (1.0–1.0)	1.0 (1.0–1.0)	0.594
Upper gastric body	2.0 (1.0–2.0)	1.0 (1.0–2.0)	0.706
Fundus	1.0 (1.0–2.0)	1.0 (1.0–1.0)	0.221
Total visibility score	5.0 (4.0–5.0)	4.0 (4.0–5.0)	0.266

Data are presented as median (interquartile range)..

Patient comfort is an important aspect of the clinical quality of endoscopy.¹⁶ In this study, we proposed a “4+2” protocol for pre-EGD fasting and conducted a double-center, parallel-group, randomized controlled trial. We found that the participants’ comfort was significantly improved during both the fasting period and EGD procedure under this modified protocol without compromising mucosal visibility and safety. To the best of our knowledge, this is the first report on the application of semifluid in the gastric preparation of EGD.

Current guidelines recommend a 6-hour fasting period for solids and 2- or 4-hour fasting period for liquids prior to EGD.^3,4 However, in practice, participants usually fast for longer hours. For procedures performed in the morning, nil per os after midnight is often required, resulting in 8-hour to 12-hour fasting periods for solids. A semifluid diet provides more energy and satiety than liquids and is associated with a significantly shorter gastric emptying time than a solid diet.¹¹ Furthermore, participants in various countries are accustomed to a semifluid diet for breakfast, including porridge, cereal, and sauce. Data on gastric emptying and adequate fasting time for semifluids are lacking. We used the mean value of 6 hours (for solids) and 2 hours (for liquids) as the fasting time for semifluids and proposed the “4+2” protocol for gastric preparation.

As measured by the VAS score (1.0 [1.0–2.0] vs 3.0 [1.0–4.0], p<0.001) and rate of satisfaction (86.8% vs 63.9%, p=0.002), the participants’ comfort during the fasting period was significantly improved under the “4+2” protocol. The European guideline states that for endoscopies in the afternoon, patient satisfaction may be increased if a small breakfast is allowed.³ The results of the present study provide evidence supporting this recommendation. Koeppe et al.¹⁰ found that participants’ level of comfort during fasting was higher during a 2-hour than during an 8-hour fast for water, including measures of anxiety, general discomfort, hunger, and weakness. De Silva et al.¹⁷ concluded that discomfort prior to EGD was significantly lower in 1-hour water fast group than in a 6-hour fasting group. Similarly, we found that the participants’ comfort during the EGD procedure was significantly improved in terms of both the VAS score (3.0 [1.25–5.0] vs 4.0 [2.0–6.25], p=0.002) and the proportion of satisfaction (59.4% vs 45.4%, p=0.039). Faria et al.¹⁸ found that the level of stress was significantly lower in the 2-hour fasting group than in the 8-hour fasting group for video cholecystectomy. Similarly, hunger, anxiety, nausea, vomiting, throat discomfort, bloating, etc., prior to and during EGD were all stressors for participants. We speculate that the “4+2” protocol could alleviate the level of stress by increasing energy supply and decreasing hunger, weakness, and anxiety, thereby making participants less vulnerable to stimuli during EGD, and therefore improving the level of comfort. However, we could not identify the exact reasons for improvements during the examination based on the current study. Consequently, a higher proportion of participants in the “4+2” protocol group (84.9% vs 75.0%, p=0.037) claimed to be willing to adopt the same fasting protocol for another EGD than the conventional group.

Fasting is also a prerequisite for clear visualization of the gastric mucosa. In another study by De Silva et al.,¹⁹ the protocol of fasting 6 hours after a rice meal led to poor endoscopic vision in 28.6% of participants, suggesting that it may be inappropriate to further shorten the fasting time for solid food. In the present study, the mucosal visibility of the “4+2” protocol group and the control group was similar (TVS: 5.0 [4.0–5.0] vs 4.0 [4.0–5.0], p=0.266), suggesting that the rice porridge meal resulted in gastric emptying within 4 hours. De Silva et al.¹⁷ found that water ingestion 1 hour prior to endoscopy resulted in good endoscopic visibility, and the volume and pH of gastric fluid were not significantly different between 1-hour and 6-hour fasted groups. Koeppe et al.¹⁰ suggested that mucosal visibility was high in both the 2-hour and the 8-hour water fast groups. Although the visibility score in the 8-hour fast group was statistically superior to that of the other groups, the difference was considered clinically irrelevant. Moreover, the examination time was also similar between the two groups in the current study, suggesting that no additional time was spent on removing the gastric fluid and cleaning the lumen. No adverse events were documented during the research period, suggesting the “4+2” protocol did not affect the safety of EGD procedures.

There are several concerns regarding the inclusion criteria during study design. First, we only included participants undergoing unsedated EGD due to concern for the risk of gastroesophageal regurgitation and aspiration under sedation. However, the “4+2” protocol resulted in a low volume of residual gastric fluid and no adverse events. Therefore, this protocol can be applied to patients undergoing sedated EGD. Second, patients with diabetes were excluded, as the ingestion of rice porridge could result in rapid hyperglycemia. However, patients with diabetes are also at an increased risk of hypoglycemia during the fasting period. Moreover, a considerable proportion of patients with diabetes also have gastroparesis. Therefore, the fasting protocol for this specific group of participants requires further investigation. Third, we included patients with EGDs scheduled late in the morning. We did not include EGDs in the early morning (8 AM to 10 AM) because adopting the “4+2” protocol would require getting up at from 4 AM to 6 AM. For EGDs in the early morning, the time between getting up and examination is around 2 hours, and the current standard protocol of 2 hours fasting for water may be more proper for this group of subjects. We suggest that individualized fasting protocols could be recommended for subjects undergoing EGD based on their time of examination and personal preference.

The limitations of this study include the weakened external validation of participants undergoing intravenous sedation or with diabetes, because of the current eligibility criteria. Furthermore, we did not objectively measure the volume and pH of the residual gastric fluid. Future studies may focus on fasting time of different types and volumes of semifluid or other sub-populations of participants undergoing EGDs to provide patient-oriented, individualized protocols for gastric preparation.

In conclusion, we propose the “4+2” protocol for gastric preparation prior to unsedated EGD to avoid prolonged fasting. Under this fasting protocol, participants’ comfort is significantly improved during both the fasting period and the EGD procedure without compromising mucosal visibility and safety. Further studies on subjects with intravenous sedation and diabetes are still required.

Supplementary materials can be accessed at https://doi.org/10.5009/gnl220170.

This work was supported by the clinical research fund of Changhai Hospital of Naval Medical University (grant number: 2019YXK009).

No potential conflict of interest relevant to this article was reported.

Study concept and design: L.X., L.W.W., Z.S.L. Data acquisition: M.X.C., Y.G., L.L., W.F., Y.L.W. Data analysis and interpretation: M.X.C., Y.G. Drafting of the manuscript: M.X.C., Y.G. Critical revision of the manuscript for important intellectual content: L.W.W., L.X. Statistical analysis: M.X.C., Y.G. Obtained funding: L.W.W., Z.S.L. Administrative, technical, or material support; study supervision: L.X., L.W.W., Z.S.L. Approval of final manuscript: all authors.