**1. Introduction**

14 Gastrointestinal Endoscopy

http://www.worldgastroenterology.org/helicobacter-pylori-in-developing-

countries.html

WGO (2007). Standards in gastroenterology training.

WHO Geneva (2008). World Health Statistics. Available from http://www.who.int/whosis/hostat/ 2008/en/index.html

> Current endoscopic guidelines advice a 6 hour fast for solids and a 4 hours fast for liquids before UGIE1, while most anaesthesia guidelines advice a 6 hour fast for solids and a 2 hour fast for clear liquids1,2. The purpose for fasting before UGIE is two fold. The first is the same as in anaesthesia prior to surgery, to prevent the aspiration of food contents3. The second, which is specific for endoscopy, is to provide clear vision4. However, prolonged fasting may result in patient discomfort and undue stress5. Due to practical delays the fasting period for endoscopy can become much longer than the stipulated six hours, thus causing even more patient discomfort. This may become especially difficult to tolerate for patients with gastroesophageal reflux disease (GERD). Physiological studies have shown that clear fluids leave the stomach rapidly6, and several anesthetic guidelines now recommend a 2-hour preoperative fast for clear fluids and a 6-h fast for solids before general anesthesia2,7. A pilot study done by us on 20 patients using real-time ultrasonography showed the time taken for a clear liquid (plain tea) or water to empty from the stomach was one hour.

> The second issue with the guidelines was that these guidelines focus on a Western type diet1. Rice is the staple diet in many Asian countries which include more than half the worlds population. Rice consumption is also increasing in many western societies8. Normally, the proximal part of the stomach acts as a reservoir for food and the distal end acts as the grinder9. Gastric emptying in normal subjects is complicated process, and is affected by,various meal related factors like the fat content, consistency and the size of the meal10. However, patient dependant factors like age, sex and body mass index (BMI) have been shown to have no significant association with gastric emptying11. In a pilot study using real time ultrasound scanning we found that the time taken for complete gastric emptying after a standard rice meal was 10 hours. To the best of our knowledge there are no published guidelines on endoscopic preparation for a rice based diet. We therefore decided to investigate whether a six hour fast after a rice based meal was sufficient prior to UGIE.

#### **2. Aims**

The aim of the first study was to determine whether a 6-hour fast for solids and a one-hour fast for water prior to UGIE gives good endoscopic vision and less patient discomfort. The aim of the second study was to determine whether a six hour fast after a standard rice based meal is sufficient to achieve good endoscopic vision during UGIE.

Evidence Based Guidelines for Preparation Before Upper Gastrointestinal Endoscopy (UGIE) 17

gastric surgery or clinically obvious motility problems were recruited from September 2007

Patients on drugs known to affect gastrointestinal motility and acid secretion were requested to stop them for two weeks prior to endoscopy. Patients who could not or refused to comply with this request were excluded from the study. Patients were then given a standard rice meal, which contained rice, dhal and an egg. All meals were isocaloric, and the caloric value of meals was calculated assuming a daily requirement of 2500 kcal

**Standard Meal Unit Energy per** 

(Parboiled) White Rice 200g per 100 g 371 742 Egg, whole, cooked, hard-boiled (one) 1 large = 50g 78 78

100g per 100 g 114 114 Total energy provided 934

U.S. Department of Agriculture, Agricultural Research Service (USDA:ARS) 1998 USDA

The patients were then randomized into two groups in preparation for UGIE: fasting for 6 hours after the rice meal (R6) or fasting for 10 hours after the rice meal (R10). All patients were given the meal at 7 am and endoscopies were performed in the afternoon. Endoscopies were done using Olympus GIF 0145 video endoscopes. All endoscopies were performed by two operators (MN and UK). The endoscopists were blinded to the period of fasting, but not to the indication for endoscopy. Another endoscopist (JM), who was also blinded to the period of fasting, was present during all endoscopies and independently graded endosocopic vison. Endoscopic vision was graded as poor, average or good. If there was any discrepancy between the two endoscopist, grading the lower was taken. All patients were

Ethical approval for this study was obtained from the Ethics Committee of the Faculty of Medicine, University of Kelaniya, Sri Lanka. Trial registration number: SLCTR/2008/004.

reviewed in the out-patient clinic one week later to detect any late complications.

**unit (kcal)\***

**Total energy (kcal)**

**Variables R6 R10 P value** 

Male : Female 47:58 56 :51 0.270\* Median age (Range) years 40(18 - 65) 44(18 - 65) 0.082\* GERD Symptoms 82 76 0.238\* Dyspeptic symptoms 20 27 0.278\* Non specic abdominal pain 3 4 0.072\*

Number of patients Randomized 105(49.5%) 107(50.5%)

Table 1. Demographic and endoscopic findings of patients

Lentils, mature seeds, cooked, boiled, with salt

**2.2.1 Ethical clearance and trial registration** 

Table 2. The standard rice based meal

Nutrient Database

to June 2008 (Table 1).

(Table 2).

#### **2.1 Methods (1)**

Consecutive patients referred for UGIE to the Professorial Medical Unit of the Colombo North Teaching Hospital, Ragama, Sri Lanka, between the ages of 18 to 65 years, who were not pregnant, and had no alarm symptoms previous gastric surgery or clinically obvious motility problems were recruited from 2005 September to 2006 August. Patients on drugs known to affect gastrointestinal motility and acid secretion were requested to stop them for two weeks prior to endoscopy. Patients who could not / refused to comply were excluded from the study. All patients gave informed written consent.

All patients were given a standard meal consisting of two slices of bread and jam 6 hours before endoscopy. Patients were then randomized, using a computer generated random allocation table, to either nil by mouth for 6 hours (Group A) or no solids for six hours but allowed to drink water according to thirst for up to one hour prior to endoscopy (Group B). Patients in Group B were requested to keep a record of the amount of water they drank. Just prior to endoscopy patients were requested to indicate any discomfort due to fasting on a visual analog scale (0 - no discomfort, to 10 - severe discomfort). The endoscopist was blinded to the period of fasting but not to the indication for endoscopy. Endoscopies were performed in the afternoon because patients were given the standard meal in the morning. Commencing at 7 am meals were given to consecutive patients at twenty minute intervals, in order to maintain uniform periods of fasting. Before endoscopy, sterile water was aspirated through the suction channel and then drained dry. All patients had throat spray before endoscopy, and pulse-oxymetry during endoscopy. Endoscopies were performed by the same operator. Endoscopic vision was graded as good, average or poor. During endoscopy, any fluid in the gastric fundus was aspirated and the volume and pH of the aspirate were measured. Endoscopies were recorded on video, and two other endoscopists (ASD and HJdeS), independently assessed the adequacy of endoscopic vision. Patients were seen in the out-patient clinic one week later to detect any late complications.

#### **2.1.1 Ethical clearance and trial registration**

Ethical approval for this study was obtained from the Ethics Committee of the Faculty of Medicine, University of Kelaniya, Sri Lanka.

#### **2.1.2 Statistics**

We assumed that 50% of patients awaiting endoscopy will have discomfort due to fasting. By allowing them to drink water up to one hour prior to endoscopy we expected to reduce this to 20%. We calculated that a sample size of 80 was needed to give the study 80% power at α= 0.05. The outcome variables were, endoscopic visibility, pH of gastric content, the number of patients with fluid in the gastric fundus, and the level of patient discomfort due to fasting. Statistical significance of the difference between the two groups was tested using the Mann-Whitney U test for continuous variables and Fisher's Exact Test for categorical variables. Kappa was used to calculate the degree of agreement between the endoscopist and the two independent assessors who viewed videos of the endoscopies. The p values are presented uncorrected for multiple testing.

#### **2.2 Methods (2)**

After informed written consent, consecutive patients referred for UGIE, to the Professorial Medical Unit of the Colombo North Teaching Hospital, Ragama, Sri Lanka, between the ages of 18 to 65 years, who were not pregnant, and had no alarm symptoms, previous


gastric surgery or clinically obvious motility problems were recruited from September 2007 to June 2008 (Table 1).

Table 1. Demographic and endoscopic findings of patients

Patients on drugs known to affect gastrointestinal motility and acid secretion were requested to stop them for two weeks prior to endoscopy. Patients who could not or refused to comply with this request were excluded from the study. Patients were then given a standard rice meal, which contained rice, dhal and an egg. All meals were isocaloric, and the caloric value of meals was calculated assuming a daily requirement of 2500 kcal (Table 2).


Table 2. The standard rice based meal

16 Gastrointestinal Endoscopy

Consecutive patients referred for UGIE to the Professorial Medical Unit of the Colombo North Teaching Hospital, Ragama, Sri Lanka, between the ages of 18 to 65 years, who were not pregnant, and had no alarm symptoms previous gastric surgery or clinically obvious motility problems were recruited from 2005 September to 2006 August. Patients on drugs known to affect gastrointestinal motility and acid secretion were requested to stop them for two weeks prior to endoscopy. Patients who could not / refused to comply were excluded

All patients were given a standard meal consisting of two slices of bread and jam 6 hours before endoscopy. Patients were then randomized, using a computer generated random allocation table, to either nil by mouth for 6 hours (Group A) or no solids for six hours but allowed to drink water according to thirst for up to one hour prior to endoscopy (Group B). Patients in Group B were requested to keep a record of the amount of water they drank. Just prior to endoscopy patients were requested to indicate any discomfort due to fasting on a visual analog scale (0 - no discomfort, to 10 - severe discomfort). The endoscopist was blinded to the period of fasting but not to the indication for endoscopy. Endoscopies were performed in the afternoon because patients were given the standard meal in the morning. Commencing at 7 am meals were given to consecutive patients at twenty minute intervals, in order to maintain uniform periods of fasting. Before endoscopy, sterile water was aspirated through the suction channel and then drained dry. All patients had throat spray before endoscopy, and pulse-oxymetry during endoscopy. Endoscopies were performed by the same operator. Endoscopic vision was graded as good, average or poor. During endoscopy, any fluid in the gastric fundus was aspirated and the volume and pH of the aspirate were measured. Endoscopies were recorded on video, and two other endoscopists (ASD and HJdeS), independently assessed the adequacy of endoscopic vision. Patients were

seen in the out-patient clinic one week later to detect any late complications.

Ethical approval for this study was obtained from the Ethics Committee of the Faculty of

We assumed that 50% of patients awaiting endoscopy will have discomfort due to fasting. By allowing them to drink water up to one hour prior to endoscopy we expected to reduce this to 20%. We calculated that a sample size of 80 was needed to give the study 80% power at α= 0.05. The outcome variables were, endoscopic visibility, pH of gastric content, the number of patients with fluid in the gastric fundus, and the level of patient discomfort due to fasting. Statistical significance of the difference between the two groups was tested using the Mann-Whitney U test for continuous variables and Fisher's Exact Test for categorical variables. Kappa was used to calculate the degree of agreement between the endoscopist and the two independent assessors who viewed videos of the endoscopies. The p values are

After informed written consent, consecutive patients referred for UGIE, to the Professorial Medical Unit of the Colombo North Teaching Hospital, Ragama, Sri Lanka, between the ages of 18 to 65 years, who were not pregnant, and had no alarm symptoms, previous

**2.1.1 Ethical clearance and trial registration** 

Medicine, University of Kelaniya, Sri Lanka.

presented uncorrected for multiple testing.

**2.1.2 Statistics** 

**2.2 Methods (2)** 

from the study. All patients gave informed written consent.

**2.1 Methods (1)** 

U.S. Department of Agriculture, Agricultural Research Service (USDA:ARS) 1998 USDA Nutrient Database

The patients were then randomized into two groups in preparation for UGIE: fasting for 6 hours after the rice meal (R6) or fasting for 10 hours after the rice meal (R10). All patients were given the meal at 7 am and endoscopies were performed in the afternoon. Endoscopies were done using Olympus GIF 0145 video endoscopes. All endoscopies were performed by two operators (MN and UK). The endoscopists were blinded to the period of fasting, but not to the indication for endoscopy. Another endoscopist (JM), who was also blinded to the period of fasting, was present during all endoscopies and independently graded endosocopic vison. Endoscopic vision was graded as poor, average or good. If there was any discrepancy between the two endoscopist, grading the lower was taken. All patients were reviewed in the out-patient clinic one week later to detect any late complications.

#### **2.2.1 Ethical clearance and trial registration**

Ethical approval for this study was obtained from the Ethics Committee of the Faculty of Medicine, University of Kelaniya, Sri Lanka. Trial registration number: SLCTR/2008/004.

Evidence Based Guidelines for Preparation Before Upper Gastrointestinal Endoscopy (UGIE) 19

62 were excluded. 128 were randomized to the two interventions, 65 to nil by mouth (Group A) and 63 to those allowed to drink water according to their thirst for up to one hour prior to endoscopy (Group B). The two groups were comparable for age and gender

32 patients (12 in group A and 20 in group B) did not complete the study; 19 had not followed instructions regarding pre-endoscopy preparation, 13 refused endoscopy after randomization (Figure 1). 96 patients completed the study; 53 in group and 43 in group B. All patients in group B had consumed at least 200 ml water (range 200 to 410) before endoscopy. Discomfort due to fasting was significantly lower in group B than in group A. According to the endoscopist, endoscopic vision was good in all 53 patients in group A and 40 patients in group B, and average in 3 patients in group B. None were graded as poor. There was good agreement between the endoscopist and independent assessors who

Fluid in the gastric fundus was noted in 11 patients in group A and 16 in group B. There were no significant differences in volume or pH of the gastric aspirate between the two

**Variables Group A Group B P value** 

Male : Female 32:33 31:32 0.9\* Median age (Range) years 48 (18-65) 49 (20-64) 0.9\*\*\*

GERD Symptoms 35 33 0.88\* Dyspeptic symptoms 23 18 0.48\* Non specific abdominal pain 7 12 0.33\*

Male: Female 29:24 24:19 0.42\* Median age (Range) years 49(19-65) 48(20-64) 0.91\*\*

for discomfort 9.7(9.6-9.8) 5.6(4.3-8.5) <0.0001\*\*

Number of Patients with fluid in the gastric fundus (%) 11(21%) 16(37%) 0.12\* Volume of gastric aspirate (ml) [Median (Range)] 8(0-65) 10 (0-78) 0.56\*\* pH of gastric aspirate[Median(Range)] 1.3(0.9-6.3) 2.9(0.9-5) 0.99\*\*

GERD 18 (34%) 24 (56%) 0.04\* Gastritis 17 (32%) 6 (14%) 0.054\* Esophageal / gastric varices 6 (11%) 1 (2%) 0.13\*

Normal 10 (19%) 12 (28%) 0.33\*

Average 0

Average 3

Average 1

Good 40 Average 3

Good 40

Good 41 Average 2

Average 3 0.00 [1.1-19]\*\*\*

groups. There were no complications attributable to the endoscopy in either group.

Number of patients randomized 65 63

Number of patients completed trial 53 43

Endoscopist Good 53

Independent Assessor 1 Good 50

Independent Assessor 2 Good 52

Gastric ulcer 2 (4%)

Table 3. Demography and indication for endoscopy in patients

viewed the videos of the endoscopies (Kappa= 0.64) (Table 3).

(Table 3).

Indications of Endoscopy

Endoscopic vision

Visual analog score [Median (IQR)]

Principal endoscopic diagnosis (%)
