Systematic Review

Duodenogastroesophageal (Biliary) Reflux Prevalence, Symptoms, Oesophageal Lesions and Treatment

Chamara Basnayake; Annelies Geeraerts; Ans Pauwels; Ger Koek; Michael Vaezi; Tim Vanuytsel; Jan Tack

Aliment Pharmacol Ther. 2021;54(6):755-778.

Abstract and Introduction

Abstract

Background: The prevalence of duodenogastroesophageal reflux (DGER) and its effect on symptoms and oesophageal lesions in gastroesophageal reflux disease (GERD) is unclear.

Aims: To conduct a systematic review to determine the prevalence of DGER among patients with GERD, the effect of DGER on symptoms and oesophageal lesions, and the treatment of DGER.

Methods: We searched Pubmed and MEDLINE for full text, English language articles until October 2020 that evaluated DGER prevalence among patients with GERD, the effect of DGER on symptoms and oesophageal lesions, and the treatment of DGER.

Results: We identified 3891 reports and included 35 which analysed DGER prevalence in GERD, 15 which evaluated its effect in non-erosive reflux disease (NERD), 17 on erosive oesophagitis, 23 in Barrett's, and 13 which evaluated the treatment of DGER. The prevalence of DGER, when evaluated by Bilitec, among all GERD patients ranged from 10% to 97%, in NERD 10%-63%, in erosive oesophagitis 22%-80% and in Barrett's 50%-100%. There were no differences in the presence or degree of DGER among patients who were asymptomatic or symptomatic on proton pump inhibitors (PPI). The most commonly evaluated treatments for DGER were PPIs and DGER reduced post-PPI therapy in all studies.

Conclusions: The prevalence of DGER increased with more advanced oesophageal lesions and did not explain persisting symptoms among patients taking PPI therapy. PPIs appear to be effective in the treatment of DGER. DGER remains an important consideration in patients with GERD and future therapies deserve more study.

Introduction

Gastroesophageal reflux disease (GERD) is defined by refluxed gastric content into the oesophagus causing symptoms and/or oesophageal lesions. Oesophageal lesions range from erosive oitis to Barrett's ous and many patients with typical GERD symptoms have a macroscopically normal appearance of the oesophagus.^[1] Traditionally, oesophageal injury in GERD patients was conceptualised as the result of chemical caustic effects from luminal factors, with the key factors in this paradigm thought to be acid and pepsin, causing direct injury to the epithelium.^[2] Since the advent of effective acid suppressive therapies, such as proton pump inhibitors (PPIs), there has been significant improvement in controlling symptoms and complications which arise from GERD.^[3] However, a substantial proportion of patients continue to suffer from symptoms and complications of GERD despite adequate PPI therapy. The contributing factors to symptoms in patients with GERD are complex, as significant numbers of patients have symptoms in the absence of erosive changes or significant acid reflux detected on 24 hours pH monitoring.^[1] Furthermore, oesophageal lesions can develop, despite an absence of gastric acid, such as in achlorhydria^[4] or after gastrectomy^[5] which counters the idea that acid is the sole mediator of oesophageal injury in GERD. In this context other injurious and symptom inducing substances in the oesophagus, besides acid, should be considered.

Duodenogastroesophageal reflux (DGER), commonly referred to as "bile reflux", is the pathophysiological entity which is the reflux of duodenal content, including bile, by way of the stomach into the oesophagus. Duodeno-gastric reflux is a normal physiological phenomenon which occurs after meals in both health and disease.^[6] GERD involves the reflux of gastric content which can include bile and other duodeno-gastric refluxate constituents.^[6] Various methods have been utilised to establish the presence of duodenal contents, such as bile acids in the oesophagus, to support the presence of DGER. Early studies demonstrated the presence of bile, trypsin and alkaline substances by aspiration,^[7,8] endoscopic biopsies^[9,10] and scintigraphy.^[11] Several studies, both in vivo and in vitro, have implicated duodenal contents in the pathogenesis of oesophageal GERD lesions. More contemporary studies have replaced aspiration measurements with ambulatory fiberoptic spectrophotometric bilirubin readings, for example, Bilitec 2000 (Medtronic), which quantify DGER by measuring bilirubin. Bilirubin, a constituent of bile, has a characteristic absorption wavelength of 450 mm which is measured by the Bilitec probe. Validation studies have shown that fiberoptic readings with the Bilitec system correlate well with aspiration measures of bilirubin.^[12]

Since the widespread availability of oesophageal pH-impedance monitoring, the use of the Bilitec 2000® has almost completely disappeared. The underlying assumption is often that impedance monitoring accounts for "non-acid" reflux detection. However, this is not an equivalent of "bile reflux" as ambulatory pH and Bilitec 2000 studies in the absence of acid suppressive therapy have shown that DGER usually occurs at an acidic pH.^[13–15] Furthermore, recent trials^[16] which have directly attempted to address bile as a constituent of GERD, have demonstrated improved symptoms and lesions in patients with GERD. Despite this knowledge, there has been a little systematic examination of the literature regarding the prevalence of DGER, its effect on symptoms and oesophageal lesions and which treatments are effective. In this systematic review, we aimed to evaluate the prevalence of DGER, the effect of DGER on symptoms and oesophageal lesions. We also aimed to evaluate the treatment of DGER.

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table 1. Prevalence of DGER among patients with symptoms of GERD
Date of study	Author	Country	Method of assessing DGER	Definition of pathologic DGER	Patient population/subgroups	Number of participants	Prevalence of bile reflux	Degree of bile reflux
1988	Johnsson et al⁸	Sweden	Oesophageal bile aspiration	Not predefined	HV vs GERD	Total participants: 28 10 HV vs 18 patients	Total DGER among patients: 94% (17/18)	Not presented
1993	Iftikhar et al⁷⁰	UK	Oesophageal bile aspiration	Total bile acid concentration, >1105 μmol/L (95% centile upper limit, healthy controls)	HV vs GERD vs Barrett's	Total participants: 60 15 HV vs 30 Barrett's vs 15 GERD	Erosive GERD 20% (3/15) Barrett's 50% (15/30)	Erosive GERD: median 817.3 μmol/L (range 123.8–5721) Barrett's: median 1351.7 μmol/L (range 0–14 011)
1994	Stein et al⁷¹	United States	Oesophageal bile aspiration	Total bile acid concentration >88 μmol/L (>95 centile for healthy controls)	HV vs GERD	Total participants: 95 43 HV vs 52 GERD	Not presented	GERD: Supine mean 107.7 μmol/L ± 24.6 SEM
1994	Champion et al³⁰	United States	Bilitec 2000	% time absorbance >0.14	HV vs GERD vs Barrett's	Total participants: 45 (35 applicable for this review) 12 HV vs 13 GERD vs 10 Barrett's (10 partial gastrectomy not reviewed)	Not presented as total number	Time absorbance was pathological: GERD absorbance mean ~12% Barrett's absorbance mean ~42%
1995	Caldwell et al⁷²	Ireland	Bilitec 2000	% time absorbance >0.15	HV vs GERD vs Barrett's	Total participants: 36 11 HV vs 13 GERD vs 12 Barrett's	Not presented	Time absorbance was pathological: GERD supine median 2.4% (range 0–37.9) Barrett's supine median % 12.9 (range 0–92.5)
1996	Vaezi et al²⁸	United States	Bilitec 2000	% time absorbance >0.14, abnormal: >1.8% time >0.14 (95th centile of healthy controls)	HV vs GERD vs Barrett's	Total participants: 70 20 HV vs 16 NERD vs 14 erosive oesophagitis vs 20 Barrett's	NERD 63% (10/16) Erosive oesophagitis 79% (11/14) Barrett's 95% (19/20)	Time absorbance was pathological: NERD median 3.2% (range 0%-37%) Erosive oesophagitis median 14.6% (range 0–44) Uncomplicated Barrett's 23% (range 0–44) Complicated Barrett's median 46% (range 21–78)
1997	Marshall et al⁴⁸	UK	Bilitec 2000	% time absorbance >0.14, or 2 min either side of symptom event having >10 s of absorbance >0.14	GERD split in two groups: With acid reflux vs no acid reflux	Total participants: 59	Not presented	Time absorbance was pathological: With acid reflux median 13.25% (IQR 5.3–26.5) Non acid reflux patients median 0.1% (IQR 0–1.7)
1997	Kauer et al⁷³	Germany	Oesophageal bile aspiration	Pathological bile reflux was defined by the 95th centile cut-off from their healthy population	HV vs GERD	Total participants: 80 43 HV vs 37 GERD	Trace bile detected in GERD patients 86% (32/37)	Not presented
1998	Marshall et al⁷⁴	UK	Bilitec 2000	% of time absorbance >0.14	HV vs GERD vs Barrett's	Total participants: 128 15 HV vs 63 NERD vs 23 erosive oesophagitis vs 27 Barrett's	Not presented	Time absorbance was pathological: NERD median 2.6% (IQR 0–8) Erosive oesophagitis median 9.1% (IQR 0.7–37.5) Barrett's median 16% (IQR 5–38.2)
1998	Stein et al²⁹	Germany	Bilitec 2000	% of time absorbance >0.25, abnormal >6.9% of time >0.25 absorbance (95th centile cut-off of healthy controls)	HV vs NERD vs erosive oesophagitis vs Barrett's no dysplasia vs Barrett's with adenocarcinoma	Total participants: 131 20 HV vs 19 NERD vs 45 erosive oesophagitis vs 33 Barrett's without dysplasia vs 12 Barrett's with early adenocarcinoma vs 2 Barrett's with High grade dysplasia	NERD 10.5% (2/19) Erosive oesophagitis 22% (10/45) Barrett's 62% (29/47)	Not presented
1998	Oberg et al¹⁹	United States	Bilitec 2000	% time absorbance >0.2 Abnormal: >1.7% time absorbance >0.2	NERD vs erosive oesophagitis vs Barrett's (short segment vs long segment)	Total participants: 262 130 NERD vs 70 erosive oesophagitis vs 30 short segment Barrett's, 32 long segment Barrett's	NERD 30% (39/130) Erosive oesophagitis 61.4% (43/70) Barrett's 79% (49/62)	NERD median 0.1% (IQR 0–2.6) Erosive oesophagitis median 4.2% (IQR 0.2–21.7) Short segment Barrett's median 7.9% (IQR 1.2–18.4) Long segment Barrett's median 15.7% (IQR 4.2–30.5)
1999	Nehra et al⁷⁵	UK	Oesophageal bile aspiration	Not predefined	HV vs GERD & Barrett's	Total participants: 40 10 HV vs 10 NERD vs 10 erosive oesophagitis vs 10 Barrett's	Not presented	NERD median peak bile acids 14 mmol/L (range 0–257) Erosive oesophagitis peak bile acid 124 mmol/L (range 6–1020) Barrett's median peak bile acids 181 mmol/L (range 30–820)
2000	Marshall et al⁷⁶	UK	Bilitec 2000	% time absorbance >0.14. Abnormal >7% of time absorbance >0.14	GERD	Total participants: 113	All participants with GERD 51% (58/113) NERD 17% (7/41) Erosive oesophagitis 64% (29/45) Barrett's 81% (22/27)	Time absorbance was pathological: All patients mean (range): Bile reflux only mean 12.2% (range 6.5–20.7) Acid and bile reflux mean 18.1% (range 3.6–61)
2000	Cuomo et al⁷⁷	Belgium	Bilitec 2000	% time absorbance >0.14, abnormal was >9.6% of time >0.14 absorbance	GERD with erosive oesophagitis	Total participants: 84	NERD 38% (17/45) Erosive oesophagitis and Barrett's 46% (18/39)	Time absorbance was pathological: NERD mean 6.1% ± 1.5 SEM Grade I-II mean 8.2% ± 2.6 SEM Grade III-IV and Barrett's mean 16.1% ± 4.2 SEM
2001	Campos et al²⁷	United States	Bilitec 2000	Abnormal absorbance >0.2 for >1.7% of time	GERD patients being evaluated for Barrett's, subset with Bilitec	Total participants: 502 (only 206 completed Bilitec) All patients GERD on pH, 174 of 502 with Barrett's	Not presented	Time absorbance was pathological: GERD: median 3.0% (IQR 0.1–10.7) Short segment Barrett's median 13.3% (IQR 2.8–26.5) Long segment Barrett's median 17.3% (IQR 6.9–34.4)
2001	Menges et al³²	Germany	Bilitec 2000	% of time absorbance >0.2	Erosive oesophagitis vs Barrett's	Total participants: 43 20 erosive oesophagitis vs 23 Barrett's	Not presented	Time absorbance was pathological: Erosive oesophagitis median 12.8% (IQR 1.9–74.9) Barrett's median 34.7% (IQR 0.3–84.6)
2001	Koek et al¹³	Belgium	Bilitec 2000	% time absorbance >0.14, >0.14 for more than 10 s, episodes >5 min >0.10	GERD	Total participants: 72	Not presented	Of 544 symptom episodes 2 min window prior to symptom episode: mean absorbance 0.11 (±0.007) 4 min window mean absorbance 0.13 (±0.007) 2 min window for symptom association probability 7% SAP for DGER with 10% SAP for mixed (Acid & DGER) reflux
2002	Osugi et al⁷⁸	Japan	Bilitec 2000	>% of time absorbance >0.15, Abnormal >5% time >0.15	HV vs GERD	Total participants: 53 10 HV vs 43 GERD	NERD 50% (3/6) Erosive oesophagitis 97% (36/37)	Not presented
2002	Zaninotto et al²⁵	Italy	Bilitec 2000	% time absorbance >0.2	HV vs GERD vs Barrett's	Total participants: 78 12 HV vs 32 GERD vs 17 long segment Barrett's vs 8 short segment Barrett's vs 9 with ultra short segment Barrett's	Not presented	Long segment Barrett's had significantly greater % time absorbance >0.2. Mean ~12.5% (number not provided)
2003	Tack et al²⁰	Belgium	Bilitec 2000	% of time absorbance >0.14 episode was considered >0.14 for greater than 10 s, upper limit of normal from healthy subjects used but not documented	HV vs GERD	Total participants: 251 40 HV vs 211 GERD	NERD • Solid meals ~10%, (of 45 patients) • Liquid meals ~30% (of 54 patients) Erosive oesophagitis • Solid meals ~30% (of 26 patients) • Liquid meals ~44% (of 34 patients)	NERD solid meals: median 7.7% (IQR 3.7–17), NERD liquid meals median 0.8% (IQR: 0–8.9) Erosive oesophagitis solid meals median 15.3% (IQR 4–42.1) Erosive oesophagitis liquid meals 3.8% (IQR 0.5–14.5)
2003	Tack et al⁷⁹	Belgium	Bilitec 2000	% of time absorbance >0.14	Patients with GERD	Total participants: 26	GERD patients 11/21 (52%)	Time absorbance was pathological: Median 5.8% (IQR 0.7%-12.6%)
2004	Cool et al⁸⁰	Belgium	Bilitec 2000	Absorbance >0.14 Pathological defined as previous identified >95th centile of HV population (not presented)	HV vs GERD	Total participants: 366 20 HV vs 346 GERD	53% (183/346)	Not presented
2006	Xu et al²³	China	Bilitec 2000	Abnormal if absorbance >0.14 for >2.53%	GERD	Total participants: 95	NERD 48% (21/44) Erosive oesophagitis 71% (36/51)	Time absorbance was pathological: NERD mean 7.26% ± 11.08 SD Erosive oesophagitis mean 19.05 ± 23.44 SD
2007	Wolfgarten et al²²	Germany	Bilitec 2000	% of time absorbance >0.25, abnormal if >1.1% of time >0.25	HV vs erosive oesophagitis vs Barrett's.	Total participants: 64 19 HV vs 24 Barrett's vs 21 erosive oesophagitis	Erosive oesophagitis 71% (15/21) Barrett's 75% (18/24)	Time absorbance was pathological: Erosive oesophagitis median 3.5% (IQR 0.1–13.5) Barrett's median 7.8% (IQR 1.6–17.8)
2007	Gadel-hak et al⁸¹	Egypt	Bilitec 2000	% time absorbance >0.14, abnormal if >0.14 for >1.9%	HV vs GERD vs Barrett's	Total participants: 91 Patients	63% (57/91) NERD patients 36% (4/11) Erosive oesophagitis 62% (44/71) Barrett's 100% (9/9)	Time absorbance was pathological: In all participants (including Barrett's) mean 8.46% ± 14.05. NERD mean 4.48 ± 8.99 SD Erosive oesophagitis mean 8.18% ± 11.28 SD Barrett's mean 15.48% ± 30.48 SD
2008	Xiong et al²¹	China	Bilitec 2000	Abnormal if absorbance >0.14 >1.9% of time	NERD	Total participants: 82	52% (43/82)	Time absorbance was pathological: Mean 12.37% ± 22.1 SD
2008	Koek et al⁴⁹	Belgium	Bilitec 2000	% time absorbance >0.14, episodes >5 min >0.14	GERD & Barrett's	Total participants: 422 Suspected GERD 422 • Erosive oesophagitis 164 • Barrett's 30	51% (215/422)	Time absorbance was pathological: All patients median 4.9% (IQR 0.8–20.2) NERD median 2.8% (IQR 0.3–12.4) Erosive oesophagitis median 8.8% median (IQR 2–25) Barrett's median 27.4% (IQR 12–4)
2020	De Bortoli et al	Italy	Bilitec 2000	% time absorbance >0.14, >1.8% of time	NERD, with acid exposure >4%, Partially responsive to PPI	Total participants: 42	50% (21/42)	Time absorbance was pathological: Median time bile reflux pathological 9.6% (IQR 8%-13%)

Abbreviations: DGER, duodenogastroesophageal reflux; GERD, gastro-oesophageal reflux disease; HV, healthy volunteers; IQR, inter-quartile range; NERD, non-erosive reflux disease; SAP, symptom association probability.

Table 2. DGER on PPI therapy
Date of study	Author	Country	Method of assessing DGER	Definition of pathologic DGER	Patient population/subgroups	Number of participants	Prevalence of bile reflux	Degree of bile reflux
2004	Tack et al⁴³	Belgium	Bilitec 2000	Absorbance >0.14 for >10 s, % time absorbance >0.14	GERD on PPI	Total participants: 65	65% (42/65)	Time absorbance was pathological: Median 11.9% (IQR 4.4–31)
2007	Pace et al⁴⁴	Italy	Bilitec 2000	Absorbance >0.14 for >10 s, abnormal >0.14 for >3.9%	GERD on PPI	Total participants: 20	45% (9/20)	Not presented
2008	Karamanolis et al⁴⁵	Belgium	Bilitec 2000	Episode was >0.14 absorbance for >10 s	GERD on PPI	Total participants: 347 (184 had Bilitec performed)	62% (114/184)	Not presented
2008	Brillantino et al⁴⁶	Italy	Bilitec 2000	% of time absorbance >0.14, normal <7% of time	GERD on PPI	Total participants: 92	Total 67% (62/92) NERD 36% (9/25) Erosive oesophagitis 80% (36/45) Barrett's 85% (17/20)	Not presented
2009	Monaco et al²⁴	Italy	Bilitec 2000	Absorbance >0.14, normal <7%	GERD & Barrett's on PPI	Total participants: 83 22 NERD 43 erosive oesophagitis 18 Barrett's	NERD 41% (9/22) Erosive oesophagitis 77% (33/43) Barrett's 83% (15/18)	Time absorbance was pathological: NERD mean 9.2% ± 5.2% SD Erosive oesophagitis: Grade A/B mean 10.9% ± 4.6 SD Grade C/D mean 16.3% ± 6.3 SD Barrett's: Short segment mean 15.8% ± 6.7 SD Long segment mean 19.9% ± 6.2 SD
2009	Gasiorowska et al¹⁷	United States	Bilitec 2000	% time absorbance >0.14, abnormal if >0.14 for >1.8% of time	Symptomatic GERD on PPI vs asymptomatic GERD on PPI	Total participants: 47 24 PPI failure vs 23 PPI success	PPI failure 67% (16/24) PPI success 69% (16/23)	Time absorbance was pathological: Mean PPI failure 11.2% ± 16.59 SD, Mean PPI success 8.49% ± 10.46 SD
2011	Hershcovici et al¹⁸	United States	Bilitec 2000	% of time absorbance >0.14 episode was >0.14 for >10 s	Symptomatic GERD on PPI vs asymptomatic GERD on PPI	Total participants: 47 23 PPI failure vs 24 PPI success	Not presented	Time absorbance was pathological: Mean PPI failure 13.4% ± 27 SD Mean PPI success 6.9% ± 19.2 SD

Table 3. Studies which evaluate the presence of DGER among non-erosive oesophageal reflux (NERD)
Date of study	Author	Country	Method of assessing bile reflux	Definition of pathologic bile reflux	Patient population/subgroups	Number of participants	Prevalence of bile reflux	Degree of bile reflux
1996	Vaezi et al²⁸	United States	Bilitec 2000	% time absorbance >0.14, abnormal: >1.8% time >0.14 (95th centile of healthy controls)	HV vs GERD vs Barrett's	Total participants: 70 20 HV vs 16 NERD vs 14 erosive oesophagitis vs 20 Barrett's	NERD 63% (10/16) Erosive oesophagitis 79% (11/14) Barrett's 95% (19/20)	Time absorbance was pathological: NERD median 3.2% (range 0%-37%) Erosive oesophagitis median 14.6% (range 0–44) Uncomplicated Barrett's 23% (range 0–44) Complicated Barrett's median 46% (range 21–78)
1998	Stein et al²⁹	Germany	Bilitec 2000	% of time absorbance >0.25, abnormal >6.9% of time >0.25 absorbance (95th centile cut-off of healthy controls)	HV vs NERD vs erosive oesophagitis vs Barrett's no dysplasia vs Barrett's with adenocarcinoma	Total participants: 131 20 HV vs 19 NERD vs 45 erosive oesophagitis vs 33 Barrett's without dysplasia vs 12 Barrett's with early adenocarcinoma vs 2 Barrett's with high grade dysplasia	NERD 10.5% (2/19) Erosive oesophagitis 22% (10/45) Barrett's 62% (29/47)	Not presented
1999	Nehra et al⁷⁵	UK	Oesophageal bile aspiration	Not predefined	HV vs GERD & Barrett's	Total participants: 40 10 HV vs 10 NERD vs 10 erosive Oesophagitis vs 10 Barrett's	Not presented	NERD median peak bile acids 14 mmol/L (range 0–257) Erosive oesophagitis peak bile acid 124 mmol/L (range 6–1020) Barrett's median peak bile acids 181 mmol/L (range 30–820)
1998	Oberg et al¹⁹	United States	Bilitec 2000	% time absorbance >0.2 Abnormal: >1.7% time absorbance >0.2	NERD vs erosive oesophagitis vs Barrett's (short segment vs long segment)	Total participants: 262 130 NERD vs 70 erosive oesophagitis vs 30 short segment Barrett's, 32 long segment Barrett's	NERD 30% (39/130) Erosive oesophagitis 61.4% (43/70) Barrett's 79% (49/62)	NERD median 0.1% (IQR 0–2.6) Erosive oesophagitis median 4.2% (IQR 0.2–21.7) Short segment Barrett's median 7.9% (IQR 1.2–18.4) Long segment Barrett's mean 15.7% (IQR 4.2–30.5)
2000	Marshall et al⁷⁶	UK	Bilitec 2000	% time absorbance >0.14. Abnormal >7% of time absorbance >0.14	GERD	Total participants: 113	All participants with GERD 51% (58/113) NERD 17% (7/41) Erosive oesophagitis 64% (29/45) Barrett's 81% (22/27)	Time absorbance was pathological: All patients mean (range): Bile reflux only mean 12.2% (range 6.5–20.7) Acid and bile reflux mean 18.1% (range 3.6–61)
2000	Cuomo et al⁷⁷	Belgium	Bilitec 2000	% time absorbance >0.14, abnormal was >9.6% of time >0.14 absorbance	GERD with erosive oesophagitis	Total participants: 84	NERD 38% (17/45) Erosive oesophagitis and Barrett's 46% (18/39)	Time absorbance was pathological: NERD mean 6.1% ± 1.5 SEM Grade I-II mean 8.2% ± 2.6 SEM Grade III-IV and Barrett's mean 16.1% ± 4.2 SEM
2002	Osugi et al⁷⁸	Japan	Bilitec 2000	>% of time absorbance >0.15, abnormal >5% time >0.15	HV vs GERD	Total participants: 53 10 HV vs 43 GERD	NERD 50% (3/6) Erosive oesophagitis 97% (36/37)	Not presented
2003	Tack et al²⁰	Belgium	Bilitec 2000	% of time absorbance >0.14 episode was considered >0.14 for greater than 10 s, upper limit of normal from healthy subjects used but not documented	HV vs GERD	Total participants: 251 40 HV vs 211 GERD	NERD • Solid meals ~10%, (of 45 patients) • Liquid meals ~30% (of 54 patients) Erosive oesophagitis • Solid meals ~30% (of 26 patients) • Liquid meals ~44% (of 34 patients)	Time absorbance was pathological: NERD solid meals: median 7.7% (IQR 3.7–17), NERD liquid meals median 0.8% (IQR: 0–8.9) Erosive oesophagitis solid meals median 15.3% (IQR 4–42.1) Erosive oesophagitis liquid meals 3.8% (IQR 0.5–14.5)
2006	Xu et al²³	China	Bilitec 2000	Abnormal if absorbance >0.14 for >2.53%	GERD	Total participants: 95	NERD 48% (21/44) Erosive oesophagitis 71% (36/51)	Time absorbance was pathological: NERD mean 7.26% ± 11.08 SD Erosive oesophagitis mean 19.05 ± 23.44 SD
2007	Gadel-hak et al⁸¹	Egypt	Bilitec 2000	% time absorbance >0.14, abnormal if >0.14 for >1.9%	HV vs GERD vs Barrett's	Total participants: 91 Patients	63% (57/91) NERD patients 36% (4/11) Erosive Oesophagitis 62% (44/71) Barrett's 100% (9/9)	Time absorbance was pathological: In all participants (including Barrett's) mean 8.46% ± 14.05. NERD mean 4.48 ± 8.99 SD Erosive oesophagitis mean 8.18% ± 11.28 SD Barrett's mean 15.48% ± 30.48 SD
2008	Xiong et al²¹	China	Bilitec 2000	Abnormal if absorbance >0.14 >1.9% of time	NERD only	Total participants: 82	52% (43/82)	Time absorbance was pathological: Mean 12.37% ± 22.1 SD
2008	Brillantino et al⁴⁶	Italy	Bilitec 2000	% of time absorbance >0.14, normal <7% of time	GERD on PPI	Total participants: 92	Total 67% (62/92) NERD 36% (9/25) Erosive oesophagitis 80% (36/45) Barrett's 85% (17/20)	Not presented
2008	Koek et al⁴⁹	Belgium	Bilitec 2000	% time absorbance >0.14, episodes >5 min >0.14	GERD & Barrett's	Total participants: 422 Suspected GERD 422 Erosive oesophagitis 164 Barrett's 30	51% (215/422)	Time absorbance was pathological: All patients median 4.9% (IQR 0.8–20.2) NERD median 2.8% (IQR 0.3–12.4) Erosive oesophagitis median 8.8% median (IQR 2–25) Barrett's median 27.4% (IQR 12–4)
2009	Monaco et al²⁴	Italy	Bilitec 2000	Absorbance >0.14, Normal <7%	GERD & Barrett's on PPI	Total participants: 83 22 NERD 43 erosive oesophagitis 18 Barrett's	NERD 41% (9/22) Erosive oesophagitis 77% (33/43) Barrett's 83% (15/18)	Time absorbance was pathological: NERD mean 9.2% ± 5.2% SD Erosive oesophagitis: Grade A/B mean 10.9% ± 4.6 SD Grade C/D mean 16.3% ± 6.3 SD Barrett's: Short segment mean 15.8% ± 6.7 SD Long segment mean 19.9% ± 6.2 SD
2020	De Bortoli et al	Italy	Bilitec 2000	% time absorbance >0.14, >1.8% of time	NERD, with acid exposure >4%, partially responsive to PPI	Total participants: 42	50% (21/42)	Time absorbance was pathological: Median time bile reflux pathological 9.6% (IQR 8%-13%)

Table 4. Studies which evaluate the presence of DGER among patients with erosive oesophageal reflux
Date of study	Author	Country	Method of assessing bile reflux	Definition of pathologic bile reflux	Patient population/subgroups	Number of participants	Prevalence of bile reflux	Degree of bile reflux
1993	Iftikhar et al⁷⁰	UK	Oesophageal bile aspiration	Total bile acid concentration, >1105 μmol/L (95% centile upper limit, healthy controls)	HV vs GERD vs Barrett's	Total participants: 60 15 HV vs 30 Barrett's vs 15 GERD	Erosive GERD 20% (3/15) Barrett's 50% (15/30)	Erosive GERD: median 817.3 μmol/L (range 123.8–5721) Barrett's: median 1351.7 μmol/L (Range 0–14011)
1996	Vaezi et al²⁸	United States	Bilitec 2000	% time absorbance >0.14, abnormal: >1.8% time >0.14 (95th centile of healthy controls)	HV vs GERD vs Barrett's	Total participants: 70 20 HV vs 16 NERD vs 14 erosive oesophagitis vs 20 Barrett's	NERD 63% (10/16) Erosive oesophagitis 79% (11/14) Barrett's 95% (19/20)	Time absorbance was pathological: NERD median 3.2% (range 0%-37%) Erosive oesophagitis median 14.6% (range 0–44) Uncomplicated Barrett's 23% (range 0–44) Complicated Barrett's median 46% (range 21–78)
1998	Marshall et al⁷⁴	UK	Bilitec 2000	% of time absorbance >0.14	HV vs GERD vs Barrett's	Total participants: 128 15 HV vs 63 NERD vs 23 erosive oesophagitis vs 27 Barrett's	Not presented	Time absorbance was pathological: NERD median 2.6% (IQR 0–8) Erosive oesophagitis median 9.1% (IQR 0.7–37.5) Barrett's median 16% (IQR 5–38.2)
1998	Stein et al²⁹	Germany	Bilitec 2000	% of time absorbance >0.25, abnormal >6.9% of time >0.25 absorbance (95th centile cut-off of healthy controls)	HV vs NERD vs erosive oesophagitis vs Barrett's no dysplasia vs Barrett's with adenocarcinoma	Total participants: 131 20 HV vs 19 NERD vs 45 erosive oesophagitis vs 33 Barrett's without dysplasia vs 12 Barrett's with early adenocarcinoma vs 2 Barrett's with high grade dysplasia	NERD 10.5% (2/19) erosive oesophagitis 22% (10/45) Barrett's 62% (29/47)	Not presented
1998	Oberg et al¹⁹	United States	Bilitec 2000	% time absorbance >0.2 Abnormal: >1.7% time absorbance >0.2	NERD vs erosive oesophagitis vs Barrett's (short segment vs long segment)	Total participants: 262 130 NERD vs 70 erosive oesophagitis vs 30 short segment Barrett's, 32 long segment Barrett's	NERD 30% (39/130) Erosive oesophagitis 61.4% (43/70) Barrett's 79% (49/62)	NERD median 0.1% (IQR 0–2.6) Erosive oesophagitis median 4.2% (IQR 0.2–21.7) Short segment Barrett's median 7.9% (IQR 1.2–18.4) Long segment Barrett's median 15.7% (IQR 4.2–30.5)
1999	Nehra et al⁷⁵	UK	Oesophageal bile aspiration	Not predefined	HV vs GERD & Barrett's	Total participants: 40 10 HV vs 10 NERD vs 10 erosive oesophagitis vs 10 Barrett's	Not presented	NERD median peak bile acids 14 mmol/L (range 0–257) Erosive oesophagitis peak bile acid 124 mmol/L (range 6–1020) Barrett's median peak bile acids 181 mmol/L (range 30–820)
2000	Marshall et al⁷⁶	UK	Bilitec 2000	% time absorbance >0.14. Abnormal >7% of time absorbance >0.14	GERD	Total participants: 113	All participants with GERD 51% (58/113) NERD 17% (7/41) Erosive oesophagitis 64% (29/45) Barrett's 81% (22/27)	Time absorbance was pathological: All patients mean (range): Bile reflux only mean 12.2% (range 6.5–20.7) Acid and bile reflux mean 18.1% (range 3.6–61)
2000	Cuomo et al⁷⁷	Belgium	Bilitec 2000	% time absorbance >0.14, abnormal was >9.6% of time >0.14 absorbance	GERD with erosive oesophagitis	Total participants: 84	NERD 38% (17/45) Erosive oesophagitis and Barrett's 46% (18/39)	Time absorbance was pathological: NERD mean 6.1% ± 1.5 SEM Grade I-II mean 8.2% ± 2.6 SEM Grade III-IV and Barrett's mean 16.1% ± 4.2 SEM
2001	Menges et al³²	Germany	Bilitec 2000	% of time absorbance >0.2	Erosive oesophagitis vs Barrett's	Total participants: 43 20 erosive oesophagitis vs 23 Barrett's	Not presented	Time absorbance was pathological: Erosive oesophagitis median 12.8% (IQR 1.9–74.9) Barrett's median 34.7% (IQR 0.3–84.6)
2002	Osugi et al⁷⁸	Japan	Bilitec 2000	>% of time absorbance >0.15, abnormal >5% time >0.15	HV vs GERD	Total participants: 53 10 HV vs 43 GERD	NERD 50% (3/6) Erosive oesophagitis 97% (36/37)	Not presented
2003	Tack et al²⁰	Belgium	Bilitec 2000	% of time absorbance >0.14 episode was considered >0.14 for greater than 10 s, upper limit of normal from healthy subjects used but not documented	HV vs GERD	Total participants: 251 40 HV vs 211 GERD	NERD • Solid meals ~10%, (of 45 patients) • Liquid meals ~30% (of 54 patients) Erosive oesophagitis • Solid meals ~30% (of 26 patients) • Liquid meals ~44% (of 34 patients)	Time absorbance was pathological: NERD solid meals: median 7.7% (IQR 3.7–17), NERD liquid meals median 0.8% (IQR: 0–8.9) Erosive oesophagitis solid meals median 15.3% (IQR 4–42.1) Erosive oesophagitis liquid meals 3.8% (IQR 0.5–14.5)
2006	Xu et al²³	China	Bilitec 2000	Abnormal if absorbance >0.14 for >2.53%	GERD	Total participants: 95	NERD 48% (21/44) Erosive oesophagitis 71% (36/51)	Time absorbance was pathological: NERD mean 7.26% ± 11.08 SD Erosive oesophagitis mean 19.05 ± 23.44 SD
2007	Wolfgarten et al²²	Germany	Bilitec 2000	% of time absorbance >0.25, abnormal if >1.1% of time >0.25	HV vs erosive oesophagitis vs Barrett's.	Total participants: 64 19 HV vs 24 Barrett's vs 21 erosive oesophagitis	Erosive oesophagitis 71% (15/21) Barrett's 75% (18/24)	Time absorbance was pathological: Erosive oesophagitis median 3.5% (IQR 0.1–13.5) Barrett's median 7.8% (IQR 1.6–17.8)
2007	Gadel-hak et al⁸¹	Egypt	Bilitec 2000	% time absorbance >0.14, abnormal if >0.14 for >1.9%	HV vs GERD vs Barrett's	Total participants: 91 patients	63% (57/91) NERD patients 36% (4/11) Erosive oesophagitis 62% (44/71) Barrett's 100% (9/9)	Time absorbance was pathological: In all participants (including Barrett's) mean 8.46% ± 14.05. NERD mean 4.48 ± 8.99 SD Erosive oesophagitis mean 8.18% ± 11.28 SD Barrett's mean 15.48% ± 30.48 SD
2008	Brillantino et al⁴⁶	Italy	Bilitec 2000	% of time absorbance >0.14, normal <7% of time	GERD on PPI	Total participants: 92	Total 67% (62/92) NERD 36% (9/25) Erosive oesophagitis 80% (36/45) Barrett's 85% (17/20)	Not presented
2008	Koek et al⁴⁹	Belgium	Bilitec 2000	% time absorbance >0.14, episodes >5 min >0.14	GERD & Barrett's	Total participants: 422 Suspected GERD 422 Erosive oesophagitis 164 Barrett's 30	51% (215/422)	Time absorbance was pathological: All patients median 4.9% (IQR 0.8–20.2) NERD median 2.8% (IQR 0.3–12.4) Erosive oesophagitis median 8.8% median (IQR 2–25) Barrett's median 27.4% (IQR 12–4)
2009	Monaco et al²⁴	Italy	Bilitec 2000	Absorbance >0.14, normal <7%	GERD & Barrett's on PPI	Total participants: 83 22 NERD 43 erosive oesophagitis 18 Barrett's	NERD 41% (9/22) Erosive oesophagitis 77% (33/43) Barrett's 83% (15/18)	Time absorbance was pathological: NERD mean 9.2% ± 5.2% SD Erosive oesophagitis: Grade A/B mean 10.9% ± 4.6 SD Grade C/D mean 16.3% ± 6.3 SD Barrett's: Short segment mean 15.8% ± 6.7 SD Long segment mean 19.9% ± 6.2 SD

Table 5. Studies which evaluate the presence of DGER among patients with Barrett's oesophagus
Date of study	Author	Country	Method of assessing bile reflux	Definition of pathologic bile reflux	Patient population/subgroups	Number of participants	Prevalence of bile reflux	Degree of bile reflux
1993	Iftikhar et al⁷⁰	UK	Oesophageal bile aspiration	Total bile acid concentration, >1105 μmol/L (95% centile upper limit, healthy controls)	HV vs GERD vs Barrett's	Total participants: 60 15 HV vs 30 Barrett's vs 15 GERD	Erosive GERD 20% (3/15) Barrett's 50% (15/30)	Erosive GERD: median 817.3 μmol/L (range 123.8–5721) Barrett's: median 1351.7 μmol/L (range 0–14011)
1994	Champion et al³⁰	United States	Bilitec 2000	% time absorbance >0.14	HV vs GERD vs Barrett's	Total participants: 45 (35 applicable for this review) 12 HV vs 13 GERD vs 10 Barrett's (10 partial gastrectomy not reviewed)	Not presented as total number	Time absorbance was pathological: GERD absorbance mean ~12% Barrett's absorbance mean ~42%
1995	Caldwell et al⁷²	Ireland	Bilitec 2000	% time absorbance >0.15	HV vs GERD vs Barrett's	Total participants: 36 11 HV vs 13 GERD vs 12 Barrett's	Not presented	Time absorbance was pathological: GERD supine median 2.4% (range 0–37.9) Barrett's supine median % 12.9 (range 0–92.5)
1995	Vaezi et al¹⁴	United States	Bilitec 2000	% time absorbance >0.14	HV vs Barrett's	Total participants: 32 12 HV vs 20 Barrett's	Not presented	Time absorbance was pathological: Uncomplicated Barrett's mean 25.9% SEM ± 5.9% Complicated Barrett's mean 48% SEM ± 4.8%
1996	Vaezi et al²⁸	United States	Bilitec 2000	% time absorbance >0.14, abnormal: >1.8% time >0.14 (95th centile of healthy controls)	HV vs GERD vs Barrett's	Total participants: 70 20 HV vs 16 NERD vs 14 erosive oesophagitis vs 20 Barrett's	NERD 63% (10/16) Erosive oesophagitis 79% (11/14) Barrett's 95% (19/20)	Time absorbance was pathological: NERD median 3.2% (range 0%-37%) Erosive oesophagitis median 14.6% (range 0–44) Uncomplicated Barrett's 23% (range 0–44) Complicated Barrett's median 46% (range 21–78)
1998	Marshall et al⁷⁴	UK	Bilitec 2000	% of time absorbance >0.14	HV vs GERD vs Barrett's	Total participants: 128 15 HV vs 63 NERD vs 23 erosive oesophagitis vs 27 Barrett's	Not presented	Time absorbance was pathological: NERD median 2.6% (IQR 0–8) Erosive oesophagitis median 9.1% (IQR 0.7–37.5) Barrett's median 16% (IQR 5–38.2)
1998	Stein et al²⁹	Germany	Bilitec 2000	% of time absorbance >0.25, abnormal >6.9% of time >0.25 absorbance (95th centile cut-off of healthy controls)	HV vs NERD vs erosive oesophagitis vs Barrett's no dysplasia vs Barrett's with adenocarcinoma	Total participants: 131 20 HV vs 19 NERD vs 45 erosive oesophagitis vs 33 Barrett's without dysplasia vs 12 Barrett's with early adenocarcinoma vs 2 Barrett's with high grade dysplasia	NERD 10.5% (2/19) Erosive oesophagitis 22% (10/45) Barrett's 62% (29/47)	Not presented
1998	Oberg et al¹⁹	United States	Bilitec 2000	% time absorbance >0.2 Abnormal: >1.7% time absorbance >0.2	NERD vs erosive oesophagitis vs Barrett's (short segment vs long segment)	Total participants: 262 130 NERD vs 70 erosive oesophagitis vs 30 short segment Barrett's, 32 long segment Barrett's	NERD 30% (39/130) Erosive oesophagitis 61.4% (43/70) Barrett's 79% (49/62)	NERD median 0.1% (IQR 0–2.6) Erosive oesophagitis median 4.2% (IQR 0.2–21.7) Short segment Barrett's median 7.9% (IQR 1.2–18.4) Long segment Barrett's mean 15.7% (IQR 4.2–30.5)
1999	Nehra et al⁷⁵	UK	Oesophageal bile aspiration	Not predefined	HV vs GERD & Barrett's	Total participants: 40 10 HV vs 10 NERD vs 10 erosive oesophagitis vs 10 Barrett's	Not presented	NERD median peak bile acids 14 mmol/L (range 0–257) Erosive oesophagitis peak bile acid 124 mmol/L (range 6–1020) Barrett's median peak bile acids 181 mmol/L (range 30–820)
2000	Marshall et al⁷⁶	UK	Bilitec 2000	% time absorbance >0.14. Abnormal >7% of time absorbance >0.14	GERD	Total participants: 113	All participants with GERD 51% (58/113) NERD 17% (7/41) Erosive oesophagitis 64% (29/45) Barrett's 81% (22/27)	Time absorbance was pathological: All patients mean (range): Bile reflux only mean 12.2% (range 6.5–20.7) Acid and bile reflux mean 18.1% (range 3.6–61)
2000	Cuomo et al⁷⁷	Belgium	Bilitec 2000	% time absorbance >0.14, abnormal was >9.6% of time >0.14 absorbance	GERD with erosive oesophagitis	Total participants: 84	NERD 38% (17/45) Erosive oesophagitis and Barrett's 46% (18/39)	Time absorbance was pathological: NERD mean 6.1% ± 1.5 SEM Grade I-II mean 8.2% ± 2.6 SEM Grade III-IV and Barrett's mean 16.1% ± 4.2 SEM
2001	Campos et al²⁷	United States	Bilitec 2000	Abnormal absorbance >0.2 for >1.7% of time	GERD patients being evaluated for Barrett's, subset with Bilitec	Total participants: 502 (only 206 completed Bilitec) All patients GERD on pH, 174 of 502 with Barrett's	Not presented	Time absorbance was pathological: GERD: median 3.0% (IQR 0.1–10.7) Short segment Barrett's median 13.3% (IQR 2.8–26.5) Long segment Barrett's median 17.3% (IQR 6.9–34.4)
2001	Menges et al³²	Germany	Bilitec 2000	% of time absorbance >0.2	Erosive oesophagitis vs Barrett's	Total participants: 43 20 erosive oesophagitis vs 23 Barrett's	Not presented	Time absorbance was pathological: Erosive oesophagitis median 12.8% (IQR 1.9–74.9) Barrett's median 34.7% (IQR 0.3–84.6)
2002	Csendes et al²⁶	Chile	Bilitec 2000	% of time absorbance >0.2 abnormal if >2% of time >0.2	Barrett's and intestinal metaplasia of cardia	Total participants: 582 408 Barrett's vs 174 intestinal metaplasia	Short segment Barrett's 66% (49/74) Long segment Barrett's 68% (39/57)	Time absorbance was pathological: Short segment Barrett's mean 18% (±22 SD) Long segment mean 15% (±18 SD)
2002	Zaninotto et al²⁵	Italy	Bilitec 2000	% time absorbance >0.2	HV vs GERD vs Barrett's	Total participants: 78 12 HV vs 32 GERD vs 17 long segment Barrett's vs 8 short segment Barrett's vs 9 with ultra short segment Barrett's	Not presented	Long segment Barrett's had significantly greater % time absorbance >0.2. Mean ~12.5% (number not provided)
2004	Sarela et al⁸²	UK	Bilitec 2000	Abnormal if absorbance >0.14 for 1.8%	Barrett's on PPI	Total participants: 32 (23 completed Bilitec testing)	43% (11/23)	Time absorbance was pathological: Median 8.7% (IQR 3.9–14.8)
2007	Wolfgarten et al²²	Germany	Bilitec 2000	% of time absorbance >0.25, abnormal if >1.1% of time >0.25	HV vs erosive oesophagitis vs Barrett's.	Total participants: 64 19 HV vs 24 Barrett's vs 21 erosive oesophagitis	Erosive oesophagitis 71% (15/21) Barrett's 75% (18/24)	Time absorbance was pathological: Erosive oesophagitis median 3.5% (IQR 0.1–13.5) Barrett's median 7.8% (IQR 1.6–17.8)
2007	Gadel-hak et al⁸¹	Egypt	Bilitec 2000	% time absorbance >0.14, abnormal if >0.14 for >1.9%	HV vs GERD vs Barrett's	Total participants: 91 patients	63% (57/91) NERD patients 36% (4/11) Erosive oesophagitis 62% (44/71) Barrett's 100% (9/9)	Time absorbance was pathological: In all participants (including Barrett's) mean 8.46% ± 14.05. NERD mean 4.48 ± 8.99 SD Erosive oesophagitis mean 8.18% ± 11.28 SD Barrett's mean 15.48% ± 30.48 SD
2008	Smythe et al³⁴	UK	Sodium Ion Selective Electrode	% of time Na⁺ ions >50 mmol/L	HV vs Barrett's	Total participants: 30	Not stated	%of time >50 mmol/L, OFF PPI upright, 12% HV vs 17% Barrett's, supine 24% HV vs 1% Barrett's
2008	Brillantino et al⁴⁶	Italy	Bilitec 2000	% of time absorbance >0.14, normal <7% of time	GERD on PPI	Total participants: 92	Total 67% (62/92) NERD 36% (9/25) Erosive oesophagitis 80% (36/45) Barrett's 85% (17/20)	Not presented
2008	Koek et al⁴⁹	Belgium	Bilitec 2000	% time absorbance >0.14, episodes >5 min >0.14	GERD & Barrett's	Total participants: 422 Suspected GERD 422 Erosive oesophagitis 164 Barrett's 30	51% (215/422)	Time absorbance was pathological: All patients median 4.9% (IQR 0.8–20.2) NERD median 2.8% (IQR 0.3–12.4) Erosive oesophagitis median 8.8% median (IQR 2–25) Barrett's median 27.4% (IQR 12–4)
2009	Monaco et al²⁴	Italy	Bilitec 2000	Absorbance >0.14, normal <7%	GERD & Barrett's on PPI	Total participants: 83 22 NERD 43 erosive oesophagitis 18 Barrett's	NERD 41% (9/22) Erosive oesophagitis 77% (33/43) Barrett's 83% (15/18)	Time absorbance was pathological: NERD mean 9.2% ± 5.2% SD Erosive oesophagitis: Grade A/B mean 10.9% ± 4.6 SD Grade C/D mean 16.3% ± 6.3 SD Barrett's: Short segment mean 15.8% ± 6.7 SD Long segment mean 19.9% ± 6.2 SD
2015	Yachimski et al³⁷	United States	Bilitec 2000	% of time absorbance >0.14 abnormal >%1.8 time	Uncomplicated Barrett's	Total participants: 29 (23 had Bilitec)	93% Abnormal *number of patients not provided	Time absorbance was pathological: Median 9.7% (IQR 5–22)

Table 6. Treatment of DGER
Author	Year	Country	Study design	Intervention	Population	DGER measured	Outcome
Champion et al³⁰	1994	USA	Observational Intervention Pre and post	Omeprazole 20 mg BD Testing post 2 weeks therapy	GERD & Barrett's	Bilitec 2000	9 patients (3 GERD and 6 Barrett's) Mean % time absorbance >0.14 Pre PPI: 32.8% ± 6.9 SEM Post-PPI: 4.7% ± 1.7% SEM
Smythe et al³⁸	1997	UK	Observational Intervention Pre and post	Cisapride 10 mg QID (7 days post-treatment)	Barrett's Erosive oesophagitis	Sodium ion electrode	67% (8 of 12) Barrett's reduced DGER 50% (2 of 4) erosive oesophagitis reduced DGER
Stein et al²⁹	1998	Germany	Observational Intervention (in subgroup) Pre and post-PPI or pre and post-Nissen fundoplication	15 patients omeprazole 20 mg BD OR 16 Nissen fundoplication (3–12 months post-surgery)	Asymptomatic healthy volunteers GERD GERD erosive Oesophagitis GERD with Barrett's Barrett's with Adenocarcinoma	Bilitec 2000	Mean DGER % time absorbance abnormal Pre PPI 16.2% and post-PPI 8.9% Mean pre surgery DGER 16.6%, mean post-surgery 2.4%
Marshall et al³¹	1998	UK	Observational Intervention Pre and post	Omeprazole 20 mg BD (6–10 weeks)	Barrett's of at least 3 cm length	Bilitec 2000	11 of 23 patients underwent oesophageal bilirubin monitoring Percent time ≥0.14 Pre omeprazole % time DGER present median 28.9% (range 5.1–67.3) Post-omeprazole % time DGER present median 2.4% (range 0–14.9)
Menges et al³²	2001	Germany	Observational Interventional study Pre and post-PPI	Omeprazole or pantoprazole. Doses ranged 20–80 mg. At least 7 days	Two groups 16 erosive oesophagitis 18 Barrett's	Bilitec 2000	Erosive oesophagitis % time absorbance abnormal Pre PPI median 21.5% (range 5.8–74.9%) Post-PPI median 0.9% (range 0%-42%) Barrett's % time absorbance abnormal Pre PPI median 29.8% (range 0.3–84.6) Post-PPI median 0.7% (range 0–41)
Netzer et al³³	2001	Switzerland	Observational Interventional study	Pantoprazole 40 mg daily Pre and 28 days post-PPI	Symptomatic patients with endoscopic erosive oesophagitis	Bilitec 2000	Number of DGER episodes >5 min Pre PPI mean 9.4% ± 1.2 (SEM) Post-PPI 6 ± 1(SEM) P = 0.04
Koek et al⁴¹	2003	Belgium	Pre and post-intervention Single cohort	Baclofen 5 mg to 20 mg TDS with 20 mg BD omeprazole	Patients with persistent GERD symptoms on PPI treatment, with normal pH (<4%) but abnormal DGER at baseline >4.6% ≥0.14 absorbance.	Bilitec 2000	Time absorbance >0.14 Pre Baclofen median 13.8% (IQR 11.8–15.5) Post-Baclofen median 6.1% (IQR 0.8–10.4), P < 0.05 Episodes lasting greater than 5 min Pre Baclofen median 5 (IQR 3–8) Post-Baclofen median 2 (IQR 0.5–4.5), P < 0.05
Smythe et al³⁹	2003	UK	Randomised controlled trial	Cisapride 20 mg BD, in addition to PPI	Barrett's oesophagus (no high grade dysplasia) on standard dose PPI (20 mg PPI)	Sodium ion electrode (>50 mmol/L presence of DGER)	Pre Cisapride, mean DGER assessment and standard error 27% ± 4% Post-Cisapride DGER mean 33% ± 3%. Pre placebo DGER mean 33% ± 5% vs Post-placebo DGER 31% ± 6%. P = NS
Xin et al⁴⁰	2003	China	Observational Interventional study Pre and post	Famotidine 40 mg IV BD.	Critically Ill ventilated patients.	Bilitec 2000	DGER % time absorbance abnormal pre and post. 24 h post-Famotidine. DGER 15.3% pre Famotidine to 0.29% post-Famotidine
Smythe et al³⁴	2008	UK	Observational Interventional study, pre and post	Omeprazole 40 mg daily	Barrett's without high grade dysplasia on 40 mg omeprazole	Sodium ion selective electrode	DGER measured on PPI then 5 days off PPI. Mean DGER upright 19 mmol/L, upright off PPI 61 mmol/L. Supine DGER on PPI 63 mmol/L, supine off PPI 88 mmol/L
Kunsch et al³⁵	2009	Germany	Observational Interventional study, pre and post Uncontrolled	Pantoprazole 80 mg Daily	Patients with uncomplicated GERD with combined pathological acid and DGER.	Bilitec 2000	DGER pre PPI absorbance >0.14 Mean 19.6% (SD ± 13.7) vs post 5.7 (SD ± 7.7), P < 0.05
Kunsch et al³⁶	2012	Germany	Observational Interventional study Uncontrolled	High dose PPI: Pantoprazole 80 mg daily for 8 weeks	GERD with ineffective PPI response (based on symptoms). 47 patients had symptoms AND either abnormal pH study or DGER at baseline. Then high dose PPI given	Bilitec 2000	Patients split into two groups. Symptom responders and non-responders to high dose PPI (% time absorbance abnormal) PPI symptom responders: Pre PPI mean (SD) 22.8% ± 22.8 vs post-PPI 6.6% ± 10.8 (P < 0.05) PPI symptom non responder: Pre PPI mean (SD) 24.5% ± 18.6 post-PPI 22.2% ± 12.7 (P = NS)
Yachimski et al³⁷	2015	USA	Observational interventional study, pre & post, uncontrolled	Baseline 1 week off PPI Rabeprazole 20 mg BD for one month Patients with persisting acid and/or DGER at one month: Rabeprazole 40 mg BD, outcomes assessed again after another one month	Non dysplastic Barrett's oesophagus >3 cm	Bilitec 2000	DGER at baseline: % time absorbance abnormal Median 9.7% (IQR 5–22.2). Abnormal DGER at baseline 93% of population (number not presented) Post 20 mg BD Rabeprazole 18/23 had normalisation of DGER. Post 40 mg BD remaining 4/5 achieved normalisation of DGER

Abbreviations: BD, twice a day; DGER, duodenogastroesophageal reflux; GERD, gastro-oesophageal reflux disease; HV, healthy volunteers; IQR, Inter-quartile range; IV, intravenous; NERD, non-erosive reflux disease; PPI, proton pump inhibitor; QID, 4 times a day; SAP, symptom association probability; SD, standard deviation; SEM, standard error of the mean.