Prevalence of Helicobacter pylori in patients with gastro-oesophageal reflux disease: systematic review -- Raghunath et al. 326 (7392): 737 Data Supplement

Details of searches for studies of Helicobacter pylori prevalence in gastro-oesophageal reflux disease

Table A Studies excluded from systematic review

Table B Prevalence of Helicobacter pylori in patients with gastro-oesphageal reflux disease and those in the comparator group

Details of searches for studies of Helicobacter pylori prevalence in gastro-oesophageal reflux disease

Medline, Embase, Cinahl, and Cochrane (controlled trials register and database of systematic reviews) were searched using broad search strategies to identify all studies and trials determining the prevalence of Helicobacter pylori in gastro-oesophageal reflux disease. Searches were run from 1983 for Medline and Cinahl and from 1988 for Embase until May 2000. A final search of Medline and Embase was undertaken in November 2001.

ASR handsearched Gut, Gastroenterology, British Medical Journal, Lancet, New England Journal of Medicine and Alimentary Pharmacology and Therapeutics from 1998. In addition, APSH and AR routinely reviewed the content of major gastroenterological and general medical journals for the year up until the end of October 2001. AR also contacted members of the Cochrane Upper Digestive and Pancreatic Group, editors of Alimentary Pharmacology and Therapeutics and Gut as well as experts in the specialty of H pylori and gastro-oesophageal reflux disease.

We also reviewed the bibliographies of retrieved papers for relevant studies not identified by the database and handsearching. Astra Zeneca, Wyeth, and Abbott were also contacted for any data on studies that had been published or were unpublished and in their archives.

Terms for prevalence of H pylori in gastro-oesophageal reflux disease

MeSH search terms

Gastro-oesophageal reflux disease and H pylori related

helicobacter pylori

gastroesophageal reflux

heartburn

esophagitis

esophageal stenosis

barrett esophagus

esophageal neoplasms

Embase subject headings

campylobacter pyloridis

barrett esophagus

esophagus cancer

esophagus carcinoma

esophagus metastasis

esophagus tumor

reflux esophagitis

gastroesophageal reflux

esophagus stricture

heartburn

Textword search terms

helicobacter pylori

Campylobacter pyloridis

Campylobacter pylori

reflux

gastroesophageal

gastro-oesophageal

gastro AND oesophageal

gastro AND esophageal

GERD

gastro-oesophageal reflux disease

heartburn

esophagitis OR oesophagitis

stricture

esophageal OR oesophageal

barrett

esophagus OR oesophagus

neoplasm OR neoplasms

cancer OR cancers

Selection criteria

Subject

Any relation or association between H pylori and oesophageal disease—that is, Barrett’s oesophagus, oesophageal cancer, reflux oesophagitis, gastro-oesophageal reflux disease. Such a relation could be mediated through a "third party,"—for example, proton pump inhibitor treatment.

Study type

Systematic review

Meta-analysis

Randomised controlled trial

Any other type of clinical trial

Case controlled study

Cohort study (retrospective or prospective)

Cost analysis study based on the above types of study

Biomedical or biological study (human, animal, biochemical, genetic, etc.)

Qualitative research study

We discarded the following study types or publication types:

Case report

Personal literature review

Expert opinion

Consensus report (unless based on selected types of study)

Editorials

Letters

Selection has been an iterative process. Database records were reviewed independently by both ASR and SC. Disagreements were discussed and a consensus reached. Complete articles were obtained for items that passed the selection criteria or for items where there was not enough information in the database record to make a decision. RG then reviewed the full articles against the selection criteria for a second time.

Experts contacted for systematic review

Dr H H Tsai (United Kingdom), Professor P Malfertheiner (Germany), Dr A G Fraser (New Zealand) Professor J Labenz (Germany), Dr N Murai (Japan), Dr N Vakil (United States), Professor K Haruma (Japan), Professor B Tepes (Slovenia).

Table A Studies excluded from systematic review

References	Type of study	Participants	Outcomes and results	Reasons for exclusion
Schenk et al 1999⁷	Cohort, prospective	Cases: >grade 1 oesophagits, Barrett’s oesophagus, and hiatus hernia	H pylori prevalence in cases 39/88 (44%). No separate prevalence data on different categories	No control group
Cheng et al 1989⁹	Descriptive	Cases: patients undergoing paired biopsies of distal oesophagus and gastric antrum during endoscopy	H pylori prevalence in cases 11/27 (41%)	No control or comparator group
McCallum et al 1992¹³[abstract]	Case-control	Cases: reflux disease (all positive on Bernstein’s test). Controls: asymptomatic, healthy, volunteers	H pylori prevalence in cases 13/21 (60%) and in controls 1/20 (5%)	Control group did not meet eligibility and quality criteria. Wide difference in mean age between cases and controls (50:30). Numbers too small in each group
Abbas et al 1995²¹	Case-control, retrospective	Cases: uncomplicated oesophagitis. Controls: Barrett’s oesphagus	H pylori prevalence in cases 18/29 (62%) and in comparator 14/29 (48%)	Comparator group inappropriate for this systematic review
Oberg et al 1999²²	Descriptive and retrospective	Cases: oesophagitis, Barrett’s oesophagus, and negative reflux disease on endoscopy	H pylori prevalence in cases 27/189 (14%). No separate prevalence data on different categories	No control group
Sekiguchi et al 1996²³	Descriptive	Cases: oesophagitis, grades 1-4	H pylori prevalence in cases 6/21 (29%). In grades 1 and 2 prevalence was 6/12 (50%) and in grades 3 and 4 it was 0/9 (0%)	No control group. Presence of reflux disease negative on endoscopy in controls could not be excluded
Macchiarelli et al 1998²⁴	Descriptive, retrospective, prevalence	Retrospective preselection of cases presenting with typical symptoms of gastro-oesophageal reflux disease who had undergone both endoscopy and 24 hour pH studies	H pylori prevalence in patients divided into reflux (12/20, 60%) and non-reflux (6/23, 26%) groups based on abnormal and normal ph-metry results	Number of cases and comparator too small. Peptic ulcer exclusion not stated
Velanovich 1995²⁵	Prospective, cohort	Cases: oesophagitis and oesophageal ulcers. Controls: normal endoscopy, some had heartburn as symptoms	H pylori prevalence in cases 37% and in controls 17%. No data on number of cases and controls	Control group did not meet eligibility and quality criteria. Presence of reflux disease negative on endoscopy in controls could not be excluded. No numerical data on actual number of cases and controls with gastro-oesophageal reflux disease
Yerra et al 1999²⁶	Case-control	Cases: all grades of oesophagitis. Controls: non-ulcer dyspepsia	H pylori prevalence in cases 20/30 (66%) and controls 19/30 (63%)	Control group did not meet eligibility and quality criteria
O’Connor and Cunnane 1994²⁷	Cohort, prospective	Cases: oesophagitis, Barrett’s oesophagus, and hiatus hernia	Overall H pylori prevalence (50/93, 54%). Age-cohort effect, with maximum prevalence in sixth decade, then tapering off. H pylori prevalence in patients with oesophagitis grade 1, 22/42 (52%), grade 2, 13/25 (52%), grade 3, 6/11 (54%), and grade 4, 8/15 (53%)	No control group
Kuipers et al 1993²⁸	Longitudinal, cohort	Cases: severe oesophagitis	H pylori prevalence in cases 26/51 (51%)	No control group
Ho and Kang 1999²⁹	Prospective, cohort	Cases: erosive oesophagitis and negative reflux disease on endoscopy. Controls: reflux symptoms, normal endoscopy	H pylori prevalence in cases 9/22 (43%) and controls 17/30, (53%). No separate prevalence data for categories	Control group did not meet eligibility and quality criteria. Presence of reflux disease negative on endoscopy in controls could not be excluded
Bate et al 1998³⁰	Cohort, prospective (eradication study)	Case: all grades (0-4) of oesophagitis. Negative reflux disease on endoscopy based on symptoms and normal endoscopy	H pylori prevalence in cases 62/153 (34%). No individual prevalence data for patients with reflux disease negative on endoscopy and oesophagitis provided	No control group
Berstad et al 1997³¹	Longitudinal, prospective cohort	Cases: preselected grades 1 and 2 oesophagitis	H pylori prevalence in cases 44/90 (49%)	No control group
Cooper and Gearty 1991³²	Case-control, retrospective	Cases: Barrett’s oesophagus, Controls: oesophagitis	H pylori prevalence in patients with Barrett’s oesophagus 14/26 (54%) and oesophagitis 11/30 (37%)	Control group inappropriate for this systematic review
Warburton-Timms et al 2001³³	Descriptive, retrospective, prevalence	Unselected cohort of patients attending for routine endoscopy in 1986	Overall H pylori prevalence 663/1485 (45%). H pylori prevalence in patients with oesophagitis 120/312 (38%) and without oesophagitis 543/1172 (46%)	No control or preselected comparator group. No data on symptoms. Gastro-oesophageal reflux disease cannot be excluded with certainty in patients without oesophagitis from available information
Koike et al 2001³⁴	Case-control, prospective	Patients were self referred and referred by doctor. Cases: patients with reflux oesophagitis. Controls: age-sex matched, randomly selected, who visited the hospital, were asymptomatic, and had normal endoscopy result	H pylori prevalence in cases 36/105 (34%) and controls 80/105 (72%)	Large overlap of patients between this study and study published in 1999.

Table B Prevalence of Helicobacter pylori in patients with gastro-oesphageal reflux disease and those in the comparator group

First author (country, year) reference	Patients in comparator group	Patients with gastro-oesophageal reflux disease	Odds ratio (95% CI)
Csendes (Chile, 1997)³⁶	38/190 (20)	40/136 (29)	1.67 (1.00 to 2.78)
Newton (United Kingdom, 1997)⁴⁶	9/25 (36)	15/36 (42)	1.27 (0.44 to 3.63)
Pieramico (Italy, 2000)⁴⁷	19/49 (39)	24/54 (44)	1.26 (0.58 to 2.77)
Gisbert (Spain, 2001)³⁹	23/44 (52)	32/56 (57)	1.22 (0.55 to 2.69)
Hackelsberger (Germany, 1998)⁴¹	89/227 (39)	50/130 (38)	0.97 (0.62 to 1.51)
Manes (Italy, 1999)⁴⁴	80/200 (40)	37/105 (35)	0.82 (0.50 to 1.33)
Vaezi (United States, 2000)⁵⁰	25/60 (42)	39/108 (36)	0.79 (0.41 to 1.51)
El-Serag (United States, 1999)³⁷	55/148 (37)	36/116 (31)	0.76 (0.45 to 1.27)
Goldblum (United States, 1998)⁴⁰	13/27 (48)	24/58 (41)	0.76 (0.30 to 1.90)
Varanasi (United States, 1998)⁵¹	89/257 (35)	24/86 (28)	0.73 (0.43 to 1.25)
Liston (United Kingdom, 1996)⁴³	27/33 (82)	28/37 (76)	0.69 (0.22 to 2.21)
Vicari (United States, 1998)⁵²	26/57 (46)	30/84 (36)	0.66 (0.33 to 1.32)
Schubert (United States, 1999)⁴⁸	17/42 (40)	9/31 (29)	0.60 (0.22 to 1.62)
Fallone (Canada, 2000)³⁸	37/78 (47)	27/81 (33)	0.55 (0.29 to 1.05)
Werdmuller (Holland, 1997)¹⁰	204/399 (51)	34/118 (29)	0.39 (0.25 to 0.60)
Shirota (Japan, 1999)⁴⁹	17/28 (61)	26/73 (36)	0.36 (0.15 to 0.88)
Wu (Hong Kong, 1999)⁵³	73/120 (61)	21/66 (32)	0.30 (0.16 to 0.57)
Mihara (Japan, 1996)⁴⁵	47/70 (67)	26/70 (37)	0.29 (0.14 to 0.58)
Haruma (Japan, 2000)⁴²	145/190 (76)	39/95 (41)	0.22 (0.13 to 0.37)
Koike (Japan, 1999)³⁵	126/175 (72)	59/175 (34)	0.20 (0.13 to 0.31)