This was a prospective single-center an open‑label parallel randomized controlled trial. This study was performed in the department of gastroenterology at Damascus hospital, Syria. We recruited appropriate candidates from patients who visited our clinic for evaluation of dyspeptic symptoms by upper gastrointestinal endoscopy between February 2020 to August 2020. Exclusion criteria were (1) younger than 18 years and older than 80 years, (2) prior eradication treatment for H pylori, (3) documented reactions to any of the studied medications ,(4) recent use of antibiotics, bismuth, or proton pump inhibitors (PPIs) in the preceding month, (5) pregnant or lactating women, (6) previous gastric surgery, (7) alcohol or opioid abuse, or (8) severe concurrent medical illnesses, such as liver failure, renal failure, or terminal malignancy.
H. pylori Detection
All patients have undergone upper gastrointestinal endoscopy. Endoscopists have taken five gastric biopsies, two from the antrum, two from the body, and one from the incisura according to the Sydney system.[20] Pathologists confirm H. Pylori infection by microscopic examination after using haematoxylin, eosin, and Giemsa stains.[21] We sent all biopsies to the central pathology laboratory of the same referral hospital.
Intervention
Eligible patients were randomized in a 1:1 ratio to receive two weeks of treatment of either bismuth quadruple D-group or concomitant levofloxacin L-group. The D-group obtained bismuth subsalicylate 524 mg q.i.d, doxycycline 100 mg, tinidazole 500 mg, esomeprazole 20 mg each b.i.d for 14 days. While The L-group obtained levofloxacin 500 mg q.d, tinidazole 500 mg, amoxicillin 1000 mg, and esomeprazole 20 mg each b.i.d for 14 days. The indication of treatment relied on the American College of gastroenterology guideline and Maastricht V/Florence consensus report [13, 15] including peptic ulcer, chronic gastritis, primary gastric MALT lymphoma, intestinal metaplasia, dyspepsia, and unexplained iron deficiency anaemia.
We used a Microsoft Excel function called (RANDBETWEEN) to generate a sequence of two randomized numbers, number one refer to the D-group, and number two refers to the L-group. We print each code on separate paper and insert it into sealed opaque envelopes in unchanged order, and hold it in a secure locker belonging to an independent medical staff member. After obtaining informed consent, the independent medical staff member took the top envelope to assign the patient to the treatment regimen. We provide written instruction to all patients on how to take the medications. We evaluated compliance by counting the number of unused medications and considered that the patient had a good complaint if he/she had taken at least 90% of the assigned treatment protocol.
At the end of the treatment course, patients revisited the clinic to investigate side effects and evaluate compliance. We reported side effects such as nausea, vomiting, diarrhoea, melena, dysgeusia, and anorexia. After eight weeks, all patients visited the central laboratory of our hospital and performed stool antigen tests by using the enzyme immunoassay method (EIA).[22] Medical laboratory workers were blinded to the treatment arm. a qualified physician collected the data in a questionnaire including (1) participants’ demographics, (2) smoking history, (3) medication history, (4) adverse events, and (5) results of stool antigen test. Numerical data were shown as mean, and qualitative data were expressed as a ratio.
Authors report the results of this research according to the CONSORT.
Outcomes:
The primary outcome was the H pylori eradication rates of the initially assigned treatment according to intention to treat analysis (ITT), and per-protocol analysis (PPA).
Sample size and statistical analysis:
Federico et al. Found that the eradication rate based on ITT was 0.922 % in concomitant levofloxacin-containing therapy.[23] While Wang et al. found that the eradication rate based on IIT was 0.6744 in doxycycline-based quadruple.[24] We used a power (1-β) of 80 %, two tails test and Significance level(α) equal to 5%, with a 1:1 allocation ratio. Each treatment arm requires thirty-seven patients in each group. [25, 26] We added two patients to each group to compensate for the predicted dropout. [27, 28]
Statistical tests to estimate the potential differences in outcomes between the two groups were chi-square test (χ2 -test) for categorical variables, and t-test for continuous data. We reported the odds ratio with a 95% confidence interval. A P-value of less than 0.05 was considered statistically significant. We performed statistical analyses using SPSS (IBM Corp. Released in 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp).