Comparative risk of the incident cancer between histamine-2 receptor antagonists (Risk of cancer between H2RAs)

Study type

• Study type:
  Non-interventional study
• Scope of the study:
  Effectiveness study (incl. comparative)

Non-interventional study

• Non-interventional study design:
  Cohort

Study drug and medical condition

• Anatomical Therapeutic Chemical (ATC) code:
  (A02BA) H2-receptor antagonists
• Medical condition to be studied:
  Neoplasm malignant
• Additional medical condition(s):
  The primary outcome was overall cancer except non-melanoma skin cancer. The secondary outcomes were overall cancer, 16 subtypes of cancer.

Population studied

• Age groups:
  Adults (18 to < 46 years)

  Adults (46 to < 65 years)

  Adults (65 to < 75 years)

  Adults (75 to < 85 years)

  Adults (85 years and over)
• Estimated number of subjects:
  200000

Study design details

• Main study objective:
  To generate evidence for comparative safety of incident cancer of ranitidine compared with other H2 blockers
• Outcomes:
  Occurrence of malignant neoplasm except non-melanoma skin cancer for the first time in the person’s history, The secondary outcomes include overall cancer, 16 subtypes of cancer (lip, oral cavity and pharynx cancer, esophageal cancer, stomach cancer, colorectal cancer, liver cancer, pancreatic cancer, lung cancer, breast cancer, cervical cancer, uterine cancer, ovary cancer, prostate cancer, bladder cancer, leukemia, thyroid cancer, gall bladder and biliary tract cancer) and death
• Data analysis plan:
  We use propensity score model to reduce potential confounding due to imbalance between the target and comparator cohorts in baseline covariates. The following covariates are used in the large-scale propensity score matching: demographics including age, gender and race, all recorded medication, medical history, exposed procedures, Charlson comorbidity index in the year prior to the index date in each database. We construct matched cohorts using 1:1 propensity score matching with a caliper of 0.2 on the logit scale. The propensity scores were estimated by L1 regularized logistic regression, tuned by 10-fold cross validation. Cox proportional hazard models will be used to assess the hazard ratios with associated 95% confidence intervals (CIs) between the two cohorts using the CohortMethod R package (https://github.com/OHDSI/CohortMethod). Random-effect model meta-analysis will be performed to calculate summary hazard ratio for pooling effect estimates across databases.