The acute effects of azithromycin use on cardiovascular mortality, as compared with amoxicillin

Study type

• Study topic:
  Human medicinal product
• Study type:
  Non-interventional study
• Scope of the study:
  Assessment of risk minimisation measure implementation or effectiveness
• Data collection methods:
  Secondary use of data

Non-interventional study

• Non-interventional study design:
  Cohort

Study drug and medical condition

• Study drug International non-proprietary name (INN) or common name:
  AZITHROMYCIN

  AMOXICILLIN

Population studied

• Short description of the study population:
  Patients must meet the following inclusion criteria to be eligible for inclusion in the study:; Inclusion Criteria:; 1. Dispensation of an outpatient prescription for azithromycin or amoxicillin between 01 Jan 1998 and 31 Dec 2012. If a patient had more than one prescription within this period, each exposure will be counted separately (thus, individuals may contribute multiple prescriptions to the analysis).; 2. Consistent with the methodology of Ray and Svanstrom, only oral prescriptions will be included (not intravenous or ophthalmic) and amoxicillin-clavulanate prescriptions will also be included in the amoxicillin group; Exclusion criteria; Patients meeting any of the following criteria will not be included in the study:; Exclusion Criteria:; 1. Missing date of birth or gender.; 2. Age < 30 or >74 years on the index date (the date of the index antibiotic prescription).; 3. Not enrolled at KPNC or KPSC during the 365 days prior to the index date (allowing gaps of < 60 days). This criterion ensures capture of potential confounders and effect measure modifiers.; 4. Gaps in prescription coverage greater than 60 days during the 365 days prior to the index date, unless there is evidence of a filled prescription in the one year prior to the index date. This criterion ensures opportunities for capturing exposures to the medications of interest, as well as confounders and effect measure modifiers.; 5. More than one type of study antibiotic prescribed on the index date, or within 10 days prior (ie, wash-out period).; 6. Hospitalization within 30 days prior to the index date. This criterion considers that medication changes and medications administered during hospital stays may not be captured.; 7. Residing in a nursing home or other residential institution on the index date or at any time in the preceding 365 days, except for stays of <30 days following hospital discharge. This includes inferred nursing home stays, defined as 2 or more outpatient encounters in the year leading up the index prescription date with procedure codes indicating nursing home place of service separated by at least 28 days. It also includes external cause of injury diagnosis code indicating place of residence was an institution. This criterion considers that the cause of death information recorded on death certificates within a nursing home setting may be less accurate.
• Age groups:
  • Adults (18 to < 46 years)
  • Adults (46 to < 65 years)
  • Adults (65 to < 75 years)
• Estimated number of subjects:
  6900000

Study design details

• Main study objective:
  The purpose of this observational study is to examine the effects of azithromycin use on cardiovascular mortality.
• Outcomes:
  Cardiovascular death, sudden cardiac death, cardiovascular death among those with history of CVD, cardiovascular death among those with high CV risk, Non cardiovascular death and all cause death, and cardiovascular death among those with COPD and CAP
• Data analysis plan:
  The regression analysis will allow for heterogeneity in association between exposures and study endpoints over time (eg, 1 – 5 days vs. 6 – 10 days vs. 11-365 days). Candidate models include a fully parametric piecewise exponential survival model and the semi parametric Cox proportional hazards model with treatment * time interaction term, both allowing for occasional right censoring, separate azithromycin/amoxicillin hazard ratios will be estimated for time intervals of interest (eg, 1 – 5 days vs. 6 – 10 days vs. 11-365 days). Propensity scores will be used to adjust for confounding variables.