Asymptomatic hyperuricemia: to treat or not to treat. A target trial emulation to assess major cardiorenal outcomes (HYPER-TTE-HARV)

Study type

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

  Hypothesis generation (including signal detection)

  Method development or testing
• Data collection methods:
  Secondary use of data

Non-interventional study

• Non-interventional study design:
  Cohort

Study drug and medical condition

• Name of medicine:
  FEBUXOSTAT
• Name of medicine, other:
  Allopurinol
• Study drug International non-proprietary name (INN) or common name:
  ALLOPURINOL

  FEBUXOSTAT
• Anatomical Therapeutic Chemical (ATC) code:
  (M04AA01) allopurinol

  allopurinol

  (M04AA03) febuxostat

  febuxostat
• Medical condition to be studied:
  Hyperuricaemia
• Additional medical condition(s):
  Asymptomatic hyperuricaemia

Population studied

• Short description of the study population:
  Subjects with age >18 and of any sex, with incident asymptomatic hyperuricemia. Those with history of symptomatic hyperuricemia, cancer and cardiorenal events (acute myocardial infarction, ischemic stroke, and kidney disease), with less than 1 previous year of follow-up registered in the database, or contraindications to receive urate-lowering drugs will be excluded. ;
• Age groups:
  Adult and elderly population (≥18 years)
• Estimated number of subjects:
  250000

Study design details

• Study design:
  A retrospective cohort study following the components of a target trial emulation
• Main study objective:
  To evaluate whether to treat asymptomatic hyperuricemia with urate-lowering drugs would reduce the incidence of cardiorenal outcomes, in particular, chronic kidney disease, ischemic stroke, and acute myocardial infarction.
• Setting:
  From 2003 to 2019, a retrospective cohort will be constructed following the components of the target trial that articulates the causal question, as specified below. All subjects fulfilling the inclusion criteria below will be included:; 1) Subjects with age >18 (some urate-lowering drugs are not indicated below this age), and of any sex.; 2) Incident asymptomatic hyperuricemia, defined as a first record of serum uric acid >6.8 mg/dl and without prior records of gout, gout flares, gout arthritis, colchicine use or a similar suggestive term.; 3) No previous history of: cancer (except non-melanoma skin cancer) within the last 3 years, acute myocardial infarction, ischemic stroke, and kidney disease (estimated glomerular filtration rate -eGFR-, albuminuria or proteinuria outside the normal range and/or a diagnosis of chronic or acute kidney disease, dialysis, or kidney transplantation).; 4) A minimum registry of 1-year with their primary care physician in the database with the standards of quality registration applied by the staff of BIFAP. The assessment of the number of previous visits to primary care will help to relax the expectation that subjects will keep active in the health system throughout the study period.; The use of urate-lowering drugs will be compared with the non-use.
• Comparators:
  New use of urate-lowering drugs (allopurinol/febuxostat); Non-use of urate-lowering drugs
• Outcomes:
  Incident cardiorenal events as acute myocardial infarction, ischemic stroke and chronic kidney disease.
• Data analysis plan:
  We will estimate the observational analogs of the intention-to-treat and per-protocol effects. Intention-to-treat needs to adjust for baseline confounders, that is, imbalanced prognostic factors at baseline. The adjustment may be performed either propensity score matching or weighting (inverse probability weighting), or standardization, among others. Due to the characteristics of the database, factors to adjust are those predictors of prescription. Per-protocol analysis needs to adjust for pre- and post-baseline non-adherence to treatments, in addition to those adjustments of intention-to-treat analysis. In per-protocol analysis, the deviation from the initial strategy assigned will result in censorship. As this may introduce post-baseline selection bias, adjustment for predictors of adherence must be performed. Risk curves under each treatment strategy will be constructed. Risks ratios and risk differences at different timepoints will be obtained and compared under different treatment strategies. A Cox proportional hazards model may be also fitted to estimate averaged hazard ratios. Subgroup analysis will be also performed by age, gender, or cardiovascular risk factors. The set of potential confounders will be selected by expert criteria and after the construction and analysis of directed acyclic graphs. The sequence of trials with the cloning-censoring-weighting approach will be applied. If computational constraints occur, a random sample of non-exposed subjects will be selected.