Rationale and background
Art.31 CHMP referral in 2016 on metformin and use in patients with reduced kidney function led to PI updates including introduction of a warning and interaction with concomitant use with iodinated agents for patients undergoing imaging procedures.
Based on the evidence reviewed in the PSUSA, it was considered premature to delete the warning and interaction in the metformin PIs, at this stage. PRAC suggested the possibility to conduct a RWD study to gather more data on this specific issue and understand better if further regulatory action would be deemed necessary Research question and objectives
1. To phenotype and characterise ICA in patients with type 2 diabetes initiating metformin
2. To phenotype AKI and CKD stage using diagnosis codes and eGFR measurements among patients
with type 2 diabetes initiating metformin
3. To characterise patients with type 2 iabetes initiating treatment of metformin in terms of:
a. Demographics (age, sex)
b. Recorded comorbidities
c. Recorded duration from first diabetes diagnosis
d. Previous procedures with ICA
e. CKD stage (most recent in past year)
4. To characterise patients with type 2 diabetes with a first procedure requiring ICA with ongoing
metformin use in terms of previous history of:
a. Demographics (age, sex)
b. Comorbidities
c. Recorded duration from first diabetes diagnosis
d. CKD stage
e. ICA type
f. Time from metformin initiation to first procedure requiring ICA
5. To quantify the occurrence of renal dysfunction and of acute diabetes decompensation among
patients with type 2 diabetes with a first procedure requiring ICA during metformin use specifically:
a. AKI
b. Lactic acidosis
c. Diabetic ketoacidosis
Methods
Study design
New user cohort study
Population
Two cohorts:
1) Metformin new users: new users of metformin between 01/01/2014 and 31/12/2024 (or latest date
available), with a diagnosis of diabetes and at least 365 days of history prior to the date of their first
metformin prescription and no prior use of metformin.
2) ICA new use: first procedure requiring an ICA between 01/01/2014 and 31/12/2024 (or latest date available), with metformin initiation during the study period, ongoing metformin use, previous diagnosis of diabetes and 365 days of prior history before metformin initiation. ICA new use population (Cohort 2) is a subset of the metformin new users (Cohort 1).
We will exclude patients with a record of AKI, ketoacidosis or diabetic acidosis within a year of index date.
Variables
Exposures: ICA, metformin
Outcomes:
• AKI
• Lactic acidosis
• Diabetic ketoacidosis
Covariates for characterisation:
• Demographics (age, sex)
• Comorbidities: Anxiety, asthma, CKD, chronic liver disease, chronic obstructive pulmonary disorder (COPD), dementia, gastroesophageal reflux disease (GERD), heart failure, human immunodeficiency virus (HIV), hypertension, hypothyroidism, inflammatory bowel disease, malignant neoplastic disease, myocardial infarction, osteoporosis, pneumonia, rheumatoid
arthritis, stroke, venous thromboembolism
• Duration of diabetes (days from initial diagnosis of type 2 diabetes to index date)
• Previous use of ICA
• CKD stage
• ICA type
• Time from metformin initiation to first procedure requiring ICA
Objective 5: Survival analyses will be stratified by
• ICA indication/type
• CKD stage (1-5)
Data source
1. SIDIAP (Spain, Primary Care Database)
2. FinOMOP-HILMO (Finland, National Registry)
3. DK-DHR (Denmark, National Registry)
4. CPRD GOLD (United Kingdom [UK], Primary Care Database)
Statistical analysis
Phenotyping and characterisation of ICA, and AKI/CKD using diagnosis codes and eGFR measurements, will be done for patients with type 2 diabetes initiating metformin.
Patient-level characterisation will be conducted in databases based on data availability.
Patient-level characterisation of new metformin users will be conducted at index date (date of first prescription), including patient demographics, comorbidities, time since diabetes diagnosis, previous use of ICA, and CKD stage.
Patient-level characterisation of first procedure requiring ICA during ongoing metformin use will be conducted at index date (date of first procedure), including patient demographics, comorbidities, time since diabetes diagnosis, CKD stage, ICA type and time from metformin initiation to first procedure requiring ICA.
We will estimate Kaplan Meier survival functions to describe the probability of outcome occurrence and median survival to outcomes of interests, specially, AKI, lactic acidosis, diabetic ketoacidosis among patients with type 2 diabetes with a first procedure requiring ICA during metformin use stratified by CKD stage, and ICA type.
For all analyses a minimum cell counts of 5 will be used when reporting results, with any smaller counts will be noted as <5.