This study was a part of the Rovaniemi Primary Care T2D Study, which is a registry-based real-world study in a primary health care setting. Rovaniemi is a city and municipality located in northern Finland with a total population of 62,000 people living in both urban and rural areas. The study population consisted of 5104 patients who had received a T2D diagnosis between November 1, 2011 and February 19, 2019 at the Rovaniemi Health Center, Rovaniemi, Finland. The T2D diagnosis was based on the International Classification of Disease (ICD- ) codes for T2D (E11.1–E11.9) or the equivalent T2D code (T90) from the International Classification of Primary Care (ICPC). Patients with at least two evaluations of eGFR with a minimum interval of six months were included in the study. The data were retrieved from patient records. In the analyses, the baseline was considered as the first measurement of eGFR and the follow-up as the last measurement of eGFR within each patient. The study period was defined as the time between the baseline and follow-up, varying individually.
Reduction of eGFR
The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation formulation, which is based on age, sex, race, and serum creatinine level (μmol/l), was used to calculate the eGFR (ml/min/1.73 m2). A substantial reduction in eGFR was defined as a reduction of ≥25% between the baseline and the follow-up in accordance with the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines in CKD evaluation and management . The “no substantial reduction” category included all others except those with substantial eGFR reduction.
Screening frequency of eGFR
The mean value of eGFR screening frequency between 2011 and 2019 was calculated and then dichotomized as yearly (365 days or under) vs. non-yearly (over 365 days), in line with the current national screening frequency recommendations [13, 19]. The first category was used as a reference.
We assessed the sex, age, achievement of treatment goals of bioclinical variables (HbA1c, LDL, and sBP), and prescription of antihyperglycemic and cardiovascular medications (Anatomical Therapeutic Chemical, ATC codes): statins (C10AA), any long-acting insulin (A10AE, A10AC), GLP-1 RAs (A10BJ), SGLT2i (A10BK), and angiotensin-converting enzyme inhibitors (ACEi) (C09A) or angiotensin II receptor blockers (ARB) (C09C) as potential covariates [2, 11, 13, 14, 16]. The preferred source of BP data was the patients’ own home measurements. If home measurements were not available, the measurements performed by a healthcare professional during the consultation visit were used [20, 21].
Age was measured at the follow-up. Achievement of the treatment targets for HbA1c, LDL, and sBP was defined according to the national guidelines as follows: HbA1c < 53 mmol/mol, LDL < 2.5 mmol/l, and sBP < 135 mmHg . This was estimated at the baseline and follow-up. The patients were divided into two groups in terms of all three variables: 1) had achieved the treatment target at both time points, and 2) had not achieved the treatment target at both time points.
Data on prescribed or renewed prescriptions of statins, any long-acting insulin, GLP-1 RAs, SGLT2i, and ACEi/ARB were collected from the national electronic prescription registry using ATC codes. Medication data (new prescriptions or prescription renewals) were further processed by calculating the midpoint of the study period for each patient. Baseline medication was defined as before whereas follow-up medication was defined as after the midpoint. If the prescription was valid at the midpoint, then it was assumed to be the same at the baseline and follow-up. Follow-up medication data were used in the logistic models.
Body mass index (BMI) was calculated by dividing the patient weight in kilograms by the square of their height in meters (kg/m2) and considered a continuous variable. In addition to the previously mentioned bioclinical variables and medications, the following measurements were gathered from the patient records: hemoglobin (Hb; g/l), diastolic BP (mmHg), and medications (ATC codes), such as calcium blockers (C08CA), beta blockers (C07AB), diuretics (C03), metformin (A10BA02), gliptins (A10BH), glitazones (A10BG), sulphonylureas (A10BB), fibrates (C10AB), multiple daily insulin injections (A10AE or A10AC and A10AB), and ezetimibe (C10AX).
Study protocol and data collection
The data were recorded as part of each patient’s routine control visits at the health care center or during other visits and collected and handled anonymously using patient IDs for scientific purposes. Being a registry-based study, no written consent from the patients was required, in accordance with current Finnish legislation. The Ethics Committee of Lapland Central Hospital, Rovaniemi, Finland approved the study protocol in May 2018.
Clinical outcome measures were presented as mean and standard deviation (SD) and categorical variables as proportions. Continuous variables were tested with the independent samples t-test, while the Pearson χ2 test was used to evaluate the difference between categorical values. A multiple-multivariable binomial logistic regression analysis with odds ratios (ORs) and 95% confidence intervals (CIs) was performed to study the potential associations between substantial eGFR reduction and the screening frequency of eGFR. The logistic models were stratified by the baseline kidney function as normal (eGFR≥60 ml/min/1.73 m2) and impaired (eGFR< 60 ml/min/1.73 m2) kidney function, as the screening frequency recommendations differ between patients with and without impaired kidney function . All statistical analyses were performed using the R software version 4.1.1. R Core Team (2020). A p-value < 0.05 was considered statistically significant.