A questionnaire was created to collect the prescribing habits for antihyperglycemic drugs in hospitalised patients in internal medicine wards. Data collection was carried out on a specific day between 20 May and 07 June 2019.

Patients with type 2 diabetes admitted during the study period to the internal medicine wards of Spanish hospitals were included. This anonymous questionnaire collected information related to the following variables: demographic data, anthropometric data, laboratory data, frailty and cognitive impairment assessment, prescription of antihyperglycemic drugs and dosage.

The primary endpoint was defined as the use of SGLT2i and GLP1-RA drugs in patients with high cardiovascular risk, or with heart failure as per the indications according to the summary of product characteristics and the funding conditions of the National Healthcare System. In Spain, the funding for SGLT2i drugs is reserved for patients with type 2 diabetes with a glomerular filtration rate higher than 45 mL/min/1.73 m2; on the other hand, the funding for GLP1-RA drugs is reserved for patients with type 2 diabetes with a body mass index greater than 30 kg/m2 and whose glomerular filtration rate is higher than 15 mL/min/1.73 m2.

Frailty: frail patients were defined as those included in groups 5 to 9 on the Rockwood frailty scale.13

Kidney failure: the presence of kidney failure was defined according to the KDIGO 2012 guidelines and was classified depending on the estimated glomerular filtration rate and the degree of proteinuria measured in a urine sample collected in the morning.14 Creatinine values, estimated glomerular filtration rates, and albuminuria/creatinine ratios were retrieved from the medical records with the patient in stable condition prior to hospital admission.

Very high cardiovascular risk: patients with very high cardiovascular risk were defined as those with prior medical history of acute myocardial infarction, acute coronary syndrome, stroke or transitory ischaemic attack, coronary revascularization procedures, aortic aneurysm, and peripheral artery disease. Patients with type 2 diabetes and proteinuria were also included in this group, as well patients with a sustained glomerular filtration rate below 30 mL/min/1.73 m2, or, in other words, those with chronic kidney disease.15

The questionnaire was sent to members of the Spanish Society for Internal Medicine Diabetes, Obesity, and Nutrition Workgroup for volunteer participation in data collection. The final sample included participation from 73 researchers (10.6%) out of a total of 684 members of the workgroup. The bias risk was estimated to be small as the study gathered data on antihyperglycemic drug prescription in patients prior to their hospital admission.

According to the literature data, the population of patients with type 2 diabetes with very high cardiovascular risk aged over 70 years that would be eligible for SGLT2i or GLP1-RA therapy is around 25–40%.9 A 95% confidence interval with a precision of  ± 5% provides an effective sample size of n ≥ 369. The prevalence of diminished glomerular filtration or the presence of proteinuria in patients with type 2 diabetes in Spain is about 35%:10 in patients with type 2 diabetes with glomerular filtration below 60 mL/min/1.73 m2, this value reached 55% in a large cross-sectional study carried out in Spain.11 A 95% confidence interval with a precision of  ± 5% provides an effective sample size of n ≥ 381.

The categorical data is presented as numbers and percentages; the continuous data with normal distribution as mean± standard deviation (SD) and the non-normally distributed continuous variables as average and interquartile range (IQR). The one-way analysis of variance (ANOVA) and the Kruskal-Wallis tests were used to compare the continuous variables between the groups according to the normality of the distribution, and the Chi-squared test was used for the categorical variables.

A logistic regression analysis was performed to identify the determinants associated with the use of SGLT2i and GLP1-RA drugs according to the funding conditions of our National Healthcare System. Two statistical models were created, one whose response variable was the presence of SGLT2i in diabetes treatment prior to admission, and another in which the response variable was the use of GLP1-RA in diabetes treatment prior to admission. Variables presenting an association with a p value of <0.10 in the univariate analysis were considered independent variables. All analyses were performed using the SPSS software (v.22). A p value of  < 0.05 was considered statistically significant.

All data were made anonymous when taken. Clinically unidentifiable information was made available to the researchers. The study was approved by the Hospital Marina Baixa Research Committee and was classified as an audit of current practices in comparison with the best available evidence, meaning it was not necessary to obtained informed patient consent.

Seventy-four hospitals out of a total of 224 belonging to the National Healthcare System participated in collecting the data, representing some 34% (95% confidence interval [95% CI]: 28.0–40.3%). During the study period, a total of 3,290 patients were admitted to the internal medicine wards, of which 908 (27.6%) had a primary or secondary diagnosis of diabetes mellitus. The main reasons for patient admission to the internal medicine ward were infectious disease (46%) or cardiovascular disease (34.5%). Only 6.4% of admissions were related to primary diagnoses of glycaemic decompensation.

The mean age of the patients included in the study was 78.9 years, and mean duration of diabetes since diagnosis was 11 years (Table 1). Mean body mass index was 29. Some 60% had established cardiovascular disease, defined by a history of coronary artery disease (27%), cerebrovascular disease (20.4%), or peripheral artery disease (13.9%); approximately one third of the patients had experienced a prior episode of heart failure. The mean estimated glomerular filtration rate was 56.4 mL/min/1.73 m2. The degree of glycaemic control measured by HbA1c value (HbA1c ≤ 8%) was acceptable in 62.7% of the patients.

Regarding the use of antihyperglycemic therapy, the most used drugs were metformin (49.1%), DPP-4 inhibitors (39%), and insulin (35.2%). The use of SGLT2i and GLP1-RA was 5.8% and 2.6%, respectively.

The use of primary treatments for cardiovascular disease was antiplatelet drugs (40.1%), ACEIs (27.6%), ARB-II (27.7%), beta blockers (36.2%), diuretics (46%) and statins or ezetimibe (53.9%).

A total of 547 (60.4%) patients showed estimated glomerular filtration rate values per the CKD-EPI formula greater than 45 mL/min/1.73 m2, the limit established in our country for funding the prescription of SGLT2i therapy. Among these patients, 65 (12.2%) had peripheral artery disease as a potential contraindication for SGLT2i use. The distribution of SGLT2i use in patients with high risk of cerebrovascular disease, defined as coronary disease, cerebrovascular disease, chronic kidney disease, and congestive heart failure, was less than 10% in all cases (Fig. 1).

Figure 1.

Flowchart of patients who received treatment with iSGLT2.

(0.13MB).

A total of 286 patients had obesity, defined as a body mass index greater than 30 kg/m2, the estimated limit for funding GLP1-RA therapy in our country. Due to the small number of observations, the GLP1-RA drugs with demonstrated cardiovascular benefit (liraglutide and semaglutide) have been grouped with those with demonstrated cardiovascular safety (long-acting exenatide, lixisenatide, and albiglutide). The distribution of GLP1-RA use in patients with very high cardiovascular risk amounts to a maximum value of 6.4% (Fig. 2).

Figure 2.

Flowchart of patients who received treatment with AR-GLP1.

(0.11MB).

Table 2 shows the association between use of SGLT2i treatment and various demographic and epidemiological variables in individuals with an estimated glomerular filtration rate higher than 45 mL/min/1.73 m2 (n = 548). In the univariate analysis, we found a statistically significant association between SGLT2i prescription and age, body mass index, and dependency level. Presence of cardiovascular disease and chronic kidney disease, history of heart failure, male sex, the presence of dementia, or glycated haemoglobin values were not associated with SGLT2i use prior to hospital admission. In the multivariate analysis, patients older than 75 years of age showed lower use of SGLT2i drugs compared to patients under the age of 65 (adjusted odds ratio: 0.28; 95% CI: 0.10–0.74; p = 0.039).

Table 3 shows the association between use of GLP1-RA treatment and various demographic and epidemiological variables in individuals with a body mass index greater than 30 kg/m2 (n = 302). In the univariate analysis, we found a statistically significant association between GLP1-RA prescription, age, and estimated glomerular filtration. Dependency level, presence of cardiovascular disease, history of heart failure, male sex, the presence of dementia, or glycated haemoglobin values were not associated with GLP1-RA use prior to hospital admission. Only age maintained statistical significance in the multivariate analysis. Patients older than 75 years of age showed lower use of GLP1-RA drugs compared to patients under the age of 65 (adjusted odds ratio: 0.09; 95% CI: 0,02–0,46; p = 0,006).