Multifactorial Control of Type 2 Diabetes and Treatment Intensity in Primary Care: A Cohort Study
Multifactorial Control of Type 2 Diabetes and Treatment Intensity in Primary Care: A Cohort Study
Main Objective: To identify changes in metabolic control of type 2 diabetes in patients of the Catalan national health system during 2007-2013.
Study design: Follow-up of a population with type 2 diabetes during 7 years (2007-2013).
Study participants: Approximately 300,000 people with a clinical record in the SIDIAP database.
Source: SIDIAP database. Variables and measurements: Age, gender, time since diagnosis of type 2 diabetes, associated cardiovascular risk factors, micro- and macrovascular complications of diabetes, causes of mortality, treatment for hyperglycaemia and for modifiable risk factors, other conditions related to diabetes and its management. The economic costs of diabetes during the study period will also be analyzed.
Analysis: The clinical variables will be described and 95% confidence intervals will be calculated for the main variables associated with the study outcomes. Missing values will count as non-existing data and missing value imputation will not be used in the analysis. If a subgroup of patients is considered of particular interest, sub analyses will be carried out. Predictive variables of good metabolic control and of complications and mortality of diabetes will be analyzed with the hazard ratios for each factor.
Results:
Applicability: The results related to control of diabetes will serve as a basis for improving metabolic control and decreasing chronic complications and to optimize cost-effectiveness of treatments.
Significance: A long-term study of a population receiving usual treatment is required to understand the changes in indicators of metabolic control, morbidity and mortality and in the management of the disease.
Limitations: Those associated with the use of large population databases.
Main Objective
• To determine the changes in metabolic control of type 2 diabetes during 2007-2013 in patients of the main Catalan public health provider (Catalan Health Service-CHS).
Secondary Objectives
Hypothesis During 2007-2013 the metabolic control of type 2 diabetes patients in the public health system of Catalonia (CHS) improves. The management of type 2 diabetes incorporates new pharmacological treatments.
Methods
Study Design Longitudinal study with cross-sectional, yearly analysis for each year of the study period (2007-2013).
Reference Population and Study Population A longitudinal study with a cross-sectional yearly analysis on June 30 of each year (DTALL) in those patients with a verified type 2 diabetes diagnosis will be carried out with SIDIAP data.
The observation period will be the calendar years 2007 (baseline), 2008, 2009, 2010, 2011, 2012 and 2013 (final).
The SIDIAP database includes the following information of the 5.8 million people from the 282 primary care centres of the Catalan Health Service linked with a unique, anonymized personal identifier:
Data Collection and Information Sources The data sources are constituted by the electronic clinical records, blood tests results, the pharmacy register of the SIDIAP database and mortality data from AQuAS (Health Quality and Evaluation Agency of Catalonia)
Since the study does not target any specific treatment and no specific medication is considered an exposure therapy under investigation, no definition of exposure to any medication is included.
Data Analysis
Data analysis will be performed on study patients meeting inclusion criteria. The clinical variables will be described in terms of absolute and relative frequency for qualitative variables. Central tendency and dispersion measures will be used to describe quantitative variables as appropriate and 95% confidence intervals will be calculated for the main variables associated with the study outcomes.
Statistical tests will be used to compare the distribution of risk factors over time (DTALL): ANOVA will be used for normally distributed quantitative variables, the non-parametric equivalent (Kruskal-Wallis) will be used for non-normally distributed variables, and chi-square test will be used for categorical variables. The P value of the linear trend over the different time periods will be calculated for each variable. The cost associated with diabetes will be calculated for each time cohort.
The likelihood-ratio test will be used to test for interaction between the study variables (eg, HbA1c or cardiovascular events) in nested models with the terms of interaction included or removed. To analyse changes over time of variables to control diabetes in the study cohorts from 2007 to 2013, repeated measures ANCOVA and multinomial logistic regression analysis will be used for quantitative and categorical variables, respectively. These models will be then be adjusted for potential confounders.
The incidence rate per year will used to analyse the frequency of events during the follow-up period (eg. ischaemic heart disease, stroke). The Cox proportional hazard model will be adjusted for potential confounders. Linearity and risk proportionality will be graphically verified in every model using mean weighted residuals.
Missing values will count as non-existing data and missing value imputation will not be used in the analysis.
Statistical comparisons will be performed using two-tail hypothesis testing with level of significance of 0.05. Data management will be conducted by IDIAP Jordi Gol personnel using the SIDIAP database and data analysis will be conducted by epidemiologists and statisticians.
Study Limitations The main limitations of the study derive from the quality of the register of big databases. Previous studies with the SIDIAP showed the validity of most of the variables used in the current protocol.
Ethics
Confidentiality The study will follow the International Guidelines for Ethical Review of Epidemiological Studies (Council for the International Organizations of Medical Sciences -CIOMS-, Geneva, 1991), and the Declaration of Helsinki (revision 64th WMA General Assembly, Fortaleza, Brazil, October 2013).
The management, communication and transfer of data of the participants will comply with the Protection of Personal Data Organic Law of December 13, 15/1999.
Amendments in the study protocol will be communicated in writing to all investigators and to the appointed Clinical Research Ethics Committee (CREC) and to the competent authorities according to the order sas / 3470/2009.
Risk-benefit assessment The current study involves solely access to a database where the data are anonymized and unlinked. It does not carry therefore any risk for the patients.
Data Confidentiality All information in the SIDIAP database is anonymized and unlinked.
Use of electronic data Data extraction will be carried out from the SIDIAP electronic database.
Follow-up and final reports A final report with the results obtained will be reviewed and approved by the IDIAP Jordi Gol researchers. No interim reports are anticipated.
The report should be written following dates in the workplan; a copy of the report will be sent to the Clinical Research Ethics Committee that has authorised the study.
Conditions of Publication The results of this retrospective, observational study will be published in international journals with mention of the Clinical Research Ethics Committee and the sources of funding. None of the investigators will communicate to any third party the results of the study before final results and interpretation have been agreed.
Clinical Research Ethics Committee (CREC) The IDIAP Jordi Gol CREC will evaluate this study.
Conflict of Interests AstraZENECA will provide funding through a contract with the IDIAP Jordi Gol Foundation.
AstraZENECA will not interfere in data extraction, analysis and presentation of reports. The funding is unlinked to the results of the study.
Inclusion Criteria:
Exclusion Criteria: