Type 2 Diabetes mellitus (T2DM) is a chronic progressive disease despite multiple treatment options. Large clinical trials have highlighted the benefits of achieving glycemic targets to reduce risks of complications for patients with T2DM. Professional Guidelines suggest glycemic control goals below 7% or 6.5% and endorse step-wise algorithms for pharmaceutical treatments. Many patients eventually need insulin replacement when OADs no longer suffice to maintain glycemic goals. The efficacy and safety of injectable glucagon-like peptide-1 receptor agonists (GLP1RA) or basal insulin (BI) have been well-established in randomized controlled trials, yet its effectiveness in the real-world clinical practice setting remains unclear.
Real world observational studies on large patient datasets in combination with advanced analytics are emerging as complimentary tools for assessing therapeutic approaches and their impact in clinical practice.
This study is a retrospective observational study to be conducted to investigate the probability over time of achieving glycemic control for patients with uncontrolled T2DM on oral antidiabetic drugs (OADs) who initiated GLP1RA or BI therapy using the CPRD database. We are to determine the probabilities of reaching glycemic control over time of 24 months of treatment. The data from this study is to help health care professionals determine when to consider T2D treatment intensification for patients on GLP1RA or BI.
Background and aims: Type 2 Diabetes mellitus (T2DM) is a chronic progressive disease despite multiple treatment options. Large interventional studies have highlighted the benefits of achieving glycemic targets to reduce risks of complications for patients with T2DM. The efficacy of injectable glucagon-like peptide-1 receptor agonists (GLP1RA) or basal insulin (BI) for the treatment (Tx) of type 2 diabetes mellitus (T2D) patients (pts) has been well established in randomised controlled clinical trials. However, there is increasing interest and need to also evaluate their clinical effectiveness in real-world (RW) practice. This study will investigate the RW effectiveness of GLP1RA and BI to achieve glycaemic control in T2D pts in the UK Clinical Practice Research Datalink (CPRD) database.
Materials and methods: T2D pts aged ?18 years (y), with HbA1c ?7.0% on oral antihyperglycaemic drugs (OADs), who initiated Tx with a GLP1RA or BI between 01/01/2010 and 30/06/2016 were identified from the UK CPRD database. The index date will be defined as first Tx date with a GLP1RA or BI (independently). Glycaemic control outcomes over a 2-y follow-up period including HbA1c change over time and cumulative probability of reaching first HbA1c <7.0% will be described. Descriptive statistics will be calculated to describe the baseline demographics and clinical characteristics of the study cohort and subcohorts. Means and standard deviations (SD) will be reported for continuous variables and percentages are provided for categorical variables.
Each quarter post-index date will defined to be 90 days, i.e., 090 days as the first quarter (03 months); 91180 days as the second quarter (36 months), etc. The descriptive statistics of HbA1c change from baseline will be calculated semi-annually following GLPRLA/BI initiation.The associations between baseline (BL) characteristics and glycaemic control outcomes will also be analysed by a Cox model.
Primary Outcome:Glycemic control outcomes post GLP-1RA or BI initiation:
a. HbA1c change over time in 720 days post GLP-1RA or BI initiation (via descriptive statistics). The 720-day post GLP-1RA or BI initiation period will be divided into consecutive 180-day periods. The mean change in HbA1c from baseline to each 180-day period will be calculated. For each time-period, patients with both baseline and a follow-up HbA1c will be included. Window periods may be adjusted to ensure adequate numbers. Multiple HbA1c measures will be included, and robust standard errors will be presented.
b. Cumulative probability of reaching HbA1c <7.0% over time in 720 days post GLP-1RA or BI initiation (via Kaplan Meier estimation); Conditional probability of reaching glycemic control every quarter post GLP1RA or BI initiation: This is defined as the probability of reaching glycemic control during a quarterly interval if the patient had NOT already reached control prior to that interval.
c. Association of baseline characteristics to glycemic control outcome post GLP-1RA or BI initiation (via Cox regression model)
Robert Lubwama - Chief Investigator - Sanofi US Services Inc (USA)
Victor Peng - Corresponding Applicant - SANOFI
- Collaborator -