Older people with type 2 diabetes commonly have different levels of frailty associated with multiple medical problems and disabilities. The risks and benefits for older patients from lowering their blood sugar levels to certain target values may be affected by their different frailty levels. We aim to use an electronic frailty index (eFI) to test the feasibility of using eFI scores for individual older patients to determine for each the blood sugar target that optimally balances benefits and risks.
To do this, we will use data collected from GP surgeries nation-wide. From this we shall extract relevant information for all patients above the age of 65 with type 2 diabetes. We will calculate individual eFI scores for patients using data such as age, disability status or co-existing medical problems. We will then look for associations between individual frailty scores and various outcomes: overall risks of death and of severe complications such as heart attacks or strokes, and some risks associated with intensive treatment, such as low blood sugars (hypoglycaemia), requiring hospital admissions. Finally, we will investigate separately for patients with different severities of diabetes (identified using sugar levels from the diabetes monitoring blood test) associations between the same outcomes and different blood sugar targets.
Our intention is to provide evidence to support the use of eFI scores to help determine appropriately differentiated blood sugar targets for older patients with type-2 diabetes, which take into account the different risks and benefits for individuals.
The technical summary should provide a succinct overview of the overarching study aim and objectives, primary exposure(s), and outcome(s), if relevant, study design, and methods including the main statistical tests to be used.
Type 2 diabetes (T2D) is an age related disease and the heterogeneity in older patients health status represent a challenge to making generalised treatment recommendations for older adults. The recent development and validation of an electronic frailty index (eFI) now offers a tool which was supported in the 2016 UK NICE multimorbidity guidelines. In this study, our research question is: Can an eFI, generated from routinely collected data, be used to predict hospitalization and mortality at an individual level of older patients with T2D?
Study design: an open cohort of patients 65 years and older, diagnosed with type 2 diabetes, followed for two years or until all-cause hospitalisation or mortality.
Methods: Multivariate Cox regression analysis will be run to predict the outcome. Model 1 will include social history, polypharmacy and the co-morbidity relevant to risk factor for hospitalization and mortality but not the eFI; Model 2 will be the same but will include the eFI. Likelihood-ratio test will be used to assess the effect of adding the eFI to the model. For both models, D statistics and Harrell's C statistics will be calculated as measures of discrimination. The addition of the eFI will be considered an improvement if a reduction in Akaike's Information Criterion is greater than 4.
The analysis will be repeated for patients grouped by different levels of glycaemic control. The groups will be formed by using quintiles of HbA1c at the study entry.
We believe findings from this study will form the basis for a wider, routine, evidence-based use of eFI when assessing and setting treatment targets for people with T2D, and will have a major impact on the quality of life of patients and in reducing the societal economic burden of treatments.
Composite of All-cause hospitalisation and All-cause mortality deriving from GP or HES data.
Iskandar Idris - Chief Investigator - University of Nottingham
Yana Vinogradova - Corresponding Applicant - University of Nottingham
Adam Gordon - Collaborator - University of Nottingham
Nadeem Qureshi - Collaborator - University of Nottingham
Thomas Crabtree - Collaborator - University of Nottingham
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Townsend Score