If you have diabetes, you're more likely to face health problems like heart and blood vessel diseases. These issues can also affect other parts of your body, like your eyes, brain, and kidneys. Having diabetes puts you at risk of getting several diseases over your lifetime. This can lead to health complications, making life shorter and less enjoyable. This research project uses data from the Clinical Practice Research Datalink to learn about how diseases develop over time in people with diabetes. The goal is to find out what factors contribute to these health issues and create tools that can predict the risk of having multiple health problems. The findings would enable healthcare professionals to offer preventive care to patients with an elevated risk of developing these diseases and to determine the best time to start treatment.
The study aims to determine the extent to which risk factors correspond with the occurrence and timing of macro- and microvascular diseases (MMVD), encompassing the conditions cardiovascular disease (CVD), stroke, vascular dementia, peripheral arterial disease, abdominal aortic aneurysm, nephropathy, retinopathy, and microangiopathy. Of particular interest is MMVD in individuals with type 2 diabetes mellitus (T2DM), who face elevated risks. We seek to determine if these risk factors influence the progression of adverse T2DM outcomes. Initially, we will assess variables in QRISK3, a CVD risk algorithm. We will subsequently expand the list of variables beyond the QRISK3 predictors using the wealth of available EHR data.
Data sources:
Data will be drawn from hospital episode statistics (HES), primary care records, and Office for National Statistics (ONS) data. These sources will provide MMVD diagnoses, fatal events, patient demographics, prescriptions, and lifestyle factors.
The study design:
A cohort study comprising individuals aged 30 at enrollment without prior MMVD history, using data from 1998 to 2021. Participants will be followed until death, MMVD diagnosis, or loss to follow-up. Stratification based on T2DM diagnosis will reflect the MMVD burden among diabetic individuals.
Proposed analyses:
Analyses will involve Cox regression and multi-state models (MSM) to predict MMVD risk, capturing transitions between health states and MMVD. The parameters of the models would estimated by maximising the likelihood function using a general purpose optimiser. MSMs will provide insight to the risk of progressing from a healthy state without MMVD, towards various disease onset, reccurance and death from other causes or MMVD. Ethnicity, sex, age, socioeconomic status, and diabetes will define subgroup indicators. Model performance will be assessed using an 80/20% split for training and testing, evaluating discrimination and calibration.
Primary outcome: time till MMVD occurrence.
Secondary outcome: occurrence of individual disease contributing the MMVD composite, recurance of MMVD, and all cause mortality.
Amand Schmidt - Chief Investigator - University College London ( UCL )
Rikesh Bhatt - Corresponding Applicant - University College London ( UCL )
HES Accident and Emergency;HES Admitted Patient Care;HES Outpatient;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation;CPRD Aurum Ethnicity Record;CPRD GOLD Ethnicity Record