Imperial College London

Department of Primary Care & Public Health, Project Lead: Professor Azeem Majeed

The Data

The dataset contains health records of 100,000 patients with diabetes in NW London. The Clinical Practice Research Datalink (CPRD), one of the largest electronic medical records databases in the world (www.cprd.com), supplied the data.

The data includes:

  • demographic information and registration status;
  • medical history and diagnoses;
  • physical measurements such as height, weight and blood pressure;
  • laboratory test results;
  • drug prescriptions; and
  • referrals to specialists.

The Project

The area of interest is the optimisation of the management of diabetes, including:

  • improving the health of people with diabetes;
  • improving patients’ experience of their primary healthcare;
  • improving health outcomes for patients; and
  • reducing health inequalities.

The specific project objective is to explore the link between quality of care and health outcomes. In England, type 2 diabetes is a major cause of disability and death and accounts for approximately 10% of the country’s health expenditure. This project aims to determine whether the achievement of primary care diabetes targets is associated with improved emergency hospital admission or mortality outcomes, for people with type 2 diabetes.

Using Atmohealth

The analysis will consider potential confounding factors such as socio-demographic indices and baseline comorbidity. It will also consider potential interactions with age, duration of disease and presence or absence of complications. This is a complex analysis and one of the reasons for using Atmohealth is that researchers and clinicians need to be able to visualise the relevant results, without requiring a deep understanding of statistical methods. This will include visualising the baseline characteristics of the patient group and the association of the risk of mortality and hospital admission with key independent variables such as age, sex, socio-economic group and clinical status.