Similarity-based approaches to identifying risk of future asthma attack using UK primary care data

Study type

• Study type:
  Non-interventional study
• Scope of the study:
  Disease epidemiology

Non-interventional study

• Non-interventional study design:
  Cohort

Study drug and medical condition

• Medical condition to be studied:
  Asthma

Population studied

• Age groups:
  • Adult and elderly population (≥18 years)

    • Adults (18 to < 65 years)

      • Adults (18 to < 46 years)
      • Adults (46 to < 65 years)
    • Elderly (≥ 65 years)

      • Adults (65 to < 75 years)
      • Adults (75 to < 85 years)
      • Adults (85 years and over)
• Estimated number of subjects:
  500000

Study design details

• Main study objective:
  To investigate similarity-based approaches to identifying the risk of asthma attack in the year following an asthma review appointment in primary care.
• Outcomes:
  risk of asthma attack in following 1 year
• Data analysis plan:
  The primary analysis will be carried out using data from OPCRD. The approach used in OPCRD will be externally validated using data from the SAIL databank. Methods of analysis are summarise below: Traditional approach: We will use methods such as logistic regression to derive a score corresponding to the risk of asthma attack in the following year. Similarity-based approaches: We will use metrics such as the Euclidean distance to quantify the similarity between patients. We will then use this metric to identify a subsample using both a neighbourhood approach (i.e. all patients within a specified similarity threshold) and a cluster approach (i.e. identify subsamples using cluster analysis). We will use methods such as logistic regression to derive a risk prediction model from the identified subsample.