Validation of a method to establish practice-based stroke and TIA registers

Bruce Willoughby1, Richard Thomson, Rakesh Chopra

  • 1Public Health, Northumberland Care Trust, Loansdean, Morpeth, Northumberland NE61 2DL, UK. bruce.willoughby@northumberlandcaretrust.nhs.uk

Insights

Establishing stroke and transient ischaemic attack (TIA) registers is crucial. A pragmatic method, involving fewer patient notes, proved nearly as effective as an extensive review, suggesting its utility for practices with limited coding resources.

Area of Science:

  • Neurology
  • Public Health
  • Health Informatics

Background:

  • Establishing accurate stroke and transient ischaemic attack (TIA) registers is challenging for practices with limited diagnostic coding capabilities.
  • Existing methods for register creation can be resource-intensive, requiring extensive review of patient records.

Purpose of the Study:

  • To compare the effectiveness and efficiency of two distinct methods for establishing stroke and TIA practice-based registers.
  • To identify a more resource-efficient approach suitable for primary care settings with limited diagnostic coding.

Main Methods:

  • Two arms were compared: an 'extensive arm' involving a detailed notes review of patients with antiplatelet/anticoagulant use, ischemic heart disease (IHD), or diabetes.
  • A 'pragmatic arm' involved a similar notes review but supplemented by direct patient questioning during IHD or diabetic clinics, focusing on stroke/TIA history.
  • The pragmatic arm required reviewing only 3% of practice notes compared to 11% in the extensive arm.

Main Results:

  • The pragmatic arm, despite reviewing significantly fewer patient notes (3% vs. 11%), achieved a yield comparable to the extensive arm.
  • This indicates a high degree of efficiency in identifying relevant cases for stroke and TIA registers using the pragmatic approach.

Conclusions:

  • The pragmatic method for establishing stroke and TIA registers is a viable and efficient alternative for practices with limited diagnostic coding.
  • This approach offers a practical solution for improving data collection and register accuracy in primary care settings.