Determining patient eligibility for a physical activity referral scheme through EHR data extraction

Area of Science:

• Public Health
• Health Informatics
• Exercise Science

Background:

• Physical Activity Referral Schemes (PARS) promote physical activity and improve patient care.
• Accurate retrospective evaluation of PARS requires precise identification of the eligible patient population.
• Current methods may overestimate eligibility due to factors like acute illness or recent surgery.

Purpose of the Study:

• To develop a refined process for accurately determining patient eligibility for physical activity referrals using retrospective electronic health record (EHR) data.
• To improve the accuracy of PARS reach estimation by excluding ineligible patient visits.

Main Methods:

• Developed inclusion criteria based on five chronic conditions identified during PARS training.
• Incorporated exercise contraindications from literature and clinical input for exclusion criteria.
• Mapped inclusion/exclusion criteria to ICD-10 codes and applied them to EHR data.

Main Results:

• An initial pool of 479,536 patient visits was extracted from the EHR.
• Inclusion criteria identified 58% of visits as potentially eligible (277,515 visits).
• Exclusion criteria further reduced the eligible pool by 23% (to 214,312 visits), a 55% overall reduction.

Conclusions:

• A novel, multi-step process was developed to accurately identify patient visits eligible for physical activity referrals from EHR data.
• This methodology enhances the retrospective evaluation of PARS reach.
• Ongoing refinement of exclusion criteria is recommended for optimal accuracy.