CMC-ID: identifying children and youth with medical complexity using an electronic health record

Area of Science:

• Pediatric Health Informatics
• Clinical Decision Support Systems
• Health Services Research

Background:

• Identifying children and youth with medical complexity (CMC) is crucial for effective care transitions.
• Existing methods for identifying CMC from electronic health records (EHR) may lack precision.
• A robust algorithm is needed to facilitate recruitment into specialized care programs.

Purpose of the Study:

• To develop and evaluate the performance of the CMC-ID algorithm for identifying CMC using EHR data.
• To assess the algorithm's utility for recruiting patients into a transition to adult care program.
• To test the algorithm's applicability in both adolescent and younger pediatric populations.

Main Methods:

• The CMC-ID algorithm was developed iteratively at a tertiary pediatric hospital using EHR data.
• It incorporated standard clinical criteria for CMC: complexity, chronicity, fragility, and technology dependence.
• The algorithm was validated in two phases: initially on older youth (17 to <18 years) and subsequently on younger children (1 year to <17 years).

Main Results:

• The final iteration of CMC-ID achieved a positive predictive value (PPV) of 85.5% in identifying older youth.
• When applied to a younger cohort, the algorithm demonstrated similar performance with a PPV of 89.1%.
• The algorithm successfully identified a substantial number of CMC patients for recruitment.

Conclusions:

• CMC-ID is a pragmatic and effective tool for identifying children and youth with medical complexity for clinical programs.
• The algorithm demonstrates high positive predictive value and consistent performance across different age groups.
• CMC-ID can support recruitment for interventions aimed at improving outcomes for CMC.