Sun Hee Rim, Jean Colbert, Emily Kobernik, Thomas B Richards, Jacqueline W Miller, Sallyann Coleman King, Julie Townsend, Michael R Nosal, Michael O'Hanlon, Rute Martins, Neelima Karipineni, Lisa C Richardson
Standards-based computable guidelines can support implementation of complex cancer screening and management recommendations within EHRs. Further development and implementation will require improved structured data capture, interoperable workflows, and sustained support. All three CDS tools require longer-term evaluation to assess usability, acceptability, and clinical outcomes.
BACKGROUND: Clinical practice guidelines for cancer screening and management are developed by multiple guideline organizations and published in narrative form. However, guideline concordance can be challenging due to several factors: rapid advances in cancer care, increasing complexity of narrative clinical guidelines, and the unstructured nature of information needed for decisions in electronic health records (EHRs). Clinical decision support (CDS) tools, or computer-interpretable versions of clinical practice guidelines, may help address these challenges when provided to clinicians at the point of care.
MATERIALS AND METHODS: The Centers for Disease Control and Prevention collaborated with MITRE Corporation and subject matter experts to translate current narrative guidelines for cervical, breast, and colorectal cancer screening and management into standards-based computable guidelines and interoperable CDS tools. The tools apply guideline logic in real-time to patient EHR data to assess screening history and risk and generate recommended actions.
RESULTS: Computable guidelines and CDS tools were developed for all three cancers. The cervical cancer CDS was integrated into EHRs and underwent a 6-month clinical pilot evaluation. Breast and colorectal cancer CDS tools progressed to executable code tested with synthetic patient data but have not undergone clinical pilot testing. All three CDS tools are open source, with code available for download.
CONCLUSIONS: Standards-based computable guidelines can support implementation of complex cancer screening and management recommendations within EHRs. Further development and implementation will require improved structured data capture, interoperable workflows, and sustained support. All three CDS tools require longer-term evaluation to assess usability, acceptability, and clinical outcomes.