Interventions to reduce missed diagnostic opportunities in cancer in primary care: an evidence and gap map (EGM)

Abstract

Abstract Introduction Missed diagnostic opportunities [MDOs] for cancer are a significant safety issue. This map identifies and characterises intervention evaluations, highlights evidence gaps and informs the focus of an effectiveness review. Content A policy and practice stakeholder group informed the work. We systematically searched 19 databases for primary studies and systematic reviews published between 2005 and 2025. Eligible evaluations addressed interventions to reduce MDOs in cancer within primary care, in high income countries. Review quality was assessed using AMSTAR-2, and the findings of the mapping exercise are presented in an online interactive map. Summary We mapped 100 primary studies and 12 systematic reviews. Studies mostly targeted skin cancer and to a lesser extent, bowel cancer, with gaps for six other error-prone cancers such as lung and pancreatic. Education, and dermoscopy/teledermoscopy were commonly evaluated for skin cancer whereas biomarkers and direct test access were common for bowel cancer. Few studies assessed education outside skin cancer, examined core service delivery functions (e.g., follow-up and tracking), patient-focused approaches, patient outcomes, and equity. Most existing systematic reviews were low quality. Stakeholders interested in English policy prioritised generalisable approaches such as decision support and direct test access, which were present but not dominant. Outlook Significant gaps remain in evidence on interventions for error-prone cancers (e.g. lung and pancreatic), on specific intervention approaches (e.g., education, core service delivery functions and patient-centred approaches), patient outcomes, equity considerations, and high-quality systematic reviews. Further research is essential for improving diagnostic safety in primary care. This map can guide research efforts.

The authors' abstract, as published at the source. Diagnosis, 2026 · DOI ↗