Open Heart. 2026 Sep 15;13(2):e004310. doi: 10.1136/openhrt-2026-004310.
ABSTRACT
BACKGROUND: Many district general hospitals (DGHs) in the UK do not have resident on-call cardiology registrars overnight; in these situations, it is becoming more common for cardiac outreach practitioners (COPs) to provide specialist input for patients presenting with suspected cardiac events. The aim of this study is to evaluate the exact diagnostic concordance between the initial working diagnoses made independently by COPs and subsequent final discharge diagnoses in patients assessed out of hours and to assess the accuracy of identifying acute coronary syndrome (ACS).
METHODS: A retrospective, single-centre service evaluation was conducted at a UK DGH between January and June 2025. Initial COP diagnoses were compared with final discharge diagnoses, with exact diagnostic concordance reported as percentage agreement. Diagnostic performance for identifying ACS was assessed using sensitivity, specificity, positive predictive value, negative predictive value, overall accuracy and Cohen's kappa (κ).
RESULTS: A total of 779 cases were screened, with 233 patients meeting inclusion criteria following staged exclusions. The overall exact diagnostic concordance was 92.7% (216/233). The 216 exact concordant cases comprised 125 concordant cardiac diagnoses and 91 concordant non-cardiac diagnoses. 17 cases had discordant diagnoses, but none were associated with delayed escalation, investigation or management. COP sensitivity for identifying ACS was 96.0%, specificity was 92.4%, positive predictive value was 85.9%, negative predictive value was 98.0% and overall classification accuracy was 93.6%. Cohen's κ was 0.858 (95% CI 0.789 to 0.927), indicating almost-perfect agreement.
CONCLUSION: COPs demonstrated a high level of exact diagnostic concordance when compared with discharge diagnosis made by consultant cardiologist. This new evidence suggests that COP-led assessments are reliable when working within a clinical pathway.
PMID:42744414 | DOI:10.1136/openhrt-2026-004310