Ear-ly Warnings: The diagnostic accuracy of Frank's sign in fatal coronary artery disease - A forensic pathology study

Globally, Coronary Artery Disease (CAD) is the leading cause of death. The mortality burden is rising in South Africa, imposing significant socioeconomic strain. In 1973, it was postulated that Frank's sign could serve as a cutaneous marker for CAD, enabling early, cost-effective, and non-invasive diagnosis with the timeous implementation of secondary preventative measures. However, its reliability as a diagnostic marker remains uncertain, and clinicians have not incorporated it into routine examinations. Given this uncertainty in the clinical context, post-mortem findings can be correlated with confirmed diagnoses in the forensic setting, which offers a unique opportunity to assess the potential association between Frank's sign and underlying CAD. This study aimed to investigate the relationship between Frank's sign and CAD in a deceased population in a forensic pathology mortuary.

A retrospective cross-sectional study was conducted over three weeks on all cases of unnatural deaths involving decedents aged 12 years and older at the Tygerberg Forensic Pathology Services Mortuary in Cape Town, South Africa. The study included 100 decedents (80 males, 20 females) with a median age of 35 [28.0-48.8] years. The presence of a diagonal earlobe crease-also known as Frank's sign-was assessed and graded using a standardized grading system ranging from Grade 0 to Grade 3.

The prevalence of Frank's sign was 27 %, and the prevalence of CAD was 28 %. Frank's sign was significantly associated with fatal CAD (p < 0.001), demonstrating high specificity (86.1 %) and negative predictive value (84.9 %) but relatively low sensitivity (60.7 %) and positive predictive value (63 %). The likelihood ratios were 4.366 (LR+) and 0.456 (LR-). Global measures of diagnostic accuracy, including the Receiver Operating Characteristic curve and area under the curve (0.76), diagnostic odds ratio (9.567), diagnostic effectiveness (0.79), and Youden's Index (0.468), suggest that Frank's sign has only fair discriminative power in detecting fatal CAD.

The absence of Frank's sign may be useful in excluding CAD mortality; however, its presence is not a reliable indicator for diagnosing the disease.