Effect of Ascertainment Bias on Estimates of Patient Mortality in Inherited Cardiac Diseases

Eline A Nannenberg,Ingrid A W van Rijsingen,Paul A van der Zwaag,Maarten P van den Berg,J Peter van Tintelen,Michael W T Tanck,Michael J Ackerman,Arthur A M Wilde,Imke Christiaans
DOI: https://doi.org/10.1161/CIRCGEN.117.001797
Abstract:Background: Accurate estimates of survival are indispensable for cardiologists, clinical geneticists, and genetic counselors dealing with families with an inherited cardiac disease. However, a bias towards a more severe disease with a worse outcome in the first publications may not accurately represent the actual survival forecast. We, therefore, evaluated the effect of ascertainment bias on survival in 3 different inherited cardiac diseases (idiopathic ventricular fibrillation, SCN5A overlap syndrome, and arrhythmogenic cardiomyopathy) caused by a founder mutation. Methods: We collected mortality data from mutation-positive subjects with either DPP6-associated idiopathic ventricular fibrillation, SCN5A overlap syndrome, and PLN-R14del-mediated arrhythmogenic cardiomyopathy >2 to 10 years of ongoing clinical/cascade genetic screening. Results: The median age of survival in DPP6 mutation-positive subjects increased from 44.6 years in the original cohort from 2008 (n=60; 95% CI, 36.8-52.4 years) to 68.2 years in the extended cohort from 2012 (n=235; 95% CI, 64.6-71.7 years; P<0.001). In the SCN5A overlap syndrome, survival increased from 56.1 years in 1999 (n=86; 95% CI, 48.0-64.2 years) to 69.7 years in 2009 (n=197; 95% CI, 61.3-78.2 years; P=0.049). In PLN-R14del positive patients, the median age of survival increased from 63.5 years in 2010 (n=89; 95% CI, 59.1-68.0 years) to 65.2 years in 2012 (n=370; 95% CI, 62.0-68.3 years; P=0.046). Conclusions: The median age of survival in 3 different inherited cardiac diseases with an established pathogenic substrate significantly increased once genetic testing and cascade screening extended, after the first publication that elucidated the discovery of the disease-susceptibility gene/mutation. This underscores the direct and negative influence of ascertainment bias on survival forecasts and the importance of ongoing clinical/genetic follow-up to establish the most accurate disease prognosis for genetically mediated heart diseases.
What problem does this paper attempt to address?