Clinical differentiation of patent foramen ovale and secundum atrial septal defect: a survey of pediatric cardiologists in Dallas, Texas, USA.

Public health birth defect surveillance registries rely on health care provider diagnosis and definition of congenital anomalies. Major anomalies are likely to have consistent diagnoses across providers; however, definition of some more common, often minor, defects can be problematic. Of particular frustration are the transient neonatal heart findings: patent ductus arteriosus, patent foramen ovale, and pulmonary artery branch stenosis. Under certain circumstances these findings may be considered true anomalies-patent foramen ovale (PFO) as a clinical finding overlaps significantly with atrial septal defect (ASD) of secundum type, the latter being considered a true congenital malformation. Some criteria must be established to separate these conditions in case ascertainment. It is therefore helpful to understand the clinical definitions of patent foramen ovale and secundum atrial septal defect. Pediatric cardiologists in the greater Dallas, Texas metropolitan area were surveyed by telephone, fax, and/or email and asked what criteria they use to distinguish a PFO from a secundum ASD. This was an open-ended question. No baseline parameters were suggested or introduced by the interviewer. Pediatric cardiology fellowship training was identified for each physician to examine the hypothesis that graduates of a given program would use the same diagnostic criteria. Responses were obtained from 22 of 23 pediatric cardiologists. Four measurement criteria were identified: size of the opening, presence or absence of a flap of septal tissue, appearance of the defect on echocardiogram and presence/absence/amount of blood shunting across through the opening. Though there was overlap, diagnostic criteria differentiating PFO and secundum ASD varied among pediatric cardiologists. Two fellowship programs were well represented by the respondent population. Eight respondents were trained at Fellowship 1 and 5 at Fellowship 2. Place of fellowship training was not a strong indicator of which diagnostic criteria were used, even when graduates were in practice together. Physicians in private practice were more likely to report objective measurements as bases for their diagnostic decision. The pronounced variability in clinical definitions will be a problem for birth defect surveillance and research based upon the resultant database. When different physicians use different diagnostic criteria for borderline defects, it is impossible to know whether a defect ascertained and coded with a standard protocol is the same across the population. Since it is unlikely that consistent diagnostic criteria can be put in place, the surveillance program is burdened with compensating for the variability.