Recent commentaries describe the dearth of evidence upon which to base most imaging decisions [1]. The lack of instructive literature is arguably an even larger deficiency in newborn medicine than in adult medicine. Unfortunately, very few imaging modalities have been subject to randomized trial evaluation to assess the value of imaging as a supplement to sound clinical assessment. The irrational drivers of increased imaging and increased use of more sophisticated modalities include fears of liability and clinicians' or patients'/parents' wishes to have the “best” and “latest” imaging technology. In adult practice in Massachusetts, a survey indicated that more than 28% of diagnostic tests were performed for fear of litigation [2]. To use modern imaging technology efficiently and effectively, the clinician is advised to apply some fundamentals of clinical epidemiology. Below we provide a simple critical approach to assessing the potential value of imaging technology in patient diagnosis and management.

Reliability: how reproducible are the results?

“Reliability” refers to the reproducibility or repeatability of results. There are two important types of reliability: intra-rater reliability and inter-rater reliability. Intra-rater reliability is the reproducibility of interpretation of a diagnostic study by one individual. In other words, it asks: “On a different day, would the radiologist/film reader give the same report?” Interrater reliability refers to the repeatability of interpretations between individuals.