This study shows that the level of interobserver agreement was very high for treatment versus no-treatment decisions, but moderate regarding the strength of the recommendations. Further studies are needed to clarify the detailed reasons for the similarities and differences in clinicians' recommendations.