ABSTRACT
<b>Objective: </b>The aim of this study was to analyze the factors influencing the addition of clinically significant adverse reactions (CSDR) section in drug package inserts in Japan.<br><b>Methods: </b>The summaries of investigation results from August 2011 to July 2014 were evaluated. The revisions were classified into revisions based only on case reports from Japan ([Revision Y]) and revisions based on other information and/or case reports from Japan ([Revision X]). The revisions were classified into MedDRA system organ class (SOC). As index of amount of information from domestic case reports, the number of accumulated cases ([Case A]), cases for which a causal relationship to the product could not be ruled out ([Case B]), and fatal cases ([Case C]) were used. In each SOC, as index of causal relationship to the product, [Index B/A] ([Case B]/[Case A]) was calculated. Relationship of [Index B/A] to [Revision X]/all revisions, or to the number of [Case A] in [Revision Y] were evaluated. Deference of drug lag between [Revision X] and [Revision Y] was evaluated.<br><b>Results: </b>Three hundreds twenty-three revisions with respect to the addition of CSDR section were identified. [Revision Y] was 203 revisions (63%). The number of [Case A], ([Case B], and ([Case C]) that were required for [Revision X] (120 revisions) were significantly lower than that were required for [Revision Y] (<i>p</i><0.0001 for all comparisons). [Index B/A] tended to inversely correlate with [Revision X]/all revisions (<i>r</i>=−0.52, <i>p</i>=0.066), and the number of [Case A] in [Revision Y] (<i>r</i>=−0.61, <i>p</i>=0.025). Drug lag of [Revision X] was significantly longer than that of [Revision Y] (<i>p</i><0.001).<br><b>Conclusions: </b>In future, it would especially needed to pay attention to adverse reactions with a low [Index B/A] of which revisions relatively depend on other information.