Numbers of articles (with the sum of reported patients in parentheses) that included analysis of variables that were predictive, not predictive, or not interpretable for the response to splenectomy. The assignment of articles to these 3 categories is described in “Methods.” If there were no articles for a variable in a given category, this was designated (—). When predictive associations were reported, responses were associated with younger age, male sex, shorter duration of illness, previous response to steroids or intravenous immunoglobulin (IVIg), higher preoperative platelet counts, less severe bleeding, platelet sequestration predominantly localized to the spleen, decreased platelet lifespan with increased platelet turnover, no megakaryocyte hyperplasia, greater and more rapidly increasing postoperative platelet counts, and the presence of splenic follicle hyperplasia. Data from case series reporting open laparotomy and laparoscopy and data from case series of adults only and adults plus children were combined for this analysis.