Background: As medical therapy improves, splenectomy has been relegated to third- or fourth-line therapy for immune thrombocytopenic purpura (ITP) in many hematologic practices. However, these medications have well-known associated morbidity and changes in treatment algorithms may affect the timing and degree of response to splenectomy as well as complications in heavily treated ITP patients. Materials and methods: This is a retrospective study of consecutive patients who underwent ITP splenectomy from January 1994 to June 2017. Nonresponders after splenectomy and those with recurrent disease were compared to complete responders. Results: The cohort included 84 patients. Median number of medications received before splenectomy was 3 (1-6). 14.3% of patients had a medication-related complication, including heart failure, adrenal insufficiency, diabetes mellitus, infection, and osteoporosis. After splenectomy, 83.5% had a complete response, 7.5% partial response, and 9% no response. Complete response was associated with response to steroids before surgery (P < 0.01). Among responders, 19% had recurrent disease, which was associated with lower platelet count at diagnosis (P < 0.01). Forty-four patients (52.0%) had nonelective splenectomies for persistent bleeding or dangerously low platelets despite maximal medical therapy. Ten patients had Clavien-Dindo grade II or higher surgical complications (11.9%). Seven of these complications were related to recurrent or refractory ITP. Conclusions: Many ITP patients have complications related to medication use, and 52.0% required nonelective splenectomy despite maximal medical therapy. Earlier splenectomy may avoid medication-related complications and may reduce the complications from splenectomy. Splenectomy remains an effective and safe treatment for ITP. (C) 2019 Elsevier Inc. All rights reserved.