Aims: To assess effects of current real-world (RW) treatment patterns on thromboembolism (TE) and mortality rates in patients with cold agglutinin disease (CAD) using the Optum®de-identified Electronic Health Record dataset from the United States, with access to approximately 95 million patients. Methods: Adult patients with ≥1 autoimmune hemolytic anemia (AIHA)-related medical encounter (2007–2018) and ≥3 CAD-related terms from physician notes were retrospectively identified from the dataset (Broome et al, Res Pract Thromb Haemost. In press). Each patient's index date (ID) was the date CAD was first mentioned in their records during the study period. Baseline period was the interval from starting medical activity in the database or study period (whichever occurred later) to the ID. Follow-up period was the interval from the ID to end of the study period or medical activity or death (whichever occurred earlier). Patients were required to have ≥1 medical encounter within 12 months before and after the ID. Mutually exclusive treatment groups were created hierarchically. First, any patient who received rituximab (as monotherapy or combined with other treatments [ie, corticosteroids (CS), chemotherapy, or immunosuppressants]) entered the rituximab group. Second, any patient who received a CS (alone or in combination with other treatments [ie, chemotherapy or immunosuppressants, but not rituximab]) joined the CS group. Third, any patient who received a CAD treatment other than rituximab or CS (eg, transfusion, splenectomy, plasmapheresis, intravenous immunoglobulin) entered the “other treatment”group. Finally, all other patients joined the “no treatment”group. TE (identified using International Classification of Diseases, Tenth Revision, codes) was assessed during the follow-up period. Mortality was captured through the Social Security Death Index or through discharge status on inpatient medical encounters and assessed during a 5-year follow-up period. Results: Of 651 identified patients with CAD (median [interquartile range] age: 72 [17] years; female: 64%; white: 85%; median [range] follow-up time: 41.6 [0.1–140.3] months), 174 were in the rituximab group, 274 were in the CS group, 45 received other treatments, and 158 received no CAD-related treatments. Overall TE and mortality rates were 29% (n = 191/651) and 14% (n = 92/651), respectively. When stratified by treatment, TE and mortality rates were 31% and 14% for the rituximab group, 32% and 12% for the CS group, 33% and 38% for the other treatment group, and 22% and 11% for patients that received no CAD-related treatment, respectively. Conclusions: Current RW therapies for CAD showed no benefit in preventing thrombosis or mortality. Newer treatments that have the potential to address these severe complications are needed for patients with this condition. Data first presented at EHA 2020, 11th-21st June 2020.