Postoperative blood loss as measured by chest drain output for intention-to-treat patients (A) and post hoc analysis patients receiving room temperature–stored platelets, 7 days cold-stored platelets, and 8 to 14 days cold-stored platelets (B). The dots represent the individual study patients. The lines show median with interquartile range. The difference between the room temperature– and cold-stored for 7 days arms was not statistically significant (P_{intention-to-treat} = 0.265, P_{post hoc} = 0.115, two-tailed Mann–Whitney U test; SPSS Statistics for Windows, version 24.0, IBM, USA). The nonconcurrent stage II cold-stored for 8 to 14 days arm was not compared with the two stage I arms.