Background: Thrombocytopenia contributes to trauma-induced coagulopathy and may require platelet transfusion. Viscoelastic assays can provide rapid surrogates of platelet count, but thresholds may be influenced by physiological derangements. We compared ROTEM-derived parameters for diagnosing thrombocytopenia and tested their robustness across selected conditions. Methods: We performed a retrospective registry-based cohort study in a level-1 trauma center (2011–2021). Severely injured adults were included if they had at least one paired ROTEM and platelet count measurement within 24 h of admission; patients receiving anticoagulants were excluded. Thrombocytopenia was defined as platelets < 100 G/L. Associations between platelet count and ROTEM parameters were assessed with Spearman correlation. Diagnostic accuracy was evaluated using ROC and precision–recall (PR) analyses; optimal thresholds were chosen by maximal F1-score. For the best parameter, performance was explored across hemoglobin level, acidosis, and hypocalcemia. Results: Of 4,395 admissions, 865 patients were included with 1,395 paired measurements; thrombocytopenia occurred in 213 samples (15.3%) and was present at admission in 35 (4%). Platelet count correlated with clot amplitude/elasticity (p<0.001). The platelet component of clot amplitude at 5 min (A5 PLAT) had the best discrimination (AUC 0.929; AUPRC 0.627). The optimal PR-based threshold was A5 PLAT ≤ 21 mm. Performance was mainly affected by severe acidosis and severe hypocalcemia. Conclusions: In severe trauma, ROTEM‑derived A5 PLAT accurately identifies platelet count < 100 G/L and outperforms other ROTEM parameters, supporting its use for rapid platelet-related decision-making. Hypocalcemia and acidosis may reduce precision and should be considered when applying ROTEM‑based thresholds during early trauma resuscitation.
David et al. (Fri,) studied this question.