Progressive postinjury thrombocytosis is associated with thromboembolic complications

• Published in: Surgery 2010-10, Vol.148 (4), p.667-675
• Main Authors: Kashuk, Jeffry L., MD, Moore, Ernest E., MD, Johnson, Jeffrey L., MD, Biffl, Walter L., MD, Burlew, Clay C., MD, Barnett, Carlton, MD, Sauaia, Angela, MD, PhD
• Format: Article
• Language: English
• Subjects: Adult; Blood Platelets; Female; Humans; Logistic Models; Male; Middle Aged; Surgery; Thrombelastography; Thrombocytosis - etiology; Thromboembolism - etiology; Thromboembolism - prevention & control; Thrombophilia - diagnosis; Thrombophilia - etiology; Wounds and Injuries - complications; Young Adult
• ISSN: 0039-6060; 1532-7361
• DOI: 10.1016/j.surg.2010.07.013

Background Our previous investigation demonstrated that despite routine chemoprophylaxis, thrombelastography, which is a comprehensive test measuring the viscoelastic properties of blood, identified a hypercoagulable state in a cohort of critically ill surgical patients that was associated with thromboemobolic events. Furthermore, because thrombelastography allows for the comprehensive assessment of coagulation status, this work suggested that platelet hyperactivity is a component of the hypercoagulable state. We hypothesized that progressive postinjury thrombocytosis contributes to a hypercoagulable state that is associated with thrombelastography. Methods One thousand four hundred and forty severely injured patients surviving >48 h were entered into a database prospectively over 12 years. The variables that were evaluated in associated with thrombocytosis (platelet count >450,000) included age, Injury Severity Score, packed red blood cell transfusions in 12 h, and thromboemobolic complications (TE) (deep venous thrombosis, pulmonary embolus, mesenteric thrombosis, stroke, and arterial thrombosis). The time frame for the development of thrombocytosis was assessed at greater or less than 7 days postinjury. Logistic regression was used to identify the independent variables predictive of thrombocytosis and to adjust the association of thrombocytosis with TE for other risk factors. C-statistic was used to assess the discriminative power of thrombocytosis for prediction of TE. Results The mean age was 37.4 ± 0.4 years. The Injury Severity Score was 29.3 ± 0.3, and mean red blood cell transfusions in 12 h was 4.4 ± 0.2 units. Injury via blunt force occurred in 76% of patients, and 72% of patients were male. Thrombocytosis was identified in 447 (31%) patients and was noted almost exclusively >7 days postinjury (98%). TE developed in 35 (8%) of the 447 patients with thrombocytosis, compared with 45 (4.5%) of the remaining 993 patients who did not develop thrombocytosis. Persistent thrombocytosis >7 days was associated with TE ( P > .0001). Logistic regression analysis indicated that when adjusted for intensive care unit duration of stay, transfusions, age, and Injury Severity Score, patients with sustained thrombocytosis more than 3 days were noted to have a 1.4 × increased risk of TE (odds ratio, 1.12; 95% confidence interval, 1.04–1.2; P = .002; C-statistic = 0.82). Conclusion Persistent thrombocytosis in critically injured patients receiving routine chemoproph