Evaluation of 4Ts score inter-rater agreement in patients undergoing evaluation for heparin-induced thrombocytopenia. Issue 5 (July 2021)

• Record Type: Journal Article
• Authors: Northam, Kalynn A.; Parker, William F.; Chen, Sheh-Li; Cicci, Jonathan D.; Lin, Feng-Chang; Rollins-Raval, Marian A.; Kasthuri, Raj S.
• Is Part Of: Blood coagulation and fibrinolysis. Volume 32:Issue 5(2021)
• Publication Date: 2021-07
• Subjects: diagnosis -- heparin -- reproducibility of results -- risk factors -- thrombocytopenia; Blood -- Coagulation -- Periodicals; Fibrinolysis -- Periodicals; Hemostasis -- Periodicals; Thrombosis -- Periodicals; Blood Coagulation -- Periodicals; Fibrinolysis -- Periodicals; Hemostasis -- Periodicals; Thrombosis -- Periodicals; 612.115
• DOI: 10.1097/MBC.0000000000001042
• Languages: English
• ISSNs: 0957-5235

Abstract

Abstract : N/A: The American Society of Hematology and American College of Chest Physicians heparin-induced thrombocytopenia guidelines recommend calculation of a pretest probability score prior to performing laboratory testing, and the 4Ts score is commonly used. Inter-rater agreement of the 4Ts score has been evaluated, but limited data are available regarding the reliability of the 4Ts score when performed by nonexpert clinicians. The purpose of this study was to Compare 4Ts scores calculated by medical teams to an expert. A single-center observational study was conducted in patients evaluated for heparin-induced thrombocytopenia over 24 months. The primary outcome was difference in mean 4Ts score calculated by the medical team compared with an expert. Secondary outcomes included inter-rater agreement in risk category assignment and the negative predictive value (NPV) of the 4Ts score. The mean total 4Ts score was significantly higher when calculated by the medical team compared with expert (4.16 ± 1.41 versus 3.42 ± 1.53; P < 0.001). There was slight agreement in risk category assignment (Cohen κ coefficient = 0.164; P = 0.005). The NPV of the 4Ts score was 0.949 (95% confidence interval 0.891–1.000) when calculated by the medical team and 0.927 (95% confidence interval 0.869–0.984) when calculated by expert. Total 4Ts scores calculated by the medical team were significantly higher with only slight inter-rater agreement compared with expert. The NPV of the 4Ts score when calculated by nonexperts may be lower than previously reported. The recommendation to forgo laboratory testing for low 4Ts score patients may need to be revisited.