Abstract
Abstract 4669
Thrombocytopenia occurs in 20–45% of critically-ill medical-surgical patients. The ‘4Ts’ HIT score (with 4 domains: Thrombocytopenia, Timing of thrombocytopenia, Thrombosis and oTher reason for thrombocytopenia) might reliably identify patients at low risk of HIT. Agreement on 4Ts scoring is uncertain in this setting.
To compare 4Ts HIT scores among research coordinators (who scored real-time), and 2 adjudicators (who scored retrospectively, independently) during an international heparin thromboprophylaxis trial.
763 of 3746 medical-surgical ICU patients in PROTECT met enrolment criteria in this HIT substudy, if any of the following occurred: platelets <50×109/L, platelets decreased to 50% of ICU admission value, venous thrombosis, or if HIT was otherwise clinically suspected. Persons independently completed 4Ts scores blinded to all laboratory HIT results and each others' scores. 475 of these patients had a central or local laboratory HIT test performed and had 4Ts scoring by adjudicators; 432 were scored by research coordinators.
Most patients (410, 86.3%) had a 4Ts score ≤3 conferring a low pretest probability (PTP) of HIT whereas 57 (12.0%) had a moderate PTP, and 8 (1.7%) had a high PTP, as determined by adjudicator consensus. Raw agreement on the 4Ts score was good for central adjudicators overall; however, agreement between adjudicated consensus scores and real time research coordinator scores was not as high. 4Ts agreement (raw and chance-corrected) is presented below.
| . | Adjudicator A vs B (N = 475) . | Adjudicator Consensus vs Research Coordinator (N=432) . | ||||
|---|---|---|---|---|---|---|
| . | Raw . | Kappa . | Weighted kappa . | Raw . | Kappa . | Weighted kappa . |
| Q1. Thrombocytopenia | 92.0% | 0.86 | 0.93 | 76.9% | 0.60 | 0.78 |
| Q2. Timing | 82.3% | 0.60 | 0.68 | 71.1% | 0.39 | 0.47 |
| Q3. Thrombosis | 92.8% | 0.86 | 0.91 | 84.0% | 0.70 | 0.78 |
| Q4. Other | 73.9% | 0.25 | 0.31 | 60.6% | 0.10 | 0.13 |
| Total (low/ intermed/high) | 80.8% | 0.33 | 0.41 | 71.5% | 0.19 | 0.23 |
| . | Adjudicator A vs B (N = 475) . | Adjudicator Consensus vs Research Coordinator (N=432) . | ||||
|---|---|---|---|---|---|---|
| . | Raw . | Kappa . | Weighted kappa . | Raw . | Kappa . | Weighted kappa . |
| Q1. Thrombocytopenia | 92.0% | 0.86 | 0.93 | 76.9% | 0.60 | 0.78 |
| Q2. Timing | 82.3% | 0.60 | 0.68 | 71.1% | 0.39 | 0.47 |
| Q3. Thrombosis | 92.8% | 0.86 | 0.91 | 84.0% | 0.70 | 0.78 |
| Q4. Other | 73.9% | 0.25 | 0.31 | 60.6% | 0.10 | 0.13 |
| Total (low/ intermed/high) | 80.8% | 0.33 | 0.41 | 71.5% | 0.19 | 0.23 |
Real time 4Ts scoring by research coordinators was reasonably close to that obtained by calibrated central adjudicators, suggesting the 4Ts score is somewhat reliable at the bedside when performed by trained personnel. The fourth domain of 4Ts (oTher causes of thrombocytopenia) generated the most disagreement. Further studies of 4Ts scoring by bedside clinicians is needed.
Heart and Stroke Foundation of Canada
Crowther:Pfizer: Consultancy, Honoraria; Leo Pharma: Consultancy, Honoraria; Bayer: Consultancy, Honoraria; BI: Honoraria; CSL Behring: Consultancy; Octaphram: Consultancy; Artisan: Consultancy. Off Label Use: Dalterparin is not indicated for prolonged prophylaxis in critically ill medical patients. Zytaruk:Pfizer: donated study drug dalteparin for PROTECT. Cook:Pfizer: donated study drug dalteparin for PROTECT. Warkentin:Sanofi-Aventis: Speakers Bureau; Pfizer Canada: Speakers Bureau; GlaxoSmithKline: Consultancy, Research Funding; GTI Diagnostics: Consultancy, Research Funding; Canyon Pharma: Consultancy, Speakers Bureau; Informa: Patents & Royalties.
Author notes
Asterisk with author names denotes non-ASH members.
