Maximum 24‐hour platelet count fall: Metric for improving the diagnosis of heparin‐induced thrombocytopenia among patients with intermediate probability 4Ts scores

Abstract

Background: Most patients with suspected heparin-induced thrombocytopenia (HIT) and an intermediate probability 4Ts score do not have HIT. We aimed to develop a metric based on the rate of platelet count fall to aid in discriminating HIT status among patients with an intermediate 4Ts score. Methods:We derived a measure of the maximum 24-hour percentage decrease in platelet count (Fall max ) in a cohort of patients with suspected HIT and an intermediate 4Ts score at the University of Pennsylvania. We validated this metric in a prospectively collected cohort of patients with suspected HIT and an intermediate 4Ts score from four hospitals in Hamilton, Ontario.Results: One hundred fifty-eight and 139 patients were included in the analysis from the derivation and validation cohorts, respectively. Fall max was significantly higher in HIT-positive patients in the derivation cohort (49.6% versus 38.6%, P = .009) and validation cohort (43.5% versus 29.3%, P = .027). The area under the receiver operating characteristic curve was 0.68 (95% confidence interval [CI] 0.57-0.78) and 0.71 (0.59-0.83) in the two cohorts, respectively. At Fall max ≥ 30%, sensitivity and specificity were 95.5% and 29.4% in the derivation cohort and 80.0% and 52.7% in the validation cohort.Conclusions: Among patients with suspected HIT and an intermediate 4Ts score, Fall max aided in discriminating HIT-negative from HIT-positive patients. Using a measure that accounts for the rate of platelet count fall may help to avoid unnecessary suspension of heparin and treatment with an alternative anticoagulant in HIT-negative patients with an intermediate probability 4Ts score, though further evaluation is warranted.

Heparin‐induced thrombocytopenia: Construction of a pretest diagnostic score derived from the analysis of a prospective multinational database, with internal validation

“…Compared with a slow decline in platelet counts, a rapid fall is more suggestive, as illustrated by the typical pattern of platelet count changes associated with HIT after cardiac surgery, 38 and as recently reported in a refinement of the 4 Ts score as well. 40 In our study, a platelet count decline ≥40%, whether over 3 days or over 6 days, was associated with HIT, but the strongest association was found for a platelet count decline ≥40% over 6 days. To further refine this item about the most suggestive pace, daily platelet counting would have been required, which is not currently recommended.…”

“…Of note, because clinical suspicion could arise in a somewhat delayed manner as discussed previously, such a decrease pace could be detected within the preceding 10 days. Compared with a slow decline in platelet counts, a rapid fall is more suggestive, as illustrated by the typical pattern of platelet count changes associated with HIT after cardiac surgery, 38 and as recently reported in a refinement of the 4 Ts score as well 40 . In our study, a platelet count decline ≥40%, whether over 3 days or over 6 days, was associated with HIT, but the strongest association was found for a platelet count decline ≥40% over 6 days.…”

“…Among the eight clinical parameters found to be independently associated with HIT, some of them (pace of platelet count decline, type and dose of heparin) are not taken into account in the 4 Ts score. They had been already considered to discriminate HIT from thrombocytopenia because of other causes 27,40 . Whether the incorporation of those parameters into the 4 Ts score or whether the score of the present study actually outperforms the 4 Ts score can only be solved with a formal prospective comparison.…”

“…They had been already considered to discriminate HIT from thrombocytopenia because of other causes. 27,40 Whether the incorporation of those parameters into the 4 Ts score or whether the score of the present study actually outperforms the 4 Ts score can only be solved with a formal prospective comparison. The currently offered Bayesian analysis based on 4 Ts and one based on the GFHT-HIT score, both combined with ELISA OD, deserve to be compared.…”

Diagnosing heparin‐induced thrombocytopenia: The need for accuracy and speed

“…Others have developed scores for particularly challenging patient populations like cardiopulmonary bypass patients who frequently have postoperative thrombocytopenia and nearly universal exposure to heparin (Lillo‐Le Louet score) 14 . Our group demonstrated that a more rapid platelet count fall in 24‐hours makes the diagnosis of HIT more likely in patients with an intermediate 4Ts score, and thus the rate of platelet count fall could be a useful tool following calculation of the 4Ts score to improve the positive predictive value of the score 15 . This approach and other modifications to the 4Ts score require validation in prospective studies.…”