HIT: 4Ts and next steps

Four items, 0–2 points each, total 0–8 (Lo 2006). Low 0–3, intermediate 4–5, high 6–8. A blank item is missing, not zero, so no band appears until all four are answered.

• Thrombocytopenia: 2 = fall over 50% with nadir 20 or more; 1 = fall 30–50% or nadir 10–19; 0 = fall under 30% or nadir under 10 (×10⁹/L). Use the peak count after heparin started; a patient can score without being thrombocytopenic (May & Cuker 2024). In the Hamilton cohort a fall over 50% caused directly by surgery scored 1, not 2 (Lo 2006 footnote); a later reproduction words this as a fall over 50% within 3 days of surgery (Alhanshani 2023).
• Timing: 2 = clear onset days 5–10, or a fall within 1 day with heparin in the past 30 days; 1 = consistent with days 5–10 but unclear, onset after day 10, or a fall within 1 day with heparin 30–100 days ago; 0 = fall before day 4 without recent exposure. After cardiac surgery the fall is often biphasic: surgical drop, rebound, then the HIT fall.
• Thrombosis or sequelae: 2 = new confirmed thrombosis, skin necrosis at injection sites, or an acute systemic reaction after an IV UFH bolus; 1 = progressive or recurrent thrombosis, non-necrotizing skin lesions, or suspected unproven thrombosis; 0 = none. Progression means objectively documented growth; recurrence means a new clot in a previously affected region. Count only thrombosis that occurs or progresses after heparin exposure, while platelets are falling or shortly before (May & Cuker 2024).
• Other causes: 2 = none apparent; 1 = possible; 0 = definite. This item drives most disagreement between scorers.

The helper computes fall % = (peak − nadir) ÷ peak and suggests a Thrombocytopenia score. You still pick the score. When counts meet both the 1-point and the 0-point wording (e.g. fall 30–50% with nadir under 10), the table does not rank them and no suggestion is made.

Missing information: ASH advises getting the data. If a key item cannot be judged, it may be prudent to err toward the higher score (ASH 2.1 remark).

Performance: in Lo 2006 antibodies were found in 1 of 64 (1.6%) and 0 of 55 low scorers, 28.6% and 7.9% at intermediate, and 100% and 21.4% at high (Hamilton, Greifswald). A low score rules HIT out well; what intermediate and high mean depends on the setting.

• 2.1 (strong): use the 4Ts rather than gestalt; immunoassay at intermediate or high. Conditional: functional assay if the immunoassay is positive and one is available. It may be skipped with a high score and a very strongly positive immunoassay (e.g. ELISA over 2.0 OD). If an ELISA is used, a low positivity threshold is preferred.
• 2.2–2.3 (strong): low score: no HIT testing and no empiric treatment, so heparin continues. Testing can be reasonable if the score itself is uncertain.
• 2.4 (intermediate, no other indication): stop heparin (strong); non-heparin anticoagulant at prophylactic intensity if high bleeding risk, therapeutic otherwise (conditional).
• 2.5 (intermediate, another indication): stop heparin (strong); therapeutic intensity (conditional).
• 2.6 (high, strong): stop heparin; therapeutic-intensity non-heparin anticoagulant.
• 2.7–2.8 (strong): negative immunoassay: stop the non-heparin agent and resume heparin if indicated. Rare false negatives: re-evaluate; consider a repeat or different immunoassay or a functional assay.
• 2.9–2.10 (strong): positive immunoassay: keep avoiding heparin and give therapeutic intensity, stepping up from prophylactic if needed.
• 3.1: acute HIT: stop heparin, start a non-heparin agent (strong); argatroban, bivalirudin, danaparoid, fondaparinux or a DOAC (conditional).
• 3.2 (strong): therapeutic rather than prophylactic intensity. 3.3: no added antiplatelet without another indication. 3.4 (strong): no routine IVC filter.
• 3.5 (strong): no VKA before platelet recovery (usually 150 ×10⁹/L or more). If already on a VKA: stop it and give IV vitamin K with the non-heparin anticoagulant.
• 3.6 (conditional): no routine platelet transfusion at average bleeding risk. Transfusion may be an option with active bleeding or high bleeding risk.
• 3.7 (conditional): isolated HIT: bilateral leg compression ultrasound; arm ultrasound only in a limb with a central venous catheter. Symptoms mean diagnostic imaging regardless.
• 3.8 (conditional): isolated HIT without DVT on screening: anticoagulate at least until platelet recovery, and not for 3 months or more unless platelets have not recovered. With thrombosis (HITT), duration is not addressed by ASH (3.8 remark); 3–6 months is typical.
• 3.9 (conditional): after platelet recovery (subacute HIT A), a DOAC rather than a VKA.

Phases: acute HIT lasts until platelet recovery (within 7 days in 90%); subacute A runs until the functional assay turns negative (median 50 days); subacute B until the immunoassay turns negative (median 85 days); then remote HIT. Emergency identifier for 3 months after HIT, not longer (8.1).

Argatroban (US label, HIT): stop heparin and get a baseline aPTT first. Start 2 mcg/kg/min by continuous infusion, no bolus. Check the aPTT 2 h after starting and after each change; target 1.5–3× baseline, not over 100 s. Maximum 10 mcg/kg/min. With moderate or severe hepatic impairment (Child-Pugh), start at 0.5 mcg/kg/min (about 4-fold lower clearance), titrate carefully, and expect steady state to take longer. No renal adjustment. Contraindicated with major bleeding or prior hypersensitivity. The label's infusion table assumes 1 mg/mL. ASH: in Child-Pugh B or C, avoid argatroban or use a reduced dose. ASH Table 2 also lists lower starting rates in heart failure, anasarca and after cardiac surgery; these are not in the US label, so agree a rate with pharmacy.

Moving to warfarin (argatroban label): only after platelet recovery (ASH 3.5). No loading dose; overlap with argatroban. At argatroban 2 mcg/kg/min or less, stop the infusion when the INR is over 4 on combined therapy, then recheck the INR in 4–6 h; if it is below range, restart argatroban and repeat daily. Above 2 mcg/kg/min the combined INR is less predictable: temporarily reduce argatroban to 2 mcg/kg/min, repeat the INR 4–6 h later, then follow the same steps.

Other agents. None of these has a US label for treating HIT. ASH 2018 Table 2 gives suggested regimens, extrapolated from VTE with limited HIT data. Apply the same contraindications as for VTE treatment; see the VTE treatment tool for label renal cutoffs and interactions.

• Bivalirudin: labeled only as the anticoagulant for PCI, including patients with HIT. ASH prefers it for PCI in acute HIT (5.1). For HIT treatment outside PCI it is off-label: 0.15 mg/kg/h with no bolus, aPTT 1.5–2.5× baseline, and a lower dose may be appropriate with renal or liver dysfunction (ASH Table 2). Short-acting: an option when critically ill or bleeding-prone.
• Fondaparinux: off-label in HIT. Reasonable when stable at average bleeding risk. ASH Table 2: 5 mg (under 50 kg), 7.5 mg (50–100 kg) or 10 mg (over 100 kg) subcutaneously once daily.
• DOACs: off-label in HIT. Reasonable when stable at average bleeding risk; a parenteral agent is preferred for life- or limb-threatening thrombosis. Rivaroxaban has the most HIT data: 15 mg twice daily for 3 weeks then 20 mg daily with thrombosis; 15 mg twice daily until platelet recovery without thrombosis, followed by 20 mg once daily if ongoing anticoagulation is indicated (3.1 remark). Apixaban: 10 mg twice daily for 1 week then 5 mg twice daily with thrombosis; 5 mg twice daily until recovery without. Dabigatran: 150 mg twice daily after at least 5 days of a parenteral non-heparin agent with thrombosis; 150 mg twice daily until recovery without (ASH Table 2 and 3.1 remark).
• Danaparoid: an ASH option, but there is no US label on DailyMed.

Liver. Child-Pugh B or C: argatroban only at a reduced start (label 0.5 mcg/kg/min; ASH: avoid or reduce), so bivalirudin is favoured. Rivaroxaban: avoid in Child-Pugh B or C (label). Apixaban: not recommended in Child-Pugh C, and the label gives no dosing advice in B.

Kidney. Argatroban needs no renal adjustment (label §12.3). Fondaparinux is contraindicated at CrCl under 30 (label §4); the tool's 15–30 band includes 30, so it is not used at 30 or under. Dabigatran has no DVT/PE dosing at CrCl 30 or under or on dialysis. Rivaroxaban: avoid at CrCl under 15, including dialysis. Apixaban: no renal dose adjustment for VTE in the label (§8.6), including dialysis, but AMPLIFY excluded CrCl under 25 and the trials did not enroll CrCl under 15 or dialysis. Bivalirudin (off-label, retrospective; Tsu & Dager 2011, 135 patients): rates needed to reach the aPTT goal were 0.13 mg/kg/h with CrCl over 60, 0.08 at 30–60, 0.05 under 30, and 0.07–0.09 on intermittent dialysis, SLEDD or CRRT. On dialysis with acute HIT, ASH suggests argatroban, danaparoid or bivalirudin for the circuit (6.1); regional citrate is not appropriate in acute HIT (6.2 remark).

Prothrombin complex concentrates. KCENTRA and BALFAXAR contain heparin, and their labels contraindicate them in HIT (§4). Plan urgent VKA reversal with the anticoagulant reversal tool.

Persisting or autoimmune HIT. If platelets do not recover on a non-heparin agent, involve hematology. IVIG 1 g/kg/day for 2 days has been reported in severe or refractory cases (off-label; May & Cuker 2024). Anticoagulate at least until platelet recovery (ASH 3.8).

• Cardiac surgery: in subacute HIT B or remote HIT, intraoperative heparin, limited to the operation (4.2, conditional); monitor platelets afterwards, since delayed-onset HIT beginning 5–10 days after intraoperative heparin has been reported (4.2 remark). In acute HIT or subacute HIT A, delay surgery if feasible (4.1a).
• PCI: in subacute HIT B or remote HIT, bivalirudin rather than UFH (5.2, conditional).
• Dialysis: in subacute or remote HIT without other anticoagulation, regional citrate for the circuit (6.2, conditional).
• VTE treatment or prophylaxis in remote HIT: a non-heparin anticoagulant rather than UFH or LMWH (7.1, strong).
• Records: keep heparin on the allergy list indefinitely (May & Cuker 2024); an emergency identifier for 3 months after HIT, not longer (8.1).

PCI dosing during acute HIT (ASH 5.1) and renal replacement circuit protocols (6.1) beyond the agent choice shown. Cardiac surgery is covered only as a summary block; for bypass in HIT see the 2026 ISTH SSC guidance and the 2026 AATS expert consensus. Screening platelet counts while on heparin is ASH 1.1. For confirmed DVT or PE, use the VTE treatment tool and the VTE guide.

Guideline check (October 2026): no ASH or CHEST HIT guideline has been published since ASH 2018 (CHEST last covered HIT in 2012). ASH reviewed its 2018 guideline in 2022 and continues to monitor the evidence. UK: the British Society for Haematology published a third-edition HIT guideline (Br J Haematol 2024;204:459–75); this tool follows ASH 2018.