The main objective of this work is to assess the intraoperator reproducibility in the calculation of the HEP score in a population of intensive care patients.

The "HEP score" (HIT Expert Probability) is also a clinical probability score, created thanks to the experience of clinicians (16). It is very recent and therefore still little studied and little used. In particular, it has not been the subject of a prospective validation study for ICU patients.

The use of this score is therefore not recommended in everyday practice.

However, the literature shows a higher diagnostic value of the HEP score especially for resuscitation-type patients, as well as when used by an operator with little experience. This could be due to greater detail given in consideration of other causes of thrombocytopenia (17).

The main objective of this work is to assess the intraoperator reproducibility in the calculation of the HEP score in a population of intensive care patients.

The secondary objectives will aim to:

• Evaluate inter-operator reproducibility
• Determine if the reproducibility of the HEP score is influenced by:

The patient's sex The type of heparin used (LMWH vs UFH) The severity of the patient (IGS2 score) A history of cardiac or orthopedic surgery

The expected consistent benefits allow validation of the HEP score in surgical intensive care (cf. EVHEP-TIH study).

This would allow a better assessment of the pre-test probabilities of TIH with the performance of biological tests.

Inclusion Criteria:

• any adult patient (age ≥ 18 years), admitted to surgical intensive care, treated with heparin (UFH or LMWH), and suspected of TIH by a clinician in the department according to the criteria of the SFAR 2002.

Exclusion Criteria:

• Non-inclusion criteria: minor patients, pregnant women and adults incapable

The cohort studied is made up of all patients admitted for surgical resuscitation between October 2011 and October 2013