Venous thromboembolism (VTE) is the second leading cause of death in patients with cancer, after disease progression. VTE is increasingly recognized as a complication in patients with hematologic malignancies and various studies have reported high rates of VTE. Critically ill patients are at high risk of VTE and should all receive thromboprophylaxis. Given the increasing number of patients with HM (hematologic malignancies) / HCT (Hematopoietic cell transplantation) who develop critical illness, and their often prolonged course, it is imperative to understand the incidence and risk factors for VTE, and to evaluate the efficacy and risks associated with both chemical and mechanical thromboprophylaxis Therefore, the investigators plan to evaluate retrospectively the VTE / PE (pulmonary embolism) incidence in HM /HCT patients at the University of Toronto, and the complications associated with it (including death). In addition, the investigators want to evaluate the use, type (mechanical or pharmacological) and timing of thromboprophylaxis. And lastly, the investigators will determine the incidence of bleeding and of complications associated with chemical and mechanical thromboprophylaxis. The investigators will describe the change in VTE incidence over the last 10 years.

The investigators know that patients with COVID-19 infection are at higher risk of thrombosis than non-COVID patients. As such, HM/HCT COVID-19 pts will comprise a subgroup, which will be compared with patients who are not not positive for COVID-19. If these numbers are low, COVID-19 status will be included as a predictive variable in our modelling.

The results of this research program will help define indications and safety of VTE prophylaxis; and will inform the development of clinical practice guidelines.

2.0 BACKGROUND AND SIGNIFICANCE Venous thrombosis is a disease of aging, with a low rate of about 1 per 10,000 annually before the fourth decade of life, rising rapidly after age 45 years, and approaching 5-6 per 1000 annually by age 801. The most common risk factors are hospitalization, cancer and surgery. Up to 20% of patients admitted to a medical service will have thrombosis and up to 40% of those admitted to a surgical service. About 10% of all deaths in hospitals are related to pulmonary embolism, and many times it was not suspected before death.

Venous thromboembolism (VTE) is the second leading cause of death in patients with cancer, after disease progression2. The incidence of venous thrombosis during the first 6 months after cancer diagnosis is 12.4 per 1000, with varying incidences depending of the type of cancer, i.e., cancers of the bone (56/1000), ovary (45/1000), uterus (38/1000), Hodgkin's lymphoma (36/1000), breast (35/1000) and brain (32/1000) had the highest incidence. Cancers of the ovary, pancreas, lung, stomach, and hematological malignancies tend to present a high incidence of venous thromboembolism in the year before the cancer diagnosis.

VTE is increasingly recognized as a complication in patients with hematologic malignancies and various studies have reported high rates of VTE. Another high risk population for bleeding and DVT is the HSCT (hematopoietic stem cell transplant), with a 3.7% 1-year incidence of VTE, and 8-20% incidence of catheter-related thrombosis (CRT)3. In addition to a risk of thrombosis, HSCT recipients also have an increased risk of bleeding because of prolonged and severe thrombocytopenia. The initiation of therapeutic anticoagulation during days 1-180 after HSCT was the strongest predictor of bleeding (odds ratio (OR) 3.1[95% confidence interval (CI) 1.8-5.5)].

The timing of hemostatic complications is also an important consideration as bleeding events are more likely to occur early in the post-transplant course when patients are profoundly thrombocytopenic, whereas thrombotic events occur more frequently after hematopoietic recovery Critically ill patients are at high risk of VTE and should all receive thromboprophylaxis. Patients with HM or HCT who are thrombocytopenic (platelets <30) are ineligible for pharmacologic prophylaxis because of the high risk of bleeding. In these patients, it has been suggested to apply mechanical antithrombotic measures 4, but not all centers have sequential compression devices (SCD). In the general population of patients admitted to the hospital, it has been suggested that SCDs do not reduce the risk of VTE 5. In critically ill patients, evidence of benefit of SCD use is lacking. Further, their use is associated with costs, discomfort, skin breakdown, and mobility restrictions. In healthcare settings lacking SCDs, patients receive no thromboprophylaxis until the platelet count recovers.

Early after chemotherapy or HCT, two recent studies in non-ICU patients suggest that rates of non catheter-related VTE are low, raising the question of whether any prophylaxis is required. In a prospective study of 1,072 consecutive adult cancer patients, only 30 VTE events were observed during follow-up, 5 (2.3%) in HM patients and 25 (2.9%) in solid tumor patients (relative risk [RR]: 0.79, 95% CI: 0.30-2.03; p=.79) (2). Four of the five episodes in HM patients were related to a catheter. Major bleeding was infrequent in both groups (1.8 vs. 3.9%; RR: 0.48, 95% CI: 0.17-1.33; p=.21), and thromboprophylaxis was not associated with an increased incidence of major bleeding.

While thromboprophylaxis may not be necessary in non-ICU patients, data in critically ill patients regarding the incidence of VTE, effectiveness of thromboprophylaxis, and bleeding risk, is lacking. Moreover, the risk of VTE may be higher later in the course, when the platelet count recovers, or in patients with graft vs host disease or sinusoidal obstructive syndrome.

Given the increasing number of patients with HM/HCT who develop critical illness, and their often prolonged course, it is imperative to understand the incidence and risk factors for VTE, and to evaluate the efficacy and risks associated with both chemical and mechanical thromboprophylaxis.

3.0 HYPOTHESES AND SPECIFIC AIMS 3.1 Specific Aim 1 Determine the use, type (mechanical or pharmacological) and timing of thromboprophylaxis.

3.2 Specific Aim 2 Determine incidence of catheter and non catheter-related VTE (upper / lower extremity DVT and PE) and explore variables that are independently associated with it.

3.3 Specific Aim 3 Determine incidence of bleeding, complications associated with SCDs, and complications associated with VTE, including death. Bleeding severity will be classified according to current accepted standards6.

3.4 Hypothesis In critically ill patients with hematologic malignancy (HM) who are thrombocytopenic, venous thromboembolism (VTE) is infrequent within the first 30 days following induction chemotherapy or hematopoietic cell transplant (HCT), rendering it possible to avoid the use of thromboprophylaxis (mechanical or pharmacological) and the associated risks (i.e. serious bleeding).

The investigators propose a program of research on VTE in HM/HCT patients. This program consists of several projects which will inform the design of a future international multi-center RCT. It will consist of three consecutive phases:

1. To conduct a systematic review to evaluate the incidence and timing of VTE, and the threshold for initiation of chemical thromboprophylaxis.
2. To evaluate retrospectively VTE incidence in HM / HCT patients at the University of Toronto.
3. To design a 2-center pilot feasibility RCT of sequential compression devices (SCD) versus no thromboprophylaxis in thrombocytopenic HM/HCT patients admitted to ICU.

Regarding the second objective, the investigators will perform a retrospective analysis of VTE incidence in HM/HCT patients at University of Toronto.

Inclusion Criteria:

• All patients 18 years and older admitted to an adult intensive care unit during the current admission episode and that also had an active HM / HCT diagnosis.

Exclusion Criteria:

• Patients not admitted to the ICU, even though they had a diagnosis of HM / HCT.

All patients 18 years and older admitted to an adult intensive care unit during the current admission episode and that also had an active HM / HCT diagnosis.

No individual patient information will be included in the analysis. All data will be anonymized.