Venous thromboembolism (VTE) is a common complication of malignancies, in particular to lung cancer. Patients with lung cancer in surgical and medical departments are at high risk of VTE development. Prophylaxis is one major way to to prevent it. Currently, VTE prophylaxis is mainly based on VTE-risk assessment. However, all patients hospitalized for cancer are at intermediate or high risk of VTE but their bleeding risk vary. To improve effect of VTE prophylaxis and reduce bleeding events in patients with lung cancer, we will conduct an open-label parallel randomized clinical tria to assess the effect of bleeding risk based prophylaxis strategy among lung cancer patients. We hypothesize that VTE prophylaxis based on bleeding risk assessment with a short post-discharge treatment course is superior to VTE propohylaxis based on VTE risk assessment among hospitalized patients with lung cancer

A sample of 3200 eligible patients will be randomized into experimental or control group with an allocation rate of 1:1. Stratified by medical/surgical units, block randomization with a varying block size of 4 or 6 will be adopted to randomize patients into experimental or control group. In experimental group, patients will undergo bleeding risk assessment and receive prophylaxis according to bleeding risk during hospitalization, and they will also receive an extended pharmacological prophylaxis of 5mg Rivaroxaban once daily for up to 15 consecutive days after discharge. In control group, patients will receive routine VTE prophylaxis, VTE risk assessment and prophylaxis if indicated during hospitalization according to current policies for hospitals in China but no further treatment prophylaxis after discharge.

Patients in both groups will be followed up for 30 days. The primary outcome is symptomatic and asymptomatic objectively proven VTE (deep vein thrombosis (DVT) and/or pulmonary embolism (PE)) within 30 days after initiation of randomization. Ultrasound and CTPA will be performed to detect DVT and PE, respectively. Clinically relevant bleeding (non-major clinically relevant and major bleeding, HIT) and death are secondary outcomes.

Randomization and sequence generation A computerized random-number generator will be used to generate the allocation sequence. In this multicenter trial involving 10 hospitals, randomization procedures will be organized centrally.

Stratified block randomization with a varying block size of 4 or 6 will be used to allocate patients into experimental or control group. Patients with lung cancer will be stratified into those under planned medical or surgical treatments.

In each stratum, patients will be blocked according to their admission sequence. Four or six patients consecutively admitted will be one block depending on the block size. In each block, patients will be randomly allocated into experimental or control group according to sequence generated in advance by software.

Allocation concealment/Blinded randomization Patient assignments will be enclosed in a sequentially numbered, opaque, sealed envelopes (SNOSE). Clinicians in charge of patient enrollment will not know the allocation sequence until eligible patients who meet inclusion and exclusion criteria are enrolled.

An independent statistician will generate the random allocation sequence. Physicians will enroll participants and assign interventions in experimental or control group.

Blinding/Open label This is an open-label trial that patients, clinicians and researchers will know allocation assignments after enrollment. But imaging experts providing the duplex ultrasound and CTPA results will be blinded in order to objectively assess the 30-day CTPA-proven VTE incidence and other outcomes in both groups. An independent data monitoring board will evaluate the trial data and safety.

Patients with primary lung cancer admitted to medical units (for chemotherapy, complications, etc) or to surgical units for operations.

Inclusion criteria

1. Aged ≥40 years
2. Have an expected hospital stay ≥72 hours for medical and/or surgical treatment
3. Confirmed lung cancer at admission or proven lung cancer within prior 6 months
4. Evidence of active lung cancer within 6 prior months
5. Written informed consent

Exclusion criteria Patient-related criteria

1. Pregnancy or breastfeeding
2. Inability to be followed-up at until 3 months after randomization
3. have participated in similar trials or are undergoing other clinical trials
4. refuse or are unable to give informed consent VTE/bleeding-related criteria
5. Incidental VTE identified on spiral CT scans which are ordered primarily for staging the malignancy at or any time before enrollment
6. Neurosurgery, vascular procedures, orthopedic surgery intended during the admission
7. Severe renal failure not receiving dialysis (creatinine clearance [CrCl] <30 mL/min), moderate to severe liver dysfunction, severe anemia
8. Uncontrolled hypertension (systolic blood pressure [BP] >180 mmHg, diastolic BP >110 mmHg)
9. severe platelet dysfunction or inherited bleeding disorder (such as haemophilia and von Willebrand's disease)
10. Acute stroke or recent stroke (within 4 weeks)
11. Recent major bleeding (within 3 months)
12. Requiring a full dose of anticoagulant treatment (e.g., recent VTE, atrial fibrillation)
13. Contraindication to heparin or rivaroxaban, e.g., heparin induced thrombocytopenia or history of documented episode of heparin or LMWH induced thrombocytopenia and/or thrombosis (HIT, HAT, or HITTS); recent central nervous system (CNS) bleed, hemorrhagic CNS metastases; active major bleeding with more than 2 units transfused in 24 hours.
14. Contraindication to mechanical prophylaxis, e.g., acute deep vein thrombosis, severe arterial insufficiency (pertains to graduated compression stockings only)
15. Concurrent use of anticoagulants known to increase the risk of bleeding (such as warfarin with INR>2).