Defining the Risk: Benefit Ratio of Venous Thromboembolism (VTE) Prophylaxis in Hospitalized Cancer Patients

Introduction: VTE affects 1.6-1.8/1000 hospitalized patients per year; active cancer increases the rate of VTE 6-fold.  Despite the high risk of VTE, studies show cancer patients receive inpatient VTE prophylaxis at a lower rate than general medical patients. In addition, VTE prophylaxis is often held for a platelet value of <50x10⁹/L, despite a lack of evidence.  Current recommendations for all inpatients with cancer (and no contraindication to blood thinners) are to receive prophylactic anticoagulation, regardless of an increased bleeding tendency in this population. A formal risk:benefit ratio of inpatient VTE prophylaxis for cancer patients has not been evaluated to date.    

Methods: The Michigan Hospital Medicine Safety Consortium, a 49 hospital quality collaborative, has prospectively collected data on VTE risk factors and outcomes in medical patients > 18 years of age.  Exclusion criteria include surgical patients, pregnancy, admission to the ICU or for palliative care, therapeutic anticoagulation, diagnosis of acute thrombus, history of VTE within 6 months, and length of stay < 2 days.  We compared rates of VTE prophylaxis, bleeding and new VTE between cancer and general medicine patients who were eligible for prophylaxis (i.e. no contraindications including active bleed within 3 months, coagulopathy, or high risk brain metastasis). Student’s t-test was used for continuous variables and chi-square for categorical data. Logistic regression was used to calculate odds ratio (OR). The number needed to treat and number needed to harm were used to derive a risk:benefit ratio.

Results:  Between 7/2012-7/2015, 86,634 admissions were captured in the cohort; 70,086 were eligible for VTE prophylaxis and included in this analysis. 22% of cases had a diagnosis of cancer.

Table 1: Demographics on Admission

	Cancer  (n= 15,166), %	General Medicine (n=54,920), %	p
Race (Caucasian)	82.3	74.3	<0.001
Age mean yrs (SD)	72.5 (14.4)	63.7 (18.8)	<0.001
Gender (female)	53.6	56.5	<0.001
Central Line	17.8	4.1	<0.001
History of VTE	7.6	5.3	<0.001
Caprini risk score ≥ 5	88.1	35.8	<0.001
Charleston Index mean (SD)	5.1 (2.5)	2.1 (2.0)	<0.001
Ever Smoked	59.6	56.3	<0.001
Obese	29.3	37.1	<0.001
Surgery w/in 30 days	3.6	2.1	<0.001
Trauma w/in 30 days	0.3	0.4	0.01

Of cancer admissions, 89.5% had solid tumors, 13.5% hematologic malignancies, 3.4% both and 20% metastatic disease.  Active treatment for cancer was delivered <6 months from index admission in 32.1%, within 6-12 mo in 3.2%, >12 mo in 44.4% and no treatment or unknown in 20.3%.

When compared to general medical admissions, cancer admissions were more likely to receive VTE prophylaxis (72.16% vs 69.21%, p<0.001), and have a new VTE out to 90 days post discharge despite prophylaxis (0.91% vs 0.45%, p<0.001, unadjusted OR 2.07 (95% CI 1.6-2.7)). There was no difference in VTE rate with regard to platelet count in cancer cases (0.43% for plt < 50 vs 1.08% for plt ≥ 50, p=0.10). Among all patients receiving VTE prophylaxis, bleeding was more common in cancer cases (major bleeding 0.84% vs 0.58%, p=0.005; minor bleeding 1.80% vs 1.36%, p=0.002). Among cancer cases, bleeding rates were higher in patients with platelet <50 vs ≥ 50 (major bleed 4.86% vs 1.88%, p<0.001; minor bleed 2.88% vs 1.7%, p=0.04).

Table 2: The Number Needed to Treat (NNT) to Prevent One VTE During Admission or 90 Days Post Discharge and Number Needed to Harm (NNH) to Cause One Major Bleed During Admission with Risk:Benefit Ratio (NNH:NNT)

General Medicine	Cancer
NNT 1428	NNT 1000
NNH 2500	NNH 277.9
NNH:NNT 1.75	NNH:NNT 0.28

Conclusions: In this prospective inpatient cohort, we compared general medicine to cancer cases and found cancer admissions received VTE prophylaxis at a higher rate. This is different than previously reported data, likely due to the exclusion of patients with contraindications to prophylaxis. However, despite prophylaxis, cancer patients had a higher rate of VTE during admission and 90 days post discharge as well as more bleeding complications. The risk:benefit ratio of VTE prophylaxis is 6 times worse in cancer patients due to bleeding. While bleeding occurs more frequently in cancer patients with platelet count <50x10⁹/L, VTE occurs at a similar rate regardless of platelet count, meaning that cancer patients are at high risk of both clotting and bleeding.  Recommendations for inpatient VTE prophylaxis for cancer patients require a targeted approach to identify a subset who would most benefit from VTE prophylaxis.