Higher Incidence of Venous Thromboembolism in the Outpatient Versus the Inpatient Setting Among U.S. Cancer Patients

Paradigm changes in cancer therapy have shifted care to primarily outpatient-based regimens. Venous thromboembolism (VTE) is a well-known complication of cancer but contemporary data regarding the burden of VTE in the outpatient versus inpatient cancer settings are lacking.

We conducted a retrospective, observational, cohort study to examine the incidence of VTE in inpatients and outpatients with cancer utilizing a linked database formed from the US Premier Perspective™ and the I3 Pharma Informatics healthcare claims databases, which provide data from the inpatient and outpatient settings, respectively. Patients with ≥1 inpatient or outpatient claims containing a diagnosis of cancer in any position (ICD-9-CM codes: 140.XX-208.99) from January 2005 to June 2009 were identified. VTE including deep vein thrombosis (DVT: ICD-9-CM codes 451 and 453) or pulmonary embolism (PE: ICD-9-CM codes 415.1–415.19) occurring in the inpatient or outpatient setting were identified over the 12 months post-index period. Data regarding demographics, clinical characteristics and cost were assessed. Multivariate analyses were conducted to adjust for differences in patient characteristics before and after the index event.

A total of 17,874 patients with cancer were identified. Over the 12 months post-index period, 996 (5.6%) of these patients had VTE (mean age 59.6 years; 46.7% male) and 16878 (94.4%) patients did not (mean age 58.3 years; 46.3% male). Patients with VTE had a higher mean Charlson Comorbidity Index (CCI) score at baseline than the non-VTE patients (0.9 vs. 0.7; P<0.001, respectively). Of patients with VTE, 731 (73.4%) had only DVT and 215 (21.6%) PE, while 50 (5.0%) had both. In multivariate analysis, predictors of VTE included primary site of cancer, particularly (odds ratio [OR], 95% confidence interval [CI]): stomach (2.24, 1.20–4.17; P<0.05); pancreas (2.29, 1.24–4.24; P<0.01); brain (1.82, 1.06–3.10; P<0.05); and testicular (3.06, 1.42–6.59; P<0.01). Increasing age (1.01, 1.00–1.02, P=0.001); increasing CCI score category (0.87, 0.72–1.06; P<0.05); history of pulmonary disease (1.26, 1.05–1.51; P<0.05); and post index use of doxorubicin (1.59, 1.15–2.21; P<0.01) were also significant predictors of VTE. A much higher proportion of VTE was diagnosed in the outpatient than the inpatient setting (78.3% versus 21.7%, P<0.0001). Of patients with outpatient VTE, 167 (21.4%) had a hospitalization within 30 days before their VTE index event. In multivariate logistic regression analysis, VTE was an independent predictor of hospitalization (OR 2.31, CI 1.92–2.78; P<0.0001). VTE was also an independent predictor of higher hospital costs (P<0.0001).Total mean (SD) annual hospital costs were twofold elevated for patients with VTE in comparison with those who did not have VTE $22,917 (SD $14,005) versus $11,250 (SD $6,592) respectively (P<0.0001) after adjusting for patient and treatment characteristics

Over three-quarters of all VTE in cancer occurs in the outpatient setting and is associated with hospitalization and increased costs. One-fifth of outpatients with VTE however also had a recent hospitalization. Public health efforts to reduce the burden of VTE in cancer will need to focus on outpatient and post-discharge thromboprophylaxis in select high-risk patients, in addition to ongoing efforts to improve compliance with inpatient prophylaxis.

Khorana:Roche/Genentech: Consultancy, Honoraria; Eisai Inc.: Honoraria, Research Funding; Ortho Biotech: Honoraria, Research Funding; Boehringer-Ingelheim: Consultancy, Honoraria; Leo Pharma: Research Funding; sanofi-aventis: Consultancy, Honoraria, Research Funding; Bayer: Consultancy, Honoraria; Daiichi Sankyo: Consultancy, Honoraria. Dalal:sanofi-aventis U.S.: Employment. Tangirala:sanofi-aventis U.S.: Employment. Miao:sanofi-aventis U.S.: Employment.