1384P - Are there differences in patient characteristics and treatment patterns by treatment setting?

Date 30 September 2012
Event ESMO Congress 2012
Session Poster presentation II
Topics Bioethics, Legal, and Economic Issues
Presenter Carolina Reyes
Authors C. Reyes1, S. Dacosta Byfield2, A. Small3
  • 1Health Outcomes, Genentech, San Francisco/US
  • 2Heor - Observational Reserach, OptumInsight, 55344 - Eden Prairie/US
  • 3Oncology, Outcomes Research, Us Meidcal Affairs, Genentech, San Francisco/US



Few studies have examined whether differences in treatment and outcomes exist among cancer patients by the setting where care is delivered. This study investigates differences in demographics, treatment patterns and health care resource use among non-Hodgkin's Lymphoma (NHL) and Chronic Lymphocytic Leukemia (CLL) patients receiving rituximab (R) or R+ chemotherapy based on site of care: office/clinic (OC) vs. hospital outpatient (HO).


Patients ≥18 years with evidence of NHL or CLL diagnoses codes at least 30 days apart and received ≥2 R claims from Jan 2007 to Mar 2011 were identified from a large US commercial insurance claims database. Patients were required to be enrolled in the health plan for at least 6 months before and after the index date (date of first R claim). The follow-up period was the date of the first infusion to 30 days after the last infusion prior to a gap of ≥7 months. Patients with evidence of multiple cancers or receipt of R at both sites of care were excluded. Cohorts were created based on site of care where R was administered. Multivariate analyses examined differences in number of infusions, ER visits and inpatient stays by cohort.


There were 2,594 OC and 286 HO patients with a mean follow-up of 242 days and 209 days, respectively. Compared to the HO cohort, the OC patients were significantly younger (71yrs vs. 63 yrs), had a lower mean baseline Charlson comorbidity index (3.88 vs. 3.40) and a lower percentage were Medicare Advantage enrollees (83% vs. 25%) [each p < 0.01]. The mean R infusion count was 7.4 (OC) vs. 5.4 (HO) with a monthly mean of 1.19 (OC) vs. 1.00 (HO) [each p < 0.01]. In adjusted analyses, R infusion counts were also lower for HO compared to the OC cohort (IRR 0.80, CI 0.75-0.86) and HO patients had higher number of inpatient stays (IRR 1.40, CI 1.07-1.84) though ER visits were not different between cohorts.


Patients treated in the OC setting compared to the HO setting had different treatment patterns and fewer inpatient stays. These results warrant further investigation to assess whether clinical outcomes differ by site of care.


C. Reyes: Has Roche stock and employment at Genentech/Roche.

S. Dacosta Byfield: Employed at OptumInsight, the entity that was paid by Genentech to conduct the study.

A. Small: Has Roche stock and employment at Genentech/Roche.