Health care utilization and costs by site of service for nonmetastatic breast cancer patients treated with trastuzumab.

Abstract

Adjuvant trastuzumab treatment is administered to early stage breast cancer patients in physician office (POV) or hospital outpatient (HOP) settings. To identify treatment patterns, utilization, and costs by site of care (POV vs. HOP) of patients with adjuvant treatment of breast cancer with trastuzumab. This retrospective analysis identified adult, female breast cancer patients who initiated trastuzumab treatment (index date) from a large U.S. claims database. Inclusion criteria also required ≥ 2 claims for both trastuzumab (from July 1, 2006, to July 31, 2012) and breast cancer (during 6-month pre-index baseline period), no evidence of metastatic breast cancer or other cancers in the baseline period, and continuous enrollment with commercial or Medicare Advantage coverage 6 months pre- and post-index, except that patients who died during follow-up were retained. Patients with evidence of trastuzumab receipt during the baseline period or more than 1 site of care during follow-up were excluded. Patients were stratified by site of care and were followed from index date to 30 days after the last trastuzumab infusion prior to a gap ≥ 90 days, death, disenrollment, or end-of-study period. Differences in treatment patterns between the POV and HOP cohorts were assessed by t-test and chi-square test. The relationship between site of care and health care costs was modeled with a generalized linear model with gamma distribution and log link, and the number of trastuzumab infusions was modeled with negative binomial regression controlling for log follow-up time. All models were adjusted for age, baseline comorbidity score, and insurance type. Of the 3,439 breast cancer patients identified, 77.6% (2,669) received adjuvant trastuzumab in the POV setting. Mean age (53.7 years) and baseline comorbidity score (3.91) were similar among cohorts; a higher percentage of POV versus HOP patients had commercial insurance (91.1% vs. 86.4%, P < 0.001). Compared with the POV cohort, HOP patients had a shorter mean duration of trastuzumab treatment (324.8 vs. 343.0 days, P < 0.001); more treatment gaps (30-59 day gap: 67.4% vs. 24.1%, P < 0.001); and fewer trastuzumab infusions per month (1.37 vs. 1.98, P < 0.001) during follow-up. In multivariate analysis, the monthly count of trastuzumab infusions in the HOP cohort was lower than the POV cohort (incidence rate ratio = 0.693; 95% CI = 0.672-0.715). Adjusted per patient per month total health care costs were 53.6 % higher in the HOP setting (cost ratio = 1.536, 95% CI = 1.472-1.604). Breast cancer patients treated with adjuvant trastuzumab in the HOP setting had a shorter duration of trastuzumab treatment and fewer trastuzumab infusions but incurred higher monthly total costs than patients treated in the POV setting.