Archives

  • 2019-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2019-11
  • 2020-03
  • 2020-07
  • 2020-08
  • 2021-03
  • br We did not make any assumptions regarding the mortality

    2020-08-12

    
    We did not make any assumptions regarding the mortality Doxorubicinol of mammography screening; instead, we predicted survival times using a Cox regression model that included age, tumor characteristics at diagnosis, and treatment. Covariate co-efficients on age, primary tumor size, and number of nodes at diagnosis in the baseline predictive survival model were esti-mated on the basis of a combined analysis of four breast cancer trials in the Cancer and Leukemia Group B cancer research cooperative group [32e35]. We then used the hazard reductions to model the additional effects on survival of treatment with tamoxifen for patients with ER- or PR-positive tumors [36] and with trastuzumab for patients with HER2-positive tumors; the r> magnitude of hazard reduction was 0.67 for hormonal therapies and 0.66 for trastuzumab [36,37]. If a woman’s estimated breast cancer survival time was shorter than her simulated natural lifetime according to life tables for the 1960 birth cohort from the US Census Bureau, she would be assumed to have died of breast cancer; otherwise, she would die from competing risks.
    Costs-Effectiveness Analysis
    We integrated cost information to the aforementioned micro-simulation model to assess the cost-effectiveness of various breast cancer screening strategies (see the Supplemental Materials for technical details). We performed cost-effectiveness analysis (CEA) from the societal perspective, as recommended by the US Public Health Panel on Cost-Effectiveness in Health and Medicine [38] and more recent good research practices guidelines [39e41]. For each woman, the model calculated cumulative medical costs related to breast cancer screening and treatment, estimated in-direct costs in the form of productivity loss from premature death caused by breast cancer, and computed quality-adjusted life-years (QALYs) throughout her lifetime. We applied an annual discount rate of 3% to both costs and QALYs. We compared a total of 10 mammography screening strategies, including no screening and 3 cessation ages (75 years, 80 years, and no upper age limit), each associated with three guideline recommendations: annual screening interval starting at the age of 40 years (ACOG, ACR, and previous ACS guidelines), biennial screening interval starting at the age of 50 years (updated USPSTF, AAFP, and ACP guidelines), and a hybrid strategy that begins with annual screening at the age of 45 years and transitions to biennial screening at the age of 55 years (updated ACS guideline).
    Medical costs captured in our model spanned from screening with digital mammography and workup procedures to breast cancer treatments and end-of-life care. Cost parameters are pre-sented in Tables 1 and 2. Specifically, we obtained costs of screening and workup from the 2017 Physician Fee Schedule [42] and the literature [43]. We analyzed the SEER-Medicare data, covering SEER up to 2011 and Medicare up to 2013, to obtain medical costs associated with real-world treatment patterns for the incident cohort of patients with breast cancer. To obtain medical cost by treatment modality, we determined from Medi-care claims whether patients had received surgery (mastectomy, lumpectomy, or none), radiation (yes/no), or chemotherapy (yes/ no) within 12 months of diagnosis, and stratified by stage at diagnosis. We obtained treatment pattern by cancer stage and age group (age <70 years vs. 70 years) from the National Cancer Database Public Benchmark Reports [44]. Following the phase of care framework, we classified cancer care into three phases: initial, continuing, and terminal care [45]. We then applied the incremental costing approach to estimate breast cancererelated costs for each care phase using the difference in mean total medical costs between patients with breast cancer and control cohorts consisting of age- and race-matched women without cancer [46,47]. Treatment patterns (by disease stage and patient age group) and costs (by stage and phase of care) for breast cancer are presented in Table 2.
    Because costs estimated from SEER-Medicare data were representative of only patients aged 65 years or older, we followed the estimation strategy in a cost report of the National Cancer Institute [45] and adjusted the SEER-Medicare estimates by a fac-tor of 1.5 for costs in terminal care phases for patients younger than 65 years to reflect more aggressive care among young women. Furthermore, we added costs of 5-year treatment with tamoxifen [48] to 70% of patients who had ER-positive tumors [49] and costs of trastuzumab for HER2-positive Doxorubicinol tumors [50]. We esti-mated indirect costs (i.e., mortality costs from lost wages) using 
    age-specific wage rates for female workers in the US labor market.