Archives

  • 2019-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2020-03
  • 2020-07
  • 2020-08
  • 2021-03
  • br Prior studies reported inconsistent

    2020-07-04


    Prior studies reported inconsistent effects of ACA implementation on cancer stage at diagnosis (Han et al., 2018; Han et al., 2016; Jemal et al., 2017; Robbins et al., 2015; Sabik and Adunlin, 2017; Smith and Fader, 2018; Soni et al., 2018). Studies that reported increased early stage diagnosis primarily focused on young adults and assessed early ACA provisions (e.g. dependent coverage expansion) (Han et al., 2016;
    Fig. 2. The effect of Affordable Care Act implementation on insurance status among patients diagnosed with first primary cancers in an underserved population,
    Tarrant County, and Texas overall.
    A. Underserved cancer population
    B. Tarrant County cancer population
    C. Texas cancer population.
    Fig. 3. The effect of Affordable Care Act implementation on stage at diagnosis among patients diagnosed with first primary cancers in an underserved population, Tarrant County, and Texas overall.
    A. Prevalence of early stage among underserved cancer patients
    B. Prevalence of early stage among cancer patients in Tarrant County
    C. Prevalence of early stage among cancer patients in Texas overall
    D. Prevalence of advanced stage among underserved cancer patients
    E. Prevalence of advanced stage among cancer patients in Tarrant County
    F. Prevalence of advanced stage among cancer patients in Texas overall.
    Robbins et al., 2015; Smith and Fader, 2018). We identified three stu-dies that assessed the effect of ACA implementation on cancer stage at diagnosis which reported an earlier stage of cancer diagnosis after ACA Methoxy-X04 among Medicaid expansion states, but these studies sug-gested smaller or no impact on cancer stage at diagnosis in states without Medicaid expansion (Han et al., 2018; Jemal et al., 2017; Soni et al., 2018). Medicaid enrollment reportedly increases the number of individuals with a usual source of care (Winkelman et al., 2018), which can increase the use of preventive services (Blewett et al., 2008) and earlier detection of cancer. States that did not expand Medicaid may not be able to influence this cascade (Griffith et al., 2017). In particular, Texas residents did not increase the use of preventive or other health-care services as much compared with other southern states that ex-panded Medicaid during the same time period (Sommers et al., 2016). We thus speculate that the lack of Medicaid expansion in Texas may partially explain the lack of improvement in earlier cancer detection in our population. An even greater reduction in uninsured individuals, which may only be feasible by large-scale policy changes such as Medicaid expansion, or longer follow-up duration may be necessary to observe effects on cancer stage at diagnosis. In addition, underserved populations have barriers to care beyond insurance such as inconsistent transportation and low health literacy (Allen et al., 2008; Chase et al., 
    2012; Escriba-Aguir et al., 2016; Fernandez and Becker, 2017). Con-sequently, increased availability of self-purchased insurance may not sufficiently overcome other barriers to earlier diagnosis.
    Several limitations should be considered when interpreting our findings. Our analysis cannot be used to infer whether ACA had an effect on cancer incidence (e.g. through detecting previously un-diagnosed cases or through screening that detected pre-malignant le-sions and prevented malignancy). Rather, our analysis addresses whe-ther ACA affected stage distribution among individuals diagnosed with cancer. Underserved individuals who acquired insurance coverage be-cause of ACA could have chosen other hospitals, and thus would not be represented in our post-ACA population. This phenomenon could manifest as a selective loss to follow-up, which could underestimate the effect of the ACA in our population. In addition, the follow-up duration for our post-ACA period is limited, which is relevant for all studies to date that assessed the effect of ACA implementation. The short post-ACA follow-up period would preclude detection of lagged effects on cancer stage at diagnosis. Lastly, interrupted time-series analysis are generally less sensitive to confounding by population characteristics than conventional analyses (Bernal et al., 2017), but other population-level interventions during this period could result in unmeasured con-founding of the observed effects. For example, low-income, uninsured,
    or under-insured women in Texas are eligible for low-cost breast and cervical cancer screening through the Breast and Cervical Cancer Ser-vices offered by Texas Department of State Health Services. Individuals in Tarrant County are eligible for free or low-cost cancer screenings through Tarrant County Indigent Health Care Program, Federally Qualified Health Centers, Moncrief Cancer Resources, or American Cancer Society North Texas Region (Texas Cancer Information). In addition, 42% of cancer patients in our hospital were enrolled in a hospital-based assistance program, which is offered to patients as an insurance supplement or as a form of primary coverage for individuals who meet certain eligibility criteria (i.e. a Tarrant County resident who has taken full advantage of state or federal assistance programs and has an annual income ≤250% of the federal poverty limit) (John Peter Smith Health Network). These programs were operational since the pre-ACA period and could have mitigated the effect of ACA implementa-tion. Nevertheless, the lack of stage migration was not limited to un-derserved cancer patients. Cancer patients in Tarrant County and Texas overall also had a lack of stage migration after ACA implementation, which suggests that competing hospital-based or other local interven-tions may not be an alternate explanation for our results.