IN-DEPTH REVIEW OF RESEARCH
Study background

One provision of the Affordable Care Act was to provide funds to expand Medicaid coverage for those who were otherwise uninsured (providing healthcare coverage for nonelderly adults whose incomes were at or below 138 percent of the federal poverty level. (This coverage applies to single individuals making under $17,775 per year or a family of four making under $36,570 per year.)

Although funds were available for all states, each state decided whether or not to expand Medicaid. Nineteen states chose not to expand Medicaid. The remaining 31 states expanded Medicaid: six states did so early (before Jan 2014), 19 states and the District of Columbia expanded the program in January 2014 and 7 states expanded late (various times after January 2014).

What impact did expanding Medicaid funding have on cancer outcomes?

Prior research linked the ACA Medicaid expansion to:

• fewer uninsured patients
• more overall cancer screenings
• cancer diagnoses at earlier stages
• more colorectal cancer screening among low-income and Black individuals
• a smaller gap in hospital access for those with lower compared to higher income
• a smaller gap between racial groups in access to high-volume hospitals for cancer surgery
• a larger gap between racial groups in access to high-quality hospitals for cancer surgery

With the changes in cancer screening and access, the study authors wondered whether a change in cancer mortality was due to the ACA expansion. On one hand, earlier detection, stage of diagnosis and improved access might lead to decreased mortality. Alternatively, increased patients may lead to overloaded healthcare systems and delays in cancer treatment that could lead to higher mortality rates.

To test these ideas, the study researchers compared mortality in expansion states to mortality in non-expansion states before and after the ACA expansion, and also within the expansion states before and after the expansion.

Researchers of this study wanted to know:

This study asked whether cancer deaths from breast, colorectal or lung cancer were reduced after Medicaid coverage was expanded under the Patient Protection and Affordable Care Act:

• Did patients die of their cancer more or less often in states that expanded Medicaid compared with states that did not expand Medicaid?
• Could detection of cancer at earlier stages account for any changes in cancer deaths?
• Were any changes in cancer deaths seen at similar rates among at-risk populations (i.e., patients living in the lowest-income neighborhoods or among Black vs White patients)?         

Populations looked at in this study

Researchers looked at information from the National Cancer Database (NCDB) for patients who were diagnosed with their first breast, colorectal or lung cancer between Jan 1, 2012 and Dec 31, 2015. They chose these cancers because they are common, easily screened and are curable in non-metastatic patients.

Patients were included in the final analysis if they:

• had breast, colorectal or lung cancer.
• were newly diagnosed with their first cancer.
• were between ages 40 and 65 (ages that are eligible for Medicaid) and
• had records that provided information on cancer stage.
• lived in rural or urban location.
• had insurance information.
• it was known whether they came from an expansion state or non-expansion state.

Expansion states:

24 states and the District of Columbia expanded their Medicaid programs by January 1, 2014:

• States with early expansion before Jan 2014 included CA, CT, DC, MN, NJ and WA.
• States with expansion on Jan 2014 included AR, AZ, CO, DE, HI, IA, IL, KY, MA, MD, ND, NM, NV, NY, OH, OR, RI, WV and VT.

Non-expansion states:

19 states did not expand their Medicaid coverage under the ACA, including AL, FL, GA, ID, KS, ME, MO, MS, NC, NE, OK, WI, SC, SD, TN, TX, WY, UT and VA.

Late expansion states:

7 states expanded Medicaid later (after January 2014). Patients from these "late expansion states" were excluded from the study because of the uncertainty about when expansion occurred relative to diagnosis and the lack of time for cancer outcomes to occur.

Patients were excluded if their records were missing data about expansion status, insurance status or cancer stage or they were under age 40.

Information about 523,802 cancer patients was included in the final analysis:

• 234,472 patients (44.8%) were from expansion states.
• 289,330 patients (55.2%) were from non-expansion states.
   
• Patients had an average age of 55 and most were women (385,739 or 74%).
  • 273,272 (52.2%) had breast cancer.
  • 111,720 (21.3%) had colorectal cancer.
  • 138,810 (26.5%) had lung cancer.

There were some differences in racial groups between the expansion and non-expansion states:

The percentage of patients with different stages of cancer was similar between expansion and non-expansion states:

Study design

This effort was a population-based study that looked at a cross-section of the population. The researchers compared data from cancer patients in expansion states to those in non-expansion states at two time periods:

• The pre-expansion period: from January 2012 to December 2013, the two years before the expansion of Medicaid.
• The post-expansion period: from January 2014 to December 2015, the two years after the expansion of Medicaid.

The researchers compared the mortality (months from cancer diagnosis to death) of one group— the patients in expansion states before and after January 2014—to mortality of another group—the patients in non-expansion states before and after Jan 2014. This type of study is called a difference-in-difference (DID) study.

The researchers chose to look at breast, colorectal and lung cancers because these cancers have standard screening and treatments for cure in non-metastatic patients. The researchers reasoned that they might detect changes that may have occurred in mortality among this group of patients. Data analysis was performed from January to May 2020.

Study findings

Patients in expansion states had lower cancer mortality in this study

• Fewer cancer patients in expansion states died from cancer compared to non-expansion states:
  • Cancers deaths were unchanged in non-expansion states during these time periods.
  • Cancer deaths were 2% lower in expansion states after the Medicaid expansion than before the Medicaid expansion.
  • Cancer deaths were lower in expansion states than in non-expansion states after the Medicaid expansion.
• This data suggests that nearly 1,400 fewer deaths occur each year if expansion states fully broaden their Medicaid coverage.

What factors contributed to the lower mortality rates in expansion states? Researchers considered whether differences in the mortality rates between the expansion states and non-expansion states were due to differences in social support, social determinants of health or limited health resources.

• The stage of cancer diagnosis was the main factor linked to decreased cancer mortality between expansion and non-expansion states.
  • In non-expansion states, when patients were diagnosed with stage 1-3 cancer, they had a 4.8% greater risk of dying from their cancer when comparing pre-expansion and post-expansion periods.
  • There was no increase in the risk of death for patients with stage 1-3 cancer in expansion states in pre- versus post-expansion. In other words, the expansion of Medicaid coverage seems to have benefited early-stage cancer patients.
     
  • For patients first diagnosed with stage 4 (metastatic) cancer, both expansion and non-expansion states had fewer cancer deaths between the pre-expansion and post-expansion time periods, but the differences were not significant.
     

Models indicate that stage of cancer diagnosis may account for nearly all of the changes in mortality improvements (expansion versus non-expansion states and pre-expansion versus post-expansion periods).

• Factors that did not affect cancer mortality in this study included:
  • age
  • sex
  • cancer type (breast, colorectal or lung)
  • location (urban/non-urban)
  • whether patients had other diseases (comorbidities)
     
• Cancer mortality was similar regardless of education, income, insurance status and race. Mortality improvements in expansion states were seen in all of the sub-groups that were studied, regardless of racial and ethnic groups, education level, income, state or insurance status.
  • Patients in geographic areas with the lowest median household incomes (lowest 25%) within expansion states showed a modest decrease in mortality after expansion. This suggests that improvements in cancer mortality were not limited to those with higher incomes.
  • Mortality improvements associated with Medicaid expansion were similar for Black and White patients.
     

The authors suggest that: "Increased Medicaid coverage may remove barriers to accessing the healthcare system for screening and timely symptom evaluation..."

Strengths and limitations

Strengths

• This study looks at a large number of patients, allowing the authors to analyze subgroups within the population.

• Individual states chose to expand or not to expand Medicaid, providing a natural comparison population for this observational study.

Limitations

• A major limitation of this study is that it was observational. Observational studies are prone to systematic bias and factors that have significant but unrecognized effects (confounding factors).  Conclusions of observational studies are only correlations (showing that two factors or findings are linked). Observational studies cannot show that one thing causes another.
• The patient information available was limited. Patient data on individuals' state of residence and eligibility for Medicaid were not available.
• Only patients who were 40 to 65 years old were included. These findings may not reflect those of younger ages.
• Patients in late expansion states were not included. These findings may not reflect people in those healthcare systems.
• This study looked at mortality (time from diagnosis to death). This could bias the results if increased screening detects more slow-growing than fast-growing tumors.
• The window of time for the analysis was fairly small: two years before and after the Medicaid expansion. This does not take into account cancer deaths beyond four  that are longer than 4 years from diagnosis. Longer follow-up of these patients would clarify this information.

Context

This study adds to our knowledge about Medicaid expansion and cancer care. A few studies have shown that among patients with cancer, expansion was associated with increased insurance coverage, an earlier stage of diagnosis, and no changes in the timeliness of treatment. This is the first study to look at mortality linked to cancer.

Previous research on a Medicaid expansion in 2000 and the Massachusetts health reform in 2006 similarly showed lower death rates after those expansions in healthcare coverage.

• In the 2000 Medicaid expansion study, researchers estimated that 1 death was prevented for every 239 to 316 adults who gained coverage.
• In the Massachusetts 2006 health reform study, researchers found that 1 death per year was prevented for 830 adults who gained coverage.
• In this study, the authors estimated that 1 death would be prevented for every 250 patients with cancer who gain coverage within 4 years after their cancer diagnosis.

Conclusions

States that expanded Medicaid access under the Affordable Care Act had fewer deaths from breast, colorectal or lung cancer than before the expansion and also had fewer deaths than states that elected not to expand Medicaid. The change in mortality (mostly among early-stage cancer patients) suggests that increased financial access may have led to earlier diagnosis and fewer deaths.

Share your thoughts on this XRAY article by taking our brief survey.
3/31/2021