Use of the NCDB to Examine Association of Health Insurance and Health Policies with Cancer Care

Insurance status was first included in the National Cancer Database (NCDB) for the 1998 diagnosis cohort. The first studies using data from the NCDB to examine associations with insurance status were published in 2007. These initial studies assessed the link between insurance status with cancer stage at diagnosis,^1-3 receipt of cancer treatment,⁴ and survival.⁵ These studies highlighted that among individuals diagnosed with certain cancers, those who were uninsured or had Medicaid coverage were more likely to be diagnosed with advanced-stage disease, received different treatments, and have decreased survival than were those with private insurance.

In 2008, a Lancet Oncology article presented broad analyses from the NCDB comparing stage at diagnosis for privately insurance vs uninsured or Medicaid-insured individuals diagnosed with 12 different types of cancer.⁶ Uninsured individuals were more likely to present with advanced (stage III/IV) cancer than were those with private insurance for 10 of the 12 cancer types examined while Medicaid-insured individuals were more likely to present with advanced cancer for 11 of the 12 cancer types examined. Findings were most prominent for patients who had cancers that could potentially be detected early by screening or symptom assessment (e.g., breast, colorectal, and lung cancer, or melanoma). A separate 2008 publication appearing in CA also used NCDB data to examine associations between insurance status and mortality for all individuals diagnosed with cancer.⁷ This study reported that patients who were uninsured or Medicaid-insured at the time of cancer diagnosis were 1.6 times as likely to die within five years of diagnosis compared with those having private insurance. Differences in survival by insurance status were consistently observed among Black, Hispanic, and White patients.

Following these initial studies, multiple published papers have used data from the NCDB to examine association of insurance status with cancer stage at diagnosis; time from diagnosis to treatment initiation or treatment completion; types of treatment received; treatment refusal; failure to receive recommended treatment; receipt of supportive/palliative care; and survival. The NCDB has also been used to examine the association of insurance status with hospital-based cancer care, including duration of hospital stay following surgery; type of hospital where surgery was performed; and likelihood of 30-day re-admission. While results from some of these studies have been mixed, the majority have demonstrated that individuals who are uninsured or have Medicaid coverage tend to be diagnosed with more advanced disease; have more limited access to high-quality cancer care, and experience worse outcomes than do those with private insurance coverage.

More recent studies have used data from the NCDB to explore potential effects of the Affordable Care Act (ACA), including the option for states to expand Medicaid coverage to adults with incomes below 138 percent of the Federal Poverty Level, on insurance coverage, stage at diagnosis, time to treatment initiation, type of treatment received, receipt of treatment at an academic facility, and survival among individuals with cancer. Results from these policy analyses have been mixed, but most found that the ACA was associated with increased insurance coverage for individuals diagnosed with cancer (particularly for those residing in states that expanded Medicaid); earlier stage at diagnosis; more rapid treatment initiation; and improved access to high-quality cancer care.

As there is now an extensive body of published literature that has well-demonstrated disparities based on insurance status in stage at diagnosis, receipt of treatment, and survival using the NCDB, innovative researchers are instead using the NCDB to explore potential strategies to address these disparities. For example, Wu and colleagues used the NCDB to examine whether guideline-concordant care could mitigate insurance disparities in survival for women with cervical cancer.⁸ Unfortunately, these researchers found that Medicaid and uninsured women who received guideline-concordant care were still at an increased risk of death compared to women with private insurance. 

Studies using data from the NCDB to examine the effects of insurance on cancer care can continue to have an important role in the development of health policy. These studies have shown critical gaps and unmet needs in care, highlighting areas where policies are needed to address disparities. Studies using the NCDB can evaluate the impacts of “natural experiments”, where two states, regions, or health care systems initiate different insurance-related policies. Similar to analyses examining outcomes in states that did vs. did not expand Medicaid following implementation of the ACA, the NCDB can be used to evaluate these policy differences, assess their impacts on patient care patterns and outcomes, and provide information for implementing new policies to improve cancer care. In addition, as new cancer treatments are approved, there will likely be variation in their insurance coverage. Studies using the NCDB to examine associations between these insurance coverage difference and the use or outcomes for new treatments will be important for policy development. Finally, gaps in studies using the NCDB also suggest policy needs. For example, there are almost no studies examining the association of insurance and receipt of supportive/palliative care using the NCDB; the study by Colibaseanu and colleagues is one of the only exceptions.⁹ This may suggest a need for policies to encourage delivery of palliative care services and collection of information on receipt of palliative care.

Going forward, the NCDB will likely incorporate new data elements to permit better assessment of the multi-disciplinary and patient-centered nature of oncology today. Researchers will continue to use this important data resource to explore strategies and assess policies related to health insurance, with a goal of enhance equity in cancer care.

DISCLAIMER: The views expressed here are those of the author and do not necessarily represent any official position of the National Cancer Institute or National Institutes of Health.

References

1. Halpern MT, Bian J, Ward EM, Schrag NM, Chen AY. Insurance status and stage of cancer at diagnosis among women with breast cancer. 2007;110(2):403-112007.
2. Chen AY, Schrag NM, Halpern MT, Stewart A, Ward EM. Health insurance and stage at diagnosis of laryngeal cancer: does insurance type predict stage at diagnosis? Arch Oto Head Neck Surg. 2007;133(8):784-90.
3. Chen AY, Schrag NM, Halpern MT, Ward EM. The impact of health insurance status on stage at diagnosis of oropharyngeal cancer. 2007;110(2):395-402.
4. Bilimoria KY, Bentrem DJ, Linn JG, et al. Utilization of total thyroidectomy for papillary thyroid cancer in the United States. Surgery. 2007 Dec;142(6):906-13; discussion 913.e1-2. doi: 10.1016/j.surg.2007.09.002. Epub 2007 Nov 5.
5. Chen AY, Halpern M. Factors predictive of survival in advanced laryngeal cancer. Arch Otolaryngol Head Neck Surg. 2007;133(12):1270-6.
6. Halpern MT, Ward EM, Pavluck AL, Schrag NM, Bian J, Chen AY. Association of insurance status and race/ethnicity with cancer stage at diagnosis: A retrospective analysis. Lancet Oncol. 2008;9(3):222-31.
7. Ward E, Halpern M, Schrag N, et al. Association of insurance with cancer care utilization and outcomes. CA Cancer J Clin. 2008;58(1):9-31.
8. Wu J, Huang Y, Tergas AI, et al. The effect of guideline-concordant care in mitigating insurance status disparities in cervical cancer. Gynecol Oncol. 2020 Nov;159(2):309-316.
9. Colibaseanu DT, Osagiede O, Spaulding AC, et al. The determinants of palliative care use in patients with colorectal cancer: A national study. Am J Hosp Palliat Care. 2018 Oct;35(10):1295-1303.