Abstract
This cohort study uses the National Cancer Database to examine changes in rates of diagnosis of thyroid cancer among Medicaid beneficiaries after Medicaid expansion in 2014.
Medicaid expansion has increased cancer screening, leading to increased early-stage cancer diagnosis in patients with Medicaid.1,2,3 However, while studies exist for many common tumors, the association of Medicaid expansion with thyroid cancer diagnosis appears to be unknown and is important to consider, given rising concerns over overdiagnosis of small tumors with broader screening.4 To determine the outcome of Medicaid expansion, this study analyzed associations between insurance coverage, thyroid cancer diagnosis rates, and tumor size before and after Medicaid expansion.
Methods
We identified patients with new diagnoses of well-differentiated (papillary and follicular) thyroid cancer in the National Cancer Database between January 2010 and December 2016. We estimated state populations using US census data from 2010 through 2016. Medicaid enrollee data were obtained from the Medicaid Analytic Extract from January 2010 to December 2012 and the Center for Medicare and Medicaid Services Medicaid Budget and Expenditure System database from January 2014 to December 2016 (the years with available data). This study was deemed exempt from institutional review board review and informed consent procedures by the University of Michigan because all data were deidentified.
Thyroid cancer incidence was calculated from cases in the National Cancer Database, assuming 70% representation.5 To determine change in incidence among Medicaid enrollees, we measured incidence among the Medicaid and non-Medicaid populations of included states. To quantify the difference, the proportion of patients with Medicaid and thyroid cancer who were younger than 65 years, adjusting for sex, race/ethnicity (as found in the National Cancer Database: White, Black, Asian American, Native American, and Hispanic), and local tumor (T) classification, was compared between January 2014 expansion and nonexpansion states using a difference-in-difference analysis. The data were analyzed from May 2019 to September 2019. All data were analyzed in Stata 15.1 (StataCorp). Significance was set at P < .05, 2 tailed.
Results
We identified 246 296 patients with well-differentiated thyroid cancer, of whom 299 168 were women (75.9%) and 205 296 were White (83.5%). The mean incidence of thyroid cancer in nonexpansion states was 10.4 per 100 000 members of the total population in 2010 to 2012 and 10.7 per 100 000 from 2014 to 2016. In nonexpansion states, the unadjusted mean incidence among Medicaid enrollees increased from 1.6 per 100 000 to 2.0 per 100 000, without significant change among patients who did not have Medicaid benefits. In Medicaid expansion states, incidence among Medicaid enrollees increased from 3.2 per 100 000 to 5.6 per 100 000, with a concurrent small increase among the non-Medicaid population (Table).
Table. Unadjusted Incidence of Thyroid Cancer by Patient Insurance and State Expansion Statusa.
| State | Incidence per 100 000 population | P value | |
|---|---|---|---|
| 2010-2012 | 2014-2016 | ||
| Overall | |||
| Never expanded | 12.3 | 12.3 | .72 |
| Expanded January 2014 | 15.9 | 16.3 | <.001 |
| Medicaid | |||
| Never expanded | 1.6 | 2.0 | <.001 |
| Expanded January 2014 | 3.2 | 5.6 | <.001 |
| Non-Medicaid | |||
| Never expanded | 12.7 | 12.7 | .18 |
| Expanded January 2014 | 16.7 | 17.5 | <.001 |
P values were calculated as paired t tests comparing pre-2014 and post-2014 unadjusted incidences.
The adjusted proportion of patients with thyroid cancer enrolled in Medicaid in nonexpansion states was 4.4% (95% CI, 4.1%-4.8%) in 2010 through 2013 and 4.5% (95% CI, 4.1%-4.9%) in 2014 through 2016. In expansion states, the adjusted proportions were 7.3% (95% CI, 7.0%-7.7%) and 12.9% (95% CI, 12.5%-13.3%). Difference-in-difference analysis showed an associated absolute difference of 5.5% (95% CI, 4.8%-6.2%) between expansion and nonexpansion states (Figure, A). The adjusted proportion of T1 and T2 disease among Medicaid enrollees did not change significantly in nonexpansion states (from 19.2% [95% CI, 16.3%-22.1%] in 2010 through 2013 to 17.9% [95% CI, 14.5%-21.2%] in 2014 through 2016) or expansion states (from 15.6% [95% CI, 13.6%-17.8%] to 16.6% [95% CI, 14.7%-18.4%]). Difference-in-difference analysis showed no significant difference (2.1% [95% CI, −3.0% to 7.4%]; Figure, B).
Figure. Adjusted Proportions of Patients With Thyroid Cancer Enrolled in Medicaid and With T1 or T2 Disease by State Medicaid Expansion Status.
A, Adjusted proportion of patients with thyroid cancer enrolled in Medicaid by state Medicaid expansion status. B, Adjusted proportion of T1 and T2 disease among Medicaid enrollees with thyroid cancer by state Medicaid expansions status. Error bars represent 95% CIs. The dotted line represents the January 2014 expansion.
Discussion
The overall incidence of thyroid cancer increased more in Medicaid expansion states compared with nonexpansion states, with a disproportionate increase among Medicaid patients. The proportion of patients with thyroid cancer covered by Medicaid increased more in Medicaid expansion states compared with nonexpansion states. These findings are similar to other cancers with increased diagnosis after Medicaid expansion.3
Increased coverage could lead to increased incidence by overdiagnosis. However, our analysis showed stable size among Medicaid enrollees, indicating no increase in overdiagnosis. These findings reinforce a study4,6 of Massachusetts health care reform, which showed increased thyroidectomies—and tumors requiring treatment—with increased access to insurance.
Incidence calculation is complicated by the assessment of Medicaid enrollees covered in the National Cancer Database; both children (not included in the sample) and patients eligible for Medicaid and Medicare (counted as Medicare patients in the database) are often included in state Medicaid enrollment data. Also, while these proportions of T1 and T2 disease are low because of missing data, the missingness is not different across states or insurance groups. Despite these limitations, our data showed that Medicaid expansion was associated with an increase in thyroid cancer diagnosis among Medicaid enrollees without a change in tumor size, suggesting appropriate health care utilization.
References
- 1.Lyu W, Wehby GL. The impacts of the ACA Medicaid expansions on cancer screening use by primary care provider supply. Med Care. 2019;57(3):202-207. doi: 10.1097/MLR.0000000000001053 [DOI] [PubMed] [Google Scholar]
- 2.Choi SK, Adams SA, Eberth JM, et al. Medicaid coverage expansion and implications for cancer disparities. Am J Public Health. 2015;105(suppl 5):S706-S712. doi: 10.2105/AJPH.2015.302876 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Soni A, Simon K, Cawley J, Sabik L. Effect of Medicaid expansions of 2014 on overall and early-stage cancer diagnoses. Am J Public Health. 2018;108(2):216-218. doi: 10.2105/AJPH.2017.304166 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Vaccarella S, Franceschi S, Bray F, Wild CP, Plummer M, Dal Maso L. Worldwide thyroid-cancer epidemic? the increasing impact of overdiagnosis. N Engl J Med. 2016;375(7):614-617. doi: 10.1056/NEJMp1604412 [DOI] [PubMed] [Google Scholar]
- 5.American College of Surgeons National Cancer Database. Updated 2020. Accessed April 20, 2020. https://www.facs.org/quality-programs/cancer/ncdb
- 6.Loehrer AP, Murthy SS, Song Z, Lubitz CC, James BC. Association of insurance expansion with surgical management of thyroid cancer. JAMA Surg. 2017;152(8):734-740. doi: 10.1001/jamasurg.2017.0461 [DOI] [PMC free article] [PubMed] [Google Scholar]