Trends in Breast, Colorectal, and Cervical Cancer Incidence Following the Affordable Care Act

Abstract

This observational study uses incidence rates before and after adoption of the Affordable Care Act (ACA) to assess whether adoption of the ACA increased screening for breast, colorectal, and cervical cancer.

The 2010 Patient Protection and Affordable Care Act (ACA) emphasizes preventive care. In addition to expanding insurance coverage, the ACA eliminates cost sharing for services graded “A” or “B” by the US Preventive Services Task Force (USPSTF). Although these policies have improved preventive care generally, their impact on cancer screening specifically is uncertain. Whereas a recent study showed more screening in Accountable Care Organizations, a health care model pioneered by the ACA, an earlier study found increases in use of medical preventive services such as blood pressure and cholesterol checks but not in cancer screening.

Although it may take years for screening to affect mortality, higher screening rates should quickly affect incidence. We hypothesized that the implementation of major ACA policies on January 1, 2014, would be followed by an increased incidence in early-stage breast, colorectal, and cervical cancer—3 malignant neoplasms with “A” or “B” screening grades from the USPSTF.

Methods

We compared age-adjusted incidence rates of early-stage breast, colorectal, and cervical cancer in the first 9 months of 2013 (pre-ACA) and the last 9 months of 2014 (post-ACA) with an intervening 6-month “wash-in” period. Incidence rates were per 100 000 person-years and age-adjusted to the 2000 US Standard Population. To assess for change between pre- and post-ACA, we computed the incidence rate ratios (IRRs) and associated 95% confidence intervals. Then, using weighted least squares (weighting by the inverse of the variance) with a log link, we ascertained whether the relative difference in IRRs (ie, ratio of IRRs) for early-stage disease varied in a statistically significant fashion compared with locally advanced/metastatic disease. The relative difference in IRRs was estimated by exponentiating the difference-in-differences (DID) of the log IRRs. To generate the incidence rates and IRRs, we used SEER*Stat Version 8.3.4. All other analyses were performed using SAS, version 9.4. The study received an institutional review board waiver from the Brigham and Women’s Hospital.

Results

From pre- to post-ACA, the incidence of early-stage breast cancer increased from 55.5 (95% CI, 54.6-56.3) to 56.9 (95% CI, 56.0-57.7) cases per 100 000 person-years, with an IRR of 1.025 (95% CI, 1.003-1.048). Furthermore, the difference in IRRs was significantly greater in early vs locally advanced/metastatic stages (DID, 1.050; 95% CI, 1.006-1.098; P = .03) (Table).

Table. Age-Adjusted Incidence Rates and the Associated Incidence Rate Ratios (IRRs)^a According to Stage at Diagnosis for Breast, Colorectal, and Cervical Cancers^b.

Cancer	Incidence Rate (95% CI), per 100 000 Person-years		IRR (95% CI)	DID (95% CI)c	P Valued
	Pre-ACA	Post-ACA
Breast cancer (age 40-64 y)
Early stage (0-IIA)	55.5 (54.6-56.3)	56.9 (56.0-57.7)	1.025 (1.003-1.048)	1.050 (1.006-1.098)
Locally advanced/metastatic stage (IIB-IV)	23.6 (23.0-24.1)	23.0 (22.5-23.6)	0.976 (0.935-1.009)	[1 Reference]	.03
Colorectal cancer (age 50-64 y)
Early stage (0-IIA)	13.5 (13.0-14.1)	15.3 (14.7-15.9)	1.132 (1.07-1.198)	1.112 (1.030-1.200)
Locally advanced/metastatic stage (IIB-V)	17.2 (16.6-17.8)	17.5 (16.9-18.1)	1.018 (0.967-1.072)	[1 Reference]	.006
Cervical cancer (age 20-64 y)
Early stage (I)	1.7 (1.6-1.9)	1.8 (1.7-2.0)	1.053 (0.962-1.153)	0.996 (0.876-1.133)
Locally advanced/metastatic stage (II-IV)	1.7 (1.6-1.9)	1.8 (1.7-1.9)	1.049 (0.958-1.149)	[1 Reference]	.95

Abbreviations: ACA, Affordable Care Act; DID, difference-in-differences.

^aThat is, change in incidence rates between the post-ACA vs pre-ACA periods.

^bIncidence rates age-adjusted per 100 000 person-years. Patients with unknown stages were analyzed and incidence was unchanged before and after ACA across all tumor types. Stages of breast, colorectal, and cervical cancer were dichotomized into early-stage vs locally advanced or metastatic. For breast cancer, we differentiated between stage 0 to IIA (early-stage) and stage IIB to IV (locally advanced/metastatic). We opted to put stage IIA in the early-stage group and stage IIB in the other group based on the fact that their difference is determined by the size of the tumor and whether the cancer has spread to the lymph nodes, and that the survival of stage IIA disease is slightly higher than the survival of stage IIB disease. For colorectal cancer, we differentiated between stage 0 to IIA (early) and stage IIB to IV (locally advanced/metastatic). This considered that for individuals with stage IIA colon cancer, the 5-year relative survival rate is approximately 87% whereas it is 63% for those with stage IIB cancer. For cervical cancer, we distinguished between stage I (early stage) and stage II to IV (locally advanced/metastatic). This considered that the 5-year survival rate of cervical cancer stage II ranges between 58% and 63% vs 80% to 93% for stage I cancer.

^cThe relative difference in IRRs (ratio of IRRs).

^dThe P value reflects the statistical significance of the difference in changes of the IRRs during the pre- vs post-ACA periods of early-stage vs locally advanced/metastatic disease (reference category).

The incidence of early-stage colorectal cancer increased from 13.5 (95% CI, 13.0-14.1) to 15.3 (95% CI, 14.7-15.9) cases per 100 000 person-years, with a pre- to post-ACA IRR of 1.132 (95% CI, 1.07-1.198). Similarly, the change in incidence rates was significantly greater in early vs locally advanced/metastatic stages (DID, 1.112; 95% CI, 1.030-1.200; P = .006). This pattern was not seen in cervical cancer.

Discussion

We found that incidence of early-stage breast and colorectal cancer increased after the adoption of the ACA, whereas it did not vary for late-stage cancer. Although screening itself was not assessed, the trend is consistent with modest but immediate increases in colorectal and breast cancer screening following the ACA. Our finding that there was no change in detection of early-stage cervical cancer is consistent with a previous report showing that the dependent coverage expansion to age 26 years did not affect the use of the Papanicolaou test in that population.

Limitations of this observational study include assessment of only 1 year pre- and post-ACA, potential for unmeasured confounders, and unrelated background epidemiological trends, as well the inherent limitations of the difference-in-differences study design.

Despite these limitations, these results are consistent with a small but positive impact of the ACA on use of recommended cancer screening, which may vary by cancer site. Recent proposals for repealing the ACA would increase the uninsured population by tens of millions. This could easily erase these modest gains.