Secondary Logo

Journal Logo

Original Articles

The Impact of Medicare Health Insurance Coverage on Lung Cancer Screening

Sun, Jiren BS*; Perraillon, Marcelo Coca PhD; Myerson, Rebecca PhD

Author Information
doi: 10.1097/MLR.0000000000001655

Abstract

Lung cancer is the leading cause of cancer death in the United States, causing >130,000 deaths each year.1 Most patients with lung cancer are diagnosed after metastasis, at which point the 5-year survival rate is only 6%.1 Identifying lung cancer at an early stage can reduce mortality. Screening for lung cancer annually using low-dose computed tomography can reduce lung cancer mortality by 20% among high-risk adults.2,3 The US Preventive Services Task Force (USPSTF) has recommended annual lung cancer screening for asymptomatic high-risk adults, defined as adults age 55–80 who have a 30 pack-year smoking history and who either currently smoke or quit within the past 15 years,4 and these screening eligibility criteria expanded in 2021.5

Despite the health benefits, uptake of lung cancer screening remains low. Data drawn from 2 multistate samples in 2017 and 2018 showed that only 15%–20% of high-risk adults received lung cancer screening.6–8 National data from 2015 found that only about 5% of high-risk adults received screening.9 Overuse of screening among lower risk adults is also a concern9 due to the potential adverse effects of radiation exposure or physical and psychological consequences of false positives.10,11 There is, therefore, a need to identify interventions to increase informed, guideline-concordant lung cancer screening among high-risk patients without escalating overscreening among lower risk patients.

Lack of health insurance and inadequate access to care could contribute to gaps in screening. Lung cancer screening rates are 3 times higher among the insured than the uninsured.8 The costs of screening are high,12–14 and health insurance reduces the cost to patients: the USPSTF level B rating ensures that qualified health insurance plans cover screening without cost-sharing for eligible, high-risk adults.15 Prior research found associations between expanded access to health insurance and uptake of colorectal, cervical, and breast cancer screening as well as early detection of these tumors.16–18

Health insurance access increases at age 65 when nearly all Americans become eligible for Medicare. This change may be relevant to lung cancer screening given that approximately half of lung cancer cases occur in individuals 55–74 years old, and Medicare has covered lung cancer screening for high-risk adults since 2015, starting after the USPSTF recommendation.19,20 Given the potential health benefits of lung cancer screening, understanding the impact of Medicare health insurance coverage on screening is important for future policy decisions.2,14

Increasing lung cancer screening among high-risk men is particularly important. Lung cancer incidence and mortality are higher among men than women.21,22 Compared with women, men are also more likely to smoke,23,24 and less likely to use recommended health care.25–27 Prior data on the impact of health insurance expansions find different effects by sex, suggesting that pooling data from men and women together may mask important heterogeneity.28,29 Gaining access to Medicare at age 65 could increase lung cancer screening among high-risk adults who use care infrequently or lack a usual source of care; we hypothesized that the majority of such individuals would be men.

This study estimated the causal effect of nearly universal access to Medicare coverage at age 65 on lung cancer screening utilization in 2017–2019. Data were stratified by sex and by the level of lung cancer risk, as defined based on the USPSTF criteria in place during our data period.

METHODS

Data

We used 2017–2019 data from the Behavioral Risk Factor Surveillance System (BRFSS), an annual cross-sectional survey representative at the state and national levels.8 Starting in 2017, questions about lung cancer screening were included in an optional module. The module was adopted by 10 states in 2017, 8 states in 2018, and 20 states in 2019 (Supplemental Table 1, Supplemental Digital Content 1, https://links.lww.com/MLR/C355).

The primary outcome was the use of lung cancer screening in the past 12 months, based on participants’ responses to the question, “In the last 12 months, did you have a CT or CAT scan?” Those who responded “Yes, to check for lung cancer” were categorized as having received lung cancer screening. Those who selected the response, “No (did not have a CT scan),” or “Had a CT scan, but for some other reason,” were categorized as not having been screened. We extracted data on respondents’ age, race, employment status, income level, education level, veteran status, state of residence, the state’s Medicaid expansion status, and year of the interview as covariates in the multivariable models.

To examine potential mechanisms underlying the results, we investigated 2 secondary outcomes: having a routine checkup in the past year and foregoing health care due to cost during the past 12 months. To assess the assumption that outcomes would have remained continuous (smooth) at age 65 in the absence of the Medicare program, we examined changes in other socioeconomic variables, including retirement status, employment status (unemployed vs. employed), veteran status, and education (college-educated vs. non–college-educated).

Stratification variables in the main analysis included sex (male vs. female) and lung cancer risk (high vs. lower lung cancer risk). High lung cancer risk was defined as meeting the USPSTF criteria for lung cancer screening eligibility in place during the data period (Supplemental Method 1, Supplemental Digital Content 1, https://links.lww.com/MLR/C355). In a supplemental analysis, we identified the full set of people considered eligible for screening under the new USPSTF guidelines released in 2021.5 Participants who turned 65 during the 12-month look-back period were excluded from the main analysis but included in robustness checks.

In supplemental analyses, we stratified the data by education (with vs. without a college education) and by geographic location (states that had vs. had not expanded Medicaid eligibility for low-income adults). The motivation for this analysis is that these groups may vary in their insurance changes at age 65.

Research Design and Statistical Analysis

Research Design

We employed a regression discontinuity design (RDD) to determine the causal relation between nearly universal access to Medicare coverage and lung cancer screening. Researchers have used this method to study the effects of policies with age-based eligibility cutoffs,30–32 including the impact of Medicare coverage on cancer screening and mortality.28 The internal validity of this design relies on the arbitrariness of age 65 as the threshold for Medicare eligibility. Medicare coverage is nearly universally available at age 65, but not at younger ages. When individuals cannot manipulate with precision the main eligibility criterion (age), the exposure (obtaining Medicare coverage) can be viewed as if it were random among participants close to the criterion threshold (age 65). This interpretation is possible because there is no deterministic relationship between turning 65 and receiving lung cancer screening (the outcome). However, because lung cancer screening increases with age, the design requires that regression models correctly specify the relationship between these 2 variables.33 The validity of the design can be checked by ensuring that the discontinuity in the outcome is not caused by a discontinuity in confounders.

Statistical Analysis

We documented the characteristics of high-risk participants before gaining Medicare coverage. We first assessed differences in demographic and health-related characteristics by sex, using χ2 tests to compare categorical variables. We then assessed the sex distribution among high-risk adults just younger than 65 who lacked a usual source of care.

The main RDD specification used a bandwidth of 9 around age 65. We estimated linear probability models for ease of interpretation, although our conclusions are the same using logistic models. The outcome was an indicator of lung cancer screening receipt in the last 12 months. The exposure of interest was an indicator variable equal to 1 for individuals who were over age 65 throughout the 12-month look-back period and 0 for those who were under 65 throughout the look-back period. Following standard practice,32,33 the models controlled for age and were centered at age 66, the age of the youngest respondents who were over 65 throughout the 12-month look-back period. The main model used a quadratic specification for age and, to increase the precision of estimates, respondent’s race, employment status, income level, education level, veteran status, indicator variables for interview year and for state of residence, and state’s Medicaid expansion status.34 We allowed differential age trends above and below the cutoff by including interactions between age and exposure. Thus, the coefficient of interest captures the discontinuity in lung cancer screening at the cutoff point.

Following previous studies,31 the RDD model used Eicker-Huber-White heteroscedasticity-robust SEs for inference, based on their superior coverage properties for a discrete running variable.35 Sample weights were incorporated to account for the complex sampling design of the BRFSS data. See Supplemental Method 2 (Supplemental Digital Content 1, https://links.lww.com/MLR/C355) for additional details.

Our main analysis used the definition of high-risk from the USPSTF recommendations in place during our data period, corresponding to the people for whom Medicare-covered lung cancer screening at the time. To provide additional context, we also extracted health insurance coverage rates among nonelderly participants eligible for screening under the 2021 guidelines.

Robustness Checks

We conducted a series of robustness checks to assess the plausibility of the assumptions underlying the model and the sensitivity of findings to model specification. First, we tested for discontinuities at age 65 in potential confounders, following similar studies.31 Second, we implemented alternate model specifications, including removing covariates from the model; using different windows of data around age 65; using linear trends, rather than a quadratic polynomial to adjust for age; using interacted models rather than stratified models; using logistic regression rather than linear regression; employing a nonparametric modeling approach, which does not assume a functional form a priori and places a higher weight on observations closer to age 6536; and including data from individuals who were partially treated during the 12-month look-back period while adding an indicator variable to flag these participants. Third, we assessed balance in the characteristics of people with and without missing data on lung cancer screening. Fourth, we examined whether comparable changes in lung cancer screening were found at ages other than 65. Fifth, we examined potential mechanisms underlying the findings by examining changes in access to care variables at age 65. Sixth, we conducted stratified analyses to examine the relationship between change in insurance coverage and change in lung cancer screening, using conceptually derived subgroups.

Analyses were conducted using Stata SE, version 16.1 software (StataCorp). We assessed statistical significance at the 0.05 level using 2-sided tests and calculated 95% confidence intervals (CIs). The research was approved by the University of Wisconsin-Madison Institutional Review Board.

RESULTS

The study sample included 11,163 high-risk and 80,501 low-risk participants, using the categorization from the USPSTF guidelines; risk level could not be calculated for 11,907 participants (11.50%). Overall, 98.10% of high-risk participants (10,951 individuals) and 84.64% of low-risk participants (68,140 individuals) replied to survey questions about their use of lung cancer screening. Women accounted for 45.7% of the high-risk respondents and 57.6% of the low-risk participants.

Table 1 reports the characteristics of the high-risk respondents in the sample at ages 56–64, just before becoming eligible for Medicare; these data show that women were more likely than men to receive preventive care and to be diagnosed with comorbid conditions. Specifically, women were more likely than men to be diagnosed with chronic obstructive pulmonary disease (41.6% vs. 26.3%, P<0.001), to be diagnosed with asthma (24.7% vs. 11.6%, P<0.001), to have a personal doctor or health care provider (89.7% vs. 81.6%, P<0.001), and to have had a routine checkup in the past year (83.3% vs. 78.2%, P=0.006). Men accounted for 69% of high-risk adults aged 56–64 who lacked a usual source of care.

TABLE 1 - Summary Statistics for People at High Risk for Lung Cancer
Raw # of Patients* (N=6560) [n (Weighted %)]
Characteristics Men (n=3504) Women (n=3056) P SMD
Ages 56–64
 Age [median (IQR)] (y) 60 (4) 60 (4) 0.27 0.027
 Race/ethnicity
  Non-Hispanic White 3055 (85.4) 2681 (88.7) 0.21 0.039
  Non-Hispanic Black 128 (7.0) 116 (4.8)
  Hispanic 67 (3.1) 44 (1.9)
  Others 193 (4.4) 175 (4.6)
 Income
  <$25,000 1153 (34.5) 1182 (40.2) 0.06 0.173
  $25,000 to <$50,000 822 (25.4) 654 (25.2)
  ≥$50,000 1122 (40.1) 773 (34.6)
 Employment status
  Unemployed 211 (8.3) 189 (12.0) 0.007 0.125
  Employed 1683 (67.2) 1094 (58.2)
  Retired 680 (24.5) 558 (29.7)
 Education
  College-educated 1507 (39.1) 1497 (42.5) 0.17 0.122
  Non–college-educated 1991 (60.9) 1550 (57.5)
 Veteran status
  Veteran 979 (26.5) 125 (4.3) <0.001 0.689
  Nonveteran 2517 (73.5) 2930 (95.7)
 Health insurance status
  Insured 3030 (88.1) 2729 (89.3) 0.45 0.088
  Uninsured 462 (11.9) 317 (10.7)
 State’s Medicaid expansion status
  Expanded 1722 (40.7) 1529 (43.2) 0.29 0.018
  Not expanded 1782 (59.3) 1527 (56.8)
 Ever diagnosed with COPD
  Yes 1027 (26.3) 1210 (41.6) <0.001 0.216
  No 2441 (73.7) 1826 (58.4)
 Ever diagnosed with asthma
  Yes 448 (11.6) 730 (24.7) <0.001 0.290
  No 3039 (88.4) 2314 (75.3)
 Having personal doctor(s) or health care provider(s)
  Yes 2873 (81.6) 2736 (89.7) <0.001 0.214
  No 622 (18.4) 317 (10.3)
 Routine checkup within the past year
  Yes 2669 (78.2) 2484 (83.3) 0.006 0.122
  No 794 (21.8) 546 (16.7)
 Skipping care because of costs in the past 12 mo
  Yes 562 (16.6) 598 (19.7) 0.09 0.093
  No 2932 (83.4) 2444 (80.3)
 Having a CT or CAT scan in the last 12 mo
  Yes 414 (12.6) 428 (15.8) 0.07 0.067
  No 3040 (87.4) 2575 (84.2)
*Weighted percentages were estimated using data from Behavioral Risk Factor Surveillance System (BRFSS) 2017–2019. Percentages have been rounded and may not total 100.
P-value for age is from the Wilcoxon rank-sum test, and P-values for other characteristics are from the χ2 test.
COPD indicates chronic obstructive pulmonary disease; CT/CAT, computed tomography; IQR, interquartile range; SMD, standardized mean difference.

Figure 1 depicts the sharp increase in health insurance coverage at age 65. Health insurance coverage increased from around 90% before age 65 to over 98% after age 65 for both men and women. Adjusted estimates, shown in Table 2, also reflect this jump in coverage.

F1
FIGURE 1:
Health insurance coverage among men and women at high risk for lung cancer, above and below age 65. These graphs show the proportion of people at high risk for lung cancer who report currently having health insurance coverage. Age 65 is the age of nearly universal access to Medicare coverage. The scatterplots were fit separately by sex, above and below this age cutoff. Data are from Behavioral Risk Factor Surveillance System (BRFSS) 2017–2019.
TABLE 2 - Changes in Health Insurance Coverage and Lung Cancer Screening at Age 65: A RD Analysis
Men Women
Age 63–64 RD at Age 65 Age 63–64 RD at Age 65
Outcome Baseline Unadjusted Adjusted Baseline Unadjusted Adjusted
Health insurance coverage
 People with high lung cancer risk (meet USPSTF criteria for screening) 90.4 8.7 (1.6–15.8) 10.4 (3.6–17.2) 90.4 9.4 (0.9–18.0) 8.8 (0.1–17.5)
   P 0.02 0.003 0.03 0.05
 People with lower risk (do not meet USPSTF criteria) 91.2 6.0 (2.6–9.3) 5.6 (2.3–8.8) 92.0 6.1 (3.3–8.8) 6.5 (3.9–9.2)
   P <0.001 <0.001 <0.001 <0.001
Lung cancer screening
 People with high lung cancer risk (meet USPSTF criteria for screening) 11.1 14.8 (0.7–28.9) 16.2 (2.4–30.0) 18.2 2.7 (−20.0 to 25.4) 1.6 (−19.8 to 23.0)
   P 0.04 0.02 0.82 0.88
 People with lower risk (do not meet USPSTF criteria) 4.0 4.3 (−0.1 to 8.6) 3.7 (−0.5 to 8.0) 4.5 3.8 (0.1–7.5) 3.9 (0.3–7.6)
   P 0.05 0.09 0.05 0.04
The columns include findings from stratified analyses, including only men or women as noted in the headlines. Models were centered at 65, so estimates apply to age 65. “Unadjusted” estimates control only for age. Models allowed age trend terms to vary above versus below the cutoff. Adjusted estimated regression discontinuities at age 65 adjusted for respondents’ age, race, employment status, income level, education level, veteran status, state of residence, state’s Medicaid expansion status, and year of the interview. 95% confidence intervals calculated using robust SEs are in parentheses.
RD indicates regression discontinuity; USPSTF, US Preventive Services Task Force.

Changes in lung cancer screening at age 65, the age of nearly universal access to Medicare coverage, and the baseline rates at slightly younger ages are shown in Table 2. In adjusted models, lung cancer screening increased among high-risk men by 16.2 percentage points (95% CI: 2.4%–30.0%; P=0.02). In contrast, high-risk women experienced no additional change at age 65 (1.6 percentage point increase, 95% CI: −19.8% to 23.0%, P=0.88). Fewer than 5% of low-risk men and women were screened before age 65. Low-risk women experienced an additional change at age 65, while low-risk men did not (3.7 percentage point increase, 95% CI: −0.5% to 8.0%, P=0.09 for low-risk men; 3.9 percentage point increase, 95% CI: 0.3% to 7.6%, P=0.04 for low-risk women, respectively). Figure 2 and Supplemental Figure 1, Supplemental Digital Content 1, https://links.lww.com/MLR/C355, depict unadjusted lung cancer screening uptake rates among high-risk and low-risk adults, respectively, corresponding to the findings reported in Table 2. These data also show that, even after nearly universal access to Medicare coverage, 80.2% of high-risk men and 80.9% of high-risk women did not receive lung cancer screening.

F2
FIGURE 2:
Lung cancer screening among men and women at high risk for lung cancer, above and below age 65. These graphs show the proportion of people at high risk for lung cancer who reported having a CT/CAT scan check for lung cancer in the past 12 months. Age 65 is the age of nearly universal eligibility for Medicare coverage. The scatterplots were fit separately by sex, above and below this age cutoff. Data from people who turned 65 during the 12-month look-back period (eg, who had partial exposure to nearly universal access to Medicare coverage) were excluded from the analysis. Data are from Behavioral Risk Factor Surveillance System (BRFSS) 2017–2019. CT/CAT indicates computed tomography.

Secondary analyses suggested the plausibility of our main findings. First, we found no significant change at age 65 in potential confounders such as the proportion of survey participants who were at high risk for lung cancer, retired, unemployed, veterans, or college-educated (Supplemental Fig. 2. Supplemental Table 2, Supplemental Digital Content 1, https://links.lww.com/MLR/C355). Second, our estimates for high-risk adults were qualitatively similar in 8 alternate specifications including omitting some or all covariates, employing interacted models, specifying the model using logistic regression or with a nonparametric specification, using a wider bandwidth or using an indicator variable to flag people partially treated during the look-back period. However, the treatment effect for high-risk men was no longer statistically significant when we used a smaller bandwidth (P=0.06), and became smaller and not statistically significant when we use a linear specification for age (P=0.05) (Supplemental Table 3, Supplemental Digital Content 1, https://links.lww.com/MLR/C355). Third, while the differences in characteristics between participants who did and did not answer the lung cancer screening question were statistically significant, they were qualitatively small; the standardized difference did not exceed 0.25 for any variable among low-risk participants, and the nonresponse rate was low (1.9%) among high-risk participants (Supplemental Table 4, Supplemental Digital Content 1, https://links.lww.com/MLR/C355). Fourth, we did not find comparable changes in lung cancer screening use at ages other than 65 (Supplemental Fig. 3, Supplemental Digital Content 1, https://links.lww.com/MLR/C355). Finally, we found subgroups with larger gains in insurance coverage also experienced larger gains in lung cancer screening (Supplemental Fig. 4, Supplemental Table 5, Supplemental Digital Content 1, https://links.lww.com/MLR/C355). Taken together, these findings support the plausibility that changes in Medicare eligibility at age 65 explain the findings.

We also examined changes in access to care at age 65. Men experienced a decline in skipping needed health care due to cost at age 65, while women did not experience a significant change in this measure (5.3 percentage point decrease, 95% CI: −10.4% to −0.3%, P=0.04 for men; 3.9 percentage point decrease, 95% CI: −9.2% to 1.4%, P=0.15 for women). The use of a routine checkup did not significantly change at age 65 in our sample (Supplemental Table 6, Supplemental Digital Content 1, https://links.lww.com/MLR/C355).

The USPSTF expanded the population for whom lung cancer screening is recommended in new guidelines issued in 2021. Among nonelderly participants who would be eligible for screening under the new guidelines, 11.2% of women (705 individuals) and 14.1% of men (910 individuals) were uninsured.

DISCUSSION

Nearly universal access to Medicare health insurance coverage increased lung cancer screening among men at high risk of lung cancer. These findings are important for public health because low-dose computed tomography can reduce lung cancer mortality by 20% among high-risk adults and because—according to our data—>1 in 10 nonelderly adults eligible for lung cancer screening are uninsured.2,3

The increase in recommended lung cancer screening was significant among high-risk men but not high-risk women. In our sample, men accounted for about 7 in 10 high-risk adults just younger than age 65 who lacked a usual source of care. Among high-risk adults just younger than age 65, women were not only more likely than men to have had a recent checkup or to have a personal doctor, but also more likely than men to be diagnosed with asthma or chronic obstructive pulmonary disease, diagnoses associated with higher use of lung cancer screening.7,8

Even after gaining access to Medicare coverage, more than three quarters of high-risk adults did not receive annual screening. These data suggest persistent barriers to screening. Patient-side contributors to low uptake may include gaps in patients’ awareness of the option to be screened, knowledge of the benefits of screening, smoking-related stigma, and lack of access to care despite having health insurance.37–39 Physicians’ lack of knowledge about screening recommendations and reimbursement may also contribute to low screening rates.6,40 Multilevel interventions beyond health insurance coverage may be needed to promote uptake of lung cancer screening. Education and outreach to physicians, promoting standardized shared decision-making, and education materials for high-risk individuals with a history of smoking may improve lung cancer screening uptake.11 Electronic medical records may also be used to systematically identify high-risk patients and prompt repeated offers of screening.41

Alternate specifications and robustness checks suggest the plausibility of our main estimates and the importance of model specification. Our estimates for high-risk adults were qualitatively similar in a range of alternate specifications, although the point estimates and, occasionally, statistical significance changed with bandwidth or the model used for age. We were limited by survey sample sizes, but the magnitude of effects was large in all model specifications used. Several other findings supported the plausibility that changes in coverage at age 65 account for our results. We found no significant discontinuity in potential confounders and no comparable changes in lung cancer screening use at ages other than 65; found that subgroups with larger gains in coverage also experienced larger gains in lung cancer screening; and documented improvements in access to care at age 65, which might result from changes in coverage. Finally, the characteristics of individuals with and without missing lung cancer screening data are well-balanced, and the response rate is high, so missing data is unlikely to bias our findings.

Among low-risk patients, gaining access to Medicare coverage significantly changed screening for women but not for men. The significance of the increase for low-risk women was inconsistent across alternate specifications, however, suggesting this finding should be interpreted with caution. Shared decision-making is required for Medicare coverage of lung cancer screening, and the Choosing Wisely campaign supports conversations between physicians and patients about the risks and potential harms of screening low-risk adults.20,42

It is likely that both the previously uninsured and previously insured benefit from gaining access to Medicare. Medicare provides a free Annual Wellness Visit as well as a Welcome to Medicare Visit at age 65 that help ensure patients are up-to-date with recommended screenings43,44 and make a personalized screening schedule.45,46 Second, Medicare provides relatively generous coverage that is useful even for the already insured.47 Patients may be more willing to be screened if they think subsequent procedures and treatment will be available and affordable.48,49 Among people with Medicaid or high-deductible health plans coverage, gaining access to Medicare could result in more utilization of high-cost services. However, we are unable to examine differential effects by type of insurance among the previously insured using our data. A key limitation of the BRFSS data is that type of insurance coverage is captured in optional state modules and therefore is only available for only 16% of our sample.

This study has other important limitations. First, data were self-reported and therefore subject to recall and social desirability bias. Yet, such biases would not account for the results unless they changed sharply at age 65. Second, subpopulations who lacked phone access may be underrepresented. We accounted for the sampling scheme by incorporating the recommended sample weights. Third, we used data from BRFSS modules that were not available in every state. We used sample weights to ensure data were representative at the state level and focused on the impacts of Medicare health insurance within the sample. Fourth, our study design identifies changes in lung cancer screening associated with Medicare insurance coverage at age 65, and the results may not generalize to people in other age groups. Finally, sample size limitations in the BRFSS data prevented us from examining variation in the findings by race or ethnicity; these topics are important to examine in future research.

In summary, inadequate insurance coverage or access to care hinders lung cancer screening, especially for high-risk men. To reach the full population with recommended screening under the new 2021 guidelines, multilevel intervention strategies including access to insurance may be needed. Further improving insurance access and access to care may reduce lung cancer mortality.

ACKNOWLEDGMENTS

R.M. acknowledges support from the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health, the Office of The Director, National Institutes of Health (OD), and the National Cancer Institute (NCI) under Award Number K12HD101368, and support from the University of Wisconsin Carbone Cancer Center Support Grant (NCI P30 CA014520). M.C.P. acknowledges support from the Population Health Share Resources and funding from the University of Colorado Cancer Center Core Support Grant (NCI P30CA046934). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

REFERENCES

1. Siegel RL, Miller KD, Fuchs HE, et al. Cancer statistics, 2021. CA Cancer J Clin. 2021;71:7–33.
2. The National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395–409.
3. de Koning HJ, van der Aalst CM, de Jong PA, et al. Reduced lung-cancer mortality with volume CT screening in a randomized trial. N Engl J Med. 2020;382:503–513.
4. Moyer VA. Screening for lung cancer: US Preventive Services Task Force Recommendation Statement. Ann Intern Med. 2014;160:330–338.
5. US Preventive Services Task Force. Screening for lung cancer: US Preventive Services Task Force Recommendation Statement. JAMA. 2021;325:962–970.
6. Narayan AK, Gupta Y, Little BP, et al. Lung cancer screening eligibility and use with low‐dose computed tomography: results from the 2018 Behavioral Risk Factor Surveillance System cross‐sectional survey. Cancer. 2021;127:748–756.
7. Richards TB, Soman A, Thomas CC, et al. Screening for lung cancer—10 States, 2017. MMWR Morb Mortal Wkly Rep. 2020;69:201–206.
8. Zahnd WE, Eberth JM. Lung cancer screening utilization: a behavioral risk factor surveillance system analysis. Am J Prev Med. 2019;57:250–255.
9. Huo J, Shen C, Volk RJ, et al. Use of CT and chest radiography for lung cancer screening before and after publication of screening guidelines: intended and unintended uptake. JAMA Intern Med. 2017;177:439.
10. Richards TB, Doria-Rose VP, Soman A, et al. Lung cancer screening inconsistent with US Preventive Services Task Force Recommendations. Am J Prev Med. 2019;56:66–73.
11. Triplette M, Thayer JH, Pipavath SN, et al. Poor uptake of lung cancer screening: opportunities for improvement. J Am Coll Radiol. 2019;16:446–450.
12. Jensen MD, Siersma V, Rasmussen JF, et al. Direct and indirect healthcare costs of lung cancer CT screening in Denmark: a registry study. BMJ Open. 2020;10:e031768.
13. Mauchley DC, Mitchell JD. Current estimate of costs of lung cancer screening in the United States. Thorac Surg Clin. 2015;25:205–215.
14. Pyenson BS, Tomicki SM. Lung cancer screening: a cost-effective public health imperative. Am J Public Health. 2018;108:1292–1293.
15. Burke A, Simmons A. Increased Coverage of Preventive Services With Zero Cost Sharing under the Affordable Care Act. Washington, DC: Department of Health and Human Services; 2014:1–8.
16. Sabik LM, Bradley CJ. The impact of near-universal insurance coverage on breast and cervical cancer screening: evidence from Massachusetts. Health Econ. 2016;25:391–407.
17. Soni A, Simon K, Cawley J, et al. Effect of Medicaid expansions of 2014 on overall and early-stage cancer diagnoses. Am J Public Health. 2018;108:216–218.
18. Lin L, Soni A, Sabik LM, et al. Early- and late-stage cancer diagnosis under 3 years of Medicaid expansion. Am J Prev Med. 2021;60:104–109.
19. de Groot PM, Wu CC, Carter BW, et al. The epidemiology of lung cancer. Transl Lung Cancer Res. 2018;7:220–233.
20. Centers for Medicare and Medicaid Services. National Coverage Determination (NCD) for lung cancer screening with low dose computed tomography (LDCT). Medicare Coverage Database; 2015. Available at: www.cms.gov/medicare-coverage-database/details/ncd-details.aspx?NCDId=364&ncdver=1&bc=AAAAEAAAAAAA&. Accessed March 16, 2021.
21. Cook MB, McGlynn KA, Devesa SS, et al. Sex disparities in cancer mortality and survival. Cancer Epidemiol Biomark Prev. 2011;20:1629–1637.
22. Rubin JB, Lagas JS, Broestl L, et al. Sex differences in cancer mechanisms. Biol Sex Differ. 2020;11:17.
23. Kaiser Family Foundation (KFF). Adults who report smoking by sex; 2020. Available at: www.kff.org/other/state-indicator/smoking-adults-by-sex/. Accessed July 13, 2021.
24. Cornelius ME. Tobacco product use among adults—United States, 2019. MMWR Morb Mortal Wkly Rep. 2020;69:1736–1742.
25. Pinkhasov RM, Wong J, Kashanian J, et al. Are men shortchanged on health? Perspective on health care utilization and health risk behavior in men and women in the United States. Int J Clin Pract. 2010;64:475–487.
26. Kaiser Family Foundation (KFF). Gender differences in health care, status, and use: spotlight on Men’s health; 2015. Available at: www.kff.org/womens-health-policy/fact-sheet/gender-differences-in-health-care-status-and-use-spotlight-on-mens-health/. Accessed July 13, 2021.
27. Vaidya V, Partha G, Karmakar M. Gender differences in utilization of preventive care services in the United States. J Womens Health (Larchmt). 2012;21:140–145.
28. Myerson RM, Tucker‐Seeley RD, Goldman DP, et al. Does Medicare coverage improve cancer detection and mortality outcomes? J Policy Anal Manage. 2020;39:577–604.
29. Wehby GL, Lyu W. The impact of the ACA Medicaid expansions on health insurance coverage through 2015 and coverage disparities by age, race/ethnicity, and gender. Health Serv Res. 2018;53:1248–1271.
30. Card D, Dobkin C, Maestas N. Does Medicare Save Lives? Q J Econ. 2009;124:597–636.
31. Card D, Dobkin C, Maestas N. The impact of nearly universal insurance coverage on health care utilization: evidence from Medicare. Am Econ Rev. 2008;98:2242–2258.
32. Venkataramani AS, Bor J, Jena AB. Regression discontinuity designs in healthcare research. BMJ. 2016;352:i1216.
33. Imbens GW, Lemieux T. Regression discontinuity designs: a guide to practice. J Econom. 2008;142:615–635.
34. Gelman A, Imbens G. Why High-Order Polynomials Should Not Be Used in Regression Discontinuity Designs. J Business Econom Stat. 2019;37:447–456.
35. Kolesár M, Rothe C. Inference in regression discontinuity designs with a discrete running variable. Am Econ Rev. 2018;108:2277–2304.
36. Nichols A. RD: Stata module for regression discontinuity estimation. Boston College Department of Economics; 2016. Available at: https://ideas.repec.org/c/boc/bocode/s456888.html. Accessed January 19, 2021.
37. Carter-Harris L, Ceppa DP, Hanna N, et al. Lung cancer screening: what do long-term smokers know and believe? Health Expect. 2017;20:59–68.
38. Cataldo JK. High-risk older smokers’ perceptions, attitudes, and beliefs about lung cancer screening. Cancer Med. 2016;5:753–759.
39. Gressard L, DeGroff AS, Richards TB, et al. A qualitative analysis of smokers’ perceptions about lung cancer screening. BMC Public Health. 2017;17:589.
40. Ersek JL, Eberth JM, McDonnell KK, et al. Knowledge of, attitudes toward, and use of low-dose computed tomography for lung cancer screening among family physicians: LDCT for lung cancer screening. Cancer. 2016;122:2324–2331.
41. Modin HE, Fathi JT, Gilbert CR, et al. Pack-year cigarette smoking history for determination of lung cancer screening eligibility. Comparison of the electronic medical record versus a shared decision-making conversation. Ann Am Thorac Soc. 2017;14:1320–1325.
42. Cassel CK. Choosing wisely: helping physicians and patients make smart decisions about their care. JAMA. 2012;307:1801.
43. Camacho F, Yao NA, Anderson R. The effectiveness of Medicare wellness visits in accessing preventive screening. J Prim Care Community Health. 2017;8:247–255.
44. Jiang M, Hughes DR, Wang W. The effect of Medicare’s Annual Wellness Visit on preventive care for the elderly. Prev Med. 2018;116:126–133.
45. US Centers for Medicare and Medicaid Services. Yearly “Wellness” Visits. Available at: www.medicare.gov/coverage/yearly-wellness-visits. Accessed July 20, 2021.
46. Cuenca AE, Kapsner S. Medicare wellness visits: reassessing their value to your patients and your practice. Fam Pract Manag. 2019;26:25–30.
47. Card D, Dobkin C, Maestas N. Does Medicare save lives? Q J Econ. 2009;124:597–636.
48. Oster E, Shoulson I, Dorsey ER. Optimal expectations and limited medical testing: evidence from Huntington disease. Am Econ Rev. 2013;103:804–830.
49. Myerson RM, Lakdawalla D, Colantonio LD, et al. Effects of Expanding Health Screening on Treatment—What Should We Expect? What Can We Learn? Cambridge, MA: National Bureau of Economic Research; 2018.
Keywords:

lung cancer; computed tomography; screening; Medicare; smoking

Supplemental Digital Content

Copyright © 2021 The Author(s). Published by Wolters Kluwer Health, Inc.