Lung Cancer Screening With Low-Dose Computed Tomography in the United States—2010 to 2015

Abstract

This study examines whether low-dose computed tomography screening for lung cancer increased following the United States Preventive Services Task Force recommendations.

Lung cancer is the most preventable and leading cause of cancer deaths in the United States, with about 155 870 deaths each year. In December 2013, the United States Preventive Services Task Force (USPSTF) recommended annual screening for lung cancer with low-dose computed tomography (LDCT) for asymptomatic persons aged 55 to 80 years who have a 30 pack or more per year smoking history and currently smoke or have quit within the past 15 years. According to the 2010 National Health Interview Survey (NHIS), only 2% to 4% of high-risk smokers received LDCT for lung cancer screening in the previous year. In this study, we examined whether LDCT screening has increased following the USPSTF recommendation.

Methods

We used the 2010 and 2015 NHIS, which included 2347 respondents who met the USPSTF criteria for LDCT. Self-reported LDCT in the past year for lung cancer screening was the primary outcome of the study. Analyses excluded respondents with unknown (n = 6) or self-reported history of lung cancer (n = 41) or were missing LDCT testing information (n = 133), leaving 2167 adults available for analyses. Weighted prevalence of LDCT for lung cancer screening in the past year was calculated by factors of interest. Multivariable prevalence ratios of LDCT in the past year were estimated using predicted margins. All statistical analyses accounted for complex sampling design and were conducted with SAS callable SUDAAN statistical software (version 9.0.3, SAS Institute). The study was based on deidentified publicly available database and exempt from institutional review board and informed consent.

Results

From 2010 to 2015, the percentage of eligible smokers who reported LCDT screening in the past 12 months remained low and constant, from 3.3% in 2010 to 3.9% in 2015 (P = .60); an even lower proportion of noneligible smokers received LDCT (Table 1). Of the 6.8 million smokers eligible for LDCT screening in 2015, only 262 700 received it. Furthermore, there was no significant increase in screening from 2010 to 2015 for any of the sociodemographic groups, nor were there significant subgroup differences in screening, except between participants with or without a history of bronchitis (Table 2). Of note, over 50% (1230/2167) of smokers meeting USPSTF recommendations for LDCT screening were uninsured or Medicaid insured (Table 1).

Table 1. Prevalence of LDCT Testing for Lung Cancer in the Past Year Among Screening-Eligible and Noneligible Smokers, National Health Interview Surveys 2010 and 2015^a,b.

Characteristic	Total		2010		2015		P Valuec
	No. (%)	(95% CI)	No. (%)	(95% CI)	No. (%)	(95% CI)
Screening-eligible smokers (n = 2167)
Weighted No. receiving LDCTd			276 700		262 700
Weighted No. eligible for LDCT			8 456 800		6 819 500
Total	2167 (3.5)	(2.6-4.8)	1036 (3.3)	(2.3-4.7)	1131 (3.9)	(2.4-6.2)	.60
Smoking history
Former, ≥30 PY, quit ≤15 years ago	1020 (4.2)	(2.7-6.5)	491 (4.0)	(2.6-6.1)	529 (4.6)e	(2.1-9.4)e	.76
Current, ≥30 PY	1147 (2.9)	(1.8-4.5)	545 (2.6)e	(1.4-4.9)e	602 (3.2)	(1.8-5.6)	.64
Age, y
55-64	1119 (2.3)	(1.5-3.6)	554 (2.8)e	(1.6-5.1)e	565 (1.7)	(1.0-3.1)	.29
65-80	1048 (5.0)	(3.3-7.6)	482 (3.8)	(2.4-6.0)	566 (6.6)e	(3.6-11.9)e	.19
Sex
Male	1245 (3.8)	(2.6-5.4)	597 (3.8)	(2.5-5.9)	648 (3.8)	(2.2-6.3)	.96
Female	922 (3.2)e	(1.7-5.7)e	439 (2.5)e	(1.2-5.0)e	483 (4.0)e	(1.6-9.5)e	.46
BMI
<25	688 (5.6)	(3.4-9.3)	320 (4.4)e	(2.4-8.0)e	368 (7.2)e	(3.3-14.7)e	.36
≥25	1400 (2.6)	(1.8-3.7)	673 (2.7)	(1.7-4.3)	727 (2.5)	(1.5-4.2)	.84
Usual place for medical care
Yes	1965 (3.9)	(2.9-5.3)	934 (3.6)	(2.5-5.2)	1031 (4.3)	(2.6-6.9)	.60
No	202 (0.2)e	(0.0-1.2)e	102e,f		100 (0.4)e	(0.1-2.6)e	f
Visited PCP in past year
Yes	1726 (4.3)	(3.1-5.9)	813 (4.1)	(2.9-5.9)	913 (4.5)	(2.7-7.4)	.78
No	440 (0.6)	(0.2-1.8)	223f		217 (1.4)	(0.5-4.1)	f
Insurance type
Uninsured or Medicaid	1230 (4.2)	(2.8-6.3)	586 (3.2)	(2.0-5.1)	644 (5.5)e	(3.0-9.9)e	.20
Medicare, private, or other	937 (2.8)	(1.7-4.4)	450 (3.4)	(1.9-6.1)	487 (2.0)e	(1.1-3.6)e	.20
Raceg
White	1787 (3.5)	(2.5-5.0)	833 (3.1)	(2.0-4.6)	954 (4.1)	(2.4-6.9)	.39
Nonwhite	380 (3.5)	(2.0-6.2)	203 (4.7)e	(2.3-9.5)e	177 (2.1)e	(1.0-4.6)e	.18
Education level
<High school or high school graduate	1216 (3.4)	(2.4-4.9)	613 (2.6)	(1.6-4.1)	603 (4.6)	(2.9-7.3)	.08
Some college or college graduate	946 (3.7)	(2.2-6.2)	420 (4.3)	(2.5-7.3)	526 (3.0)e	(1.1-8.3)e	.51
Income, $
<35 000	1130 (3.9)	(2.8-5.3)	543 (3.9)	(2.5-6.1)	587 (3.8)	(2.3-6.2)	.97
≥35 000	926 (3.3)	(2.0-5.4)	446 (2.8)	(1.5-5.0)	480 (3.9)e	(1.8-8.1)e	.51
Family history of lung cancer
Yes	362 (4.5)e	(2.4-8.2)e	161 (4.8)e	(2.0-10.8)e	201 (4.1)e	(2.1-8.0)e	.76
No	1709 (3.3)	(2.3-4.8)	812 (2.8)	(1.9-4.4)	897 (3.9)	(2.1-6.9)	.42
Attempted to quit smoking in the past 12 monthsh
Yes	363 (4.1)e	(2.1-8.0)e	164 (3.3)e	(1.2-8.8)e	199 (5.1)e	(2.1-12.3)e	.52
No	784 (2.3)	(1.3-3.9)	381 (2.3)e	(1.0-5.2)e	403 (2.2)e	(1.1-4.3)e	.93
Ever diagnosed with emphysema
Yes	321 (8.9)	(5.8-13.4)	169 (9.6)	(5.8-15.5)	152 (7.9)e	(3.8-15.8)e	.64
No	1844 (2.6)	(1.7-3.9)	866 (2.0)	(1.2-3.4)	978 (3.2)e	(1.7-5.9)e	.30
Ever diagnosed with bronchitis
Yes	272 (11.2)	(6.4-18.8)	135 (11.5)	(6.5-19.7)	137 (10.7)e	(3.6-27.7)e	.90
No	1895 (2.4)	(1.7-3.5)	901 (2.1)	(1.3-3.3)	994 (2.9)	(1.8-4.6)	.30
Ever diagnosed with asthma
Yes	327 (6.2)	(3.7-10.1)	184 (8.0)	(4.4-14.0)	143 (3.2)e	(1.3-7.3)e	.08
No	1838 (3.1)	(2.1-4.5)	851 (2.3)	(1.5-3.7)	987 (4.0)	(2.3-6.7)	.16
Noneligible smokers (n = 6632)i
Total	6632 (2.4)	(1.9-2.9)	2632 (2.0)	(1.5-2.9)	3989 (2.7)	(2.1-3.6)	.12
Former, <30 PY, quit ≤15 years ago	932 (2.3)	(1.3-4.1)	378 (3.1)	(1.5-6.3)	554 (1.7)	(0.7-4.4)	.36
Former, ≥30 PY, quit >15 years ago	740 (4.0)	(2.5-6.2)	339 (2.5)	(1.1-5.4)	401 (5.8)	(2.9-11.3)	.17
Former, <30 PY, quit ≥15 years ago	3334 (1.6)	(1.2-2.3)	1255 (1.5)	(0.9-2.5)	2079 (1.7)	(1.2-2.6)	.68
Current, <30 PY	1626 (3.3)	(2.3-4.6)	671 (2.0)	(1.2-3.5)	955 (4.4)	(2.8-6.6)	.04

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); LDCT, low-dose computerized tomography; PCP, primary care physician; PY, pack-years.

^aThe following number of respondents were missing data for these items and are shown in parentheses: income (111), BMI (79), PCP visits (1), education (5), immigration status (1), family history of lung cancer (96), emphysema (2), asthma (2). Respondents with missing information were included in the model, but data are not shown.

^bPercentages are weighted.

^cP value compares 2010 vs 2015.

^dWeighted numbers take into account the assigned sampling weights of respondents.

^eUnreliable estimates as a result of relative standard errors exceeding 30%.

^fUnable to generate estimate owing to small denominator.

^gWhite includes non-Hispanic whites, nonwhite includes: Hispanic, Asian, Black, Native American/Alaskan Native and other race and/or ethnicities.

^hAmong current smokers only.

ⁱIncludes former and current smokers who do not meet the US Preventive Services Task Force Recommendations.

Table 2. Adjusted Prevalence Ratios and 95% CIs of LDCT Testing for Lung Cancer in the Past Year Among Screening-Eligible Respondents, National Health Interview Survey 2010 and 2015 (n = 2167)^a,b.

Characteristic	PR (95% CI)
Year
2010	1 [Reference]
2015	1.28 (0.66-2.47)
Age, y
55-64	1 [Reference]
65-80	1.34 (0.62-2.88)
Sex
Male	1 [Reference]
Female	0.61 (0.26-1.4)
BMI
<25	1 [Reference]
≥25	0.36 (0.16-0.8)
Usual place for medical care
Yes	1 [Reference]
No	0.12 (0.01-1.78)
Insurance type
Uninsured or medicaid	1 [Reference]
Medicare, private, or other	0.94 (0.43-2.06)
Racec
White	1 [Reference]
Nonwhite	1.31 (0.51-3.33)
Education level
<High school or high school graduate	1 [Reference]
Some college or college graduate	1.13 (0.49-2.62)
Family history of lung cancer
Yes	1 [Reference]
No	0.84 (0.32-2.21)
Smoking history
Former, ≥30 PY, quit ≤15 years ago	1.27 (0.53-3.05)
Current, ≥30 PY	1 [Reference]
Attempted to quit smoking in the past 12 monthsd
Yes	1 [Reference]
No	0.55 (0.17-1.71)
Ever diagnosed with emphysema
Yes	1 [Reference]
No	0.60 (0.19-1.90)
Ever diagnosed with bronchitis
Yes	1 [Reference]
No	0.27 (0.09-0.83)
Ever diagnosed with asthma
Yes	1 [Reference]
No	0.91 (0.35-2.35)

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); LDCT, low-dose computerized tomography; PR, prevalence ratio; PY, pack-years.

^a192 participants were not included in the model owing to missing data. Visiting a physician in the past year and income level were not included in the model owing to instability of estimates.

^bPrevalence Ratios are adjusted for: age, sex, race, smoking history, family of lung cancer, chronic respiratory conditions, and BMI.

^cWhite includes non-Hispanic whites, nonwhite includes Hispanic, Asian, Black, Native American/Alaskan Native and other race/Ethnicities.

^dAmong current smokers only.

Discussion

Screening for lung cancer using LDCT among eligible current and former smokers remained low and unchanged in 2015 following the 2013 USPSTF recommendation for annual screening. Reasons for exceptionally low uptake of screening may include gaps in smokers’ knowledge regarding LDCT, lack of access to care as well as physicians’ knowledge about screening recommendations and reimbursement. For example, according to a 2015 survey of physicians in South Carolina, 36% of physicians correctly stated that LCDT screening should be conducted annually in high-risk individuals, and 63% of physicians did not know that Medicare covers LDCT for lung cancer screening. It is also possible that physicians may be aware of LDCT screening, but have limited access to the high-volume, and high-quality radiology centers, a recommendation set forth by public health organizations and a stipulation on Medicare reimbursement. The decrease in the number of screening-eligible smokers from 8.4 million in 2010 to 6.8 million in 2015 reflects progress in tobacco control, and this has implications for the future provision of LDCT screening. Receipt of LDCT and smoking history were self-reported and subject to recall bias and the limited time following the USPSTF recommendation and Medicare-reimbursement are limitations of our study. Despite this, our study provides the first national estimate of LDCT following the USPSTF recommendation.

In conclusion, annual LCDT screening among heavy current and former smokers remains low and unchanged following the USPSTF recommendation despite the potential to avert thousands of lung cancer deaths each year. This underscores the need to educate clinicians and smokers about the benefit and risks of lung cancer screening for informed decision making.