Sage Journals: Discover world-class research

Abstract

Objective

Current lung cancer screening criteria based primarily on outcomes from the National Lung Screening Trial may not adequately capture all subgroups of the population at risk. We aimed to evaluate the efficacy of lung cancer screening criteria recommended by the United States Preventive Services Task Force, Centers for Medicare and Medicaid Services, and the National Comprehensive Cancer Network in identifying known cases of lung cancer.

Methods

An investigation of the Stony Brook Cancer Center Lung Cancer Evaluation Center's database identified 1207 eligible, biopsy-proven lung cancer cases diagnosed between January 1996 and March 2016. Age at diagnosis, smoking history, and other known risk factors for lung cancer were used to determine the proportion of cases that would have met current United States Preventive Services Task Force, Centers for Medicare and Medicaid Services, and National Comprehensive Cancer Network eligibility requirements for lung cancer screening.

Results

Of the 1046 ever smokers in the study, 40% did not meet the National Lung Screening Trial age requirements, 20% did not have a ≥30 pack year smoking history, and approximately one-third quit smoking >15 years before diagnosis, thus deeming them ineligible for screening. Applying the United States Preventive Services Task Force, Centers for Medicare and Medicaid Services, and National Comprehensive Cancer Network eligibility criteria to the Stony Brook Cancer Center's Lung Cancer Evaluation Center cases, 49.2, 46.3, and 69.8%, respectively, would have met the current lung cancer screening guidelines.

Conclusions

The United States Preventive Services Task Force and Centers for Medicare and Medicaid Services eligibility criteria for lung cancer screening captured less than 50% of lung cancer cases in this investigation. These findings highlight the need to reevaluate the efficacy of current guidelines and may have major public health implications.

Keywords

Lung cancer screening guidelines lung cancer risk factors early detection low-dose computed tomography

Introduction

Lung cancer mortality is higher than all other cancer types, with an overall five-year survival rate of 17%.¹ Early detection has the potential for a dramatic improvement in survival, with rates exceeding 90% when lung cancer is detected at the earliest clinical stage.² The landmark National Lung Screening Trial (NLST) was the first large randomized controlled trial to demonstrate a significant reduction in mortality among individuals undergoing low-dose computed tomography compared with chest X-ray.³ The NLST included more than 53,000 individuals aged 55–74 with at least a 30 pack year smoking history, and the study findings precipitated the United States Preventive Services Task Force (USPSTF) to recommend screening for individuals at increased risk of developing lung cancer.⁴ Consequently, in February 2015, the Centers for Medicare and Medicaid Services (CMS) approved reimbursement criteria similar to those applied in the NLST.

Results from several studies that evaluated the performance of the NLST guidelines for identifying cases of lung cancer indicate that the NLST criteria may underestimate the risk of lung cancer in some individuals and may therefore not sufficiently capture the full array of vulnerable subgroups comprising the at-risk population.^5–8 As the USPSTF, CMS, and others have established recommendations for lung cancer screening based primarily on the NLST criteria, with some noted variability, there is a need to evaluate the adequacy of their guidelines. We aimed to quantify the percentage of cases that would have met the current USPSTF, CMS, and National Comprehensive Cancer Network (NCCN) eligibility criteria for lung cancer screening, to determine the efficacy of these recommendations for standard clinical practice.

Methods

All biopsy-proven lung cancer cases seen at the Stony Brook Cancer Center's Lung Cancer Evaluation Center (LCEC) between January 1996 and March 2016 were reviewed. Eligible subjects included LCEC cases with available data for age at diagnosis, smoking status (current, former, never), and number of pack years and quit date, if appropriate. Other data abstracted from the LCEC database included history of lung disease, history of other cancer, family history of lung cancer in a parent or other first degree relative, and exposure history to asbestos or silica.

Table 1 presents the current lung cancer screening criteria recommended by each association. The USPSTF and CMS guidelines related to tobacco use are similar; individuals eligible for screening must have at least a 30 pack year smoking history, and either be current smokers or have quit within the past 15 years. Both sets of guidelines have a minimum age requirement of 55; however, the upper age limit is 80 for the USPSTF and 77 for the CMS. The NCCN guidelines provide a risk stratification approach to categorizing individuals at increased risk for lung cancer. Those deemed to be at high risk are recommended for screening, and eligibility is defined by either of two sets of criteria. The first high-risk group (NCCN1) is categorized using the exact protocols employed by the NLST (age 55–74, smoking history ≥30 pack years, and current tobacco user or quit within the past 15 years). The second high-risk group (NCCN2) is defined by a lower age cutoff (≥50) and reduced smoking history (≥20 pack years), as well as no restrictions on years since quitting, but requires one additional risk factor for eligibility. Such risk factors include an exposure history (to radon, silica, asbestos, cadmium, arsenic, beryllium, chromium, diesel fume, or nickel), personal history of other cancer, family history of lung cancer, or personal history of lung disease (emphysema, chronic obstructive pulmonary disease (COPD) or, pulmonary fibrosis). Lung cancer cases meeting either the NCCN1 or NCCN2 criteria are classified as having met the network’s eligibility guidelines for screening.

Table 1.

Current lung cancer screening recommendations.

USPSTF	CMS	NCCN 1	NCCN 2
• Age at diagnosis 55–80	• Age at diagnosis 55–77	• Age at diagnosis 55–74	• Age at diagnosis ≥ 50
• ≥ 30 pack years	• ≥ 30 pack years	• ≥ 30 pack years	• ≥ 20 pack years
• Year quit < 15 years prior to age of diagnosis	• Year quit < 15 years prior to age of diagnosis	• Year quit < 15 years prior to age of diagnosis	• 1 additional risk factor^a

CMS: Centers for Medicare and Medicaid Services; NCCN: National Comprehensive Cancer Network; USPSTF: United States Preventive Services Task Force.

Additional risk factors as defined by NCCN. History of lung disease, personal history of cancer, family history of lung cancer, exposure history (asbestos, silica, radon, cadmium, arsenic, beryllium, chromium, diesel fumes, nickel).

Statistical analyses

The distribution of demographic and other potential risk factors for lung cancer among never smokers and ever smokers (current or former) is presented as frequencies and percentages, with comparisons between the two groups evaluated by chi-square tests. P-values<0.05 are considered to be statistically significant. Among ever smokers, we present the number and percentage of cases meeting each individual screening criterion, as well as the number and proportion of those meeting all of the required criteria for each association. Risk profile scores in this investigation are constructed based on cutoff values for each of the eligibility criteria from the NCCN guidelines, which categorize individuals into three risk groups (high, moderate, and low).⁹ This risk stratification schema is consistent with other published risk prediction models for lung cancer,^10–13 with quantifiable scores ranging from 1 to 3, representing low, moderate, and high risk, respectively. As any of the other specified risk factors (beyond age and smoking history) are independently considered contributors of “high risk” according to the NCCN criteria, the presence of one or more of these factors is assigned a risk weighting value of 3 for that category. A composite score, cumulated over all criteria, including age, number of pack years, duration since tobacco cessation, and additional risk factors, was used to quantify total risk for each case. For comparative purposes, absolute risks were also computed using the Liverpool Lung Project (LLP) risk prediction models incorporating age, gender, smoking duration/pack years, emphysema/COPD, family history of lung cancer, and personal history of cancer as potential risk factors.

The Stony Brook Institutional Review Board approved this investigation and waived the need for consent.

Results

Our database included 1724 cases, dating back to 1996. Smoking history was typically incomplete for many of the early cases, as year quit, for example, was not consistently recorded. As current screening criteria require detailed duration of time since quitting to determine eligibility, patients with these missing data were excluded from the analyses. Therefore, 1207 biopsy-confirmed lung cancer cases were included in this investigation. Table 2 presents the distribution of demographic, clinical, and other variables, stratified by smoking status at the time of diagnosis. A total of 161 (13.3%) lung cancer cases never smoked, and this group was more likely to be diagnosed at a younger age than those with a history of tobacco use (p = 0.008). Never smokers were also more likely to be female (p = 0.002) and less likely to have a history of exposure to asbestos or silica (p = 0.001). Among the 1046 smokers, approximately two-fifths were actively smoking at the time of diagnosis, and 80% had a ≥ 30 pack year smoking history. One-third of the ever smokers had a prior history of lung disease, which was twice as high as the proportion found among the never smokers (p < 0.001). A personal history of any other cancer was noted in one-fourth to one-third of all LCEC lung cancer cases.

Table 2.

Demographic and clinical characteristics of N = 1207 patients with biopsy-proven lung cancer.

Characteristic	Never smoker	Ever smoker	p-value
	(N = 161)	(N = 1046)
	n (%)	n (%)
Age at diagnosis (years)			0.008
≤49	21 (13.0)	69 (6.6)
50–54	19 (11.8)	87 (8.3)
55–74	84 (52.2)	629 (60.1)
75–77	9 (5.6)	105 (10.0)
78–80	9 (5.6)	64 (6.1)
81+	19 (11.8)	92 (8.8)
Gender (male)	57 (35.4)	507 (48.5)	0.002
Smoking status at diagnosis			N/A
Current	N/A	395 (37.8)
Former	N/A	651 (62.2)
Never	161 (100)	N/A
Number of pack years at diagnosis			N/A
0	161 (100)	N/A
1–19	N/A	102 (9.8)
20–29	N/A	111 (10.6)
30+	N/A	832 (79.6)
History of lung disease	27 (16.8)	342 (32.7)	<0.001
History of other cancer	53 (32.9)	269 (25.7)	0.06
Family history of lung cancer
Mother or father	13 (8.1)	135 (12.9)	0.09
Other first degree relative	6 (3.7)	91 (8.7)	0.03
Exposure history^a	6 (3.7)	124 (11.9)	0.001

Includes asbestos and silica.

Table 3 presents the distribution of lung cancer cases with a history of tobacco use who would have met the current lung cancer screening recommendations for each set of organization guidelines. Using a minimum age cutoff of 55, 15% of cases would not have met the age eligibility requirement specified by the USPSTF or CMS. An additional 15% of cases were older than the CMS upper age limit (77) and would therefore not have been eligible for screening. Similarly, 20% of the smokers with lung cancer would not have been captured by the current screening criterion requiring a smoking history of ≥30 pack years, and 28.3% would not have been eligible for screening because they quit smoking >15 years prior to diagnosis. Overall, only 49 and 46% of LCEC lung cancer cases would have been eligible for screening based on the current USPSTF and CMS guidelines, respectively.

Table 3.

Distribution of N = 1046 ever smokers with lung cancer who meet current screening recommendations.

Screening criteria	USPSTF N (%)	CMS N (%)	NCCN1 N (%)	NCCN2 N (%)	NCCN1/NCCN2 N (%)
Age					N/A
Under requirement	156 (14.9)	156 (14.9)	156 (14.9)	69 (6.6)
Met requirement	798 (76.3)	734 (70.2)	629 (60.1)	977 (93.4)
Above requirement	92 (8.8)	156 (14.9)	261 (25.0)	0 (0.0)
Pack years					N/A
Did not meet requirement	213 (20.4)	213 (20.4)	213 (20.4)	102 (9.8)
Met requirement	833 (79.6)	833 (79.6)	833 (79.6)	944 (90.2)
Time since quit				N/A	N/A
Did not meet requirement	296 (28.3)	296 (28.3)	296 (28.3)
Met requirement	750 (71.7)	750 (71.7)	750 (71.7)
Risk factors	N/A	N/A	N/A		N/A
Did not meet requirement				352 (33.7)
Met requirement				694 (66.3)
Total recommended
Met requirements	515 (49.2)	484 (46.3)	429 (41.0)	590 (56.4)	730 (69.8)

CMS: Centers for Medicare and Medicaid Services; NCCN: National Comprehensive Cancer Network; USPSTF: United States Preventive Services Task Force.

The NCCN defines high-risk individuals eligible for screening using two distinct protocols. The first protocol (NCCN1) uses criteria identical to those used in the NLST and resulted in 41% of LCEC cases meeting such guidelines. The second protocol (NCCN2), which includes smokers with a lesser smoking history (≥20 pack years), wider age range (≥50), and without restrictions on years since quitting, along with one additional risk factor, captured the highest proportion of eligible cases for screening (56%). Approximately two-thirds of LCEC lung cancer cases had at least one additional risk factor (beyond age and smoking history) that would qualify them as a candidate for screening. Taken collectively (NCCN1 or NCCN2), 730/1046 (69.8%) cases would have been eligible for screening.

Table 4 provides risk profile scoring assignments based on the distribution of cases and clinical interpretation for each of the eligibility criteria. Applying the Stony Brook Lung Cancer Screening (SBLCS) risk assessment model to the study cohort, Table 5 presents the distribution of total risk scores among the 1046 ever smokers in LCEC. A score ≥9 would have successfully captured more than three quarters of cases, while a score ≥8 would have deemed 89% of individuals eligible for screening. Applying LLP risk prediction models to the LCEC data resulted in absolute risk percentages ranging from 1.41 to 20.36%, with a median of 5.13%. In total 96% of cases had a risk >3%, 81% had >4% risk, 54% had >5% risk, and 15% had >7% risk.

Table 4.

Stony Brook Lung Cancer Screening (SBLCS) risk profile scoring criteria.

Age (years)	Assigned risk	# Pack years	Assigned risk	Years since quit	Assigned risk	Additional risk factors^a	Assigned risk
<50	1	1–19	1	0–14	3	Yes	3
50–54	2	20–29	2	15–20	2	No	0
≥55	3	≥30	3	≥20	1

Additional risk factors as defined by NCCN include history of lung disease, personal history of cancer, family history of lung cancer, exposure history (asbestos, silica, radon, cadmium, arsenic, beryllium, chromium, diesel fumes, nickel).

Table 5.

Distribution of Stony Brook Lung Cancer Screening (SBLCS) risk profile scores among N = 1046 lung cancer cases.

Total risk score	Cases stratified by total risk score		Cumulative cases captured by total risk score
Total risk score	N	(%)	N	(%)
≤5	21	(2.0)	1046	100.0
6	24	(2.3)	1025	98.0
7	68	(6.5)	1001	95.7
8	112	(10.7)	933	89.2
9	222	(21.2)	821	78.5
10	120	(11.5)	599	57.3
11	113	(10.8)	479	45.8
12	366	(35.0)	366	35.0
Total	1046	(100.0)	–	–

Discussion

In our study, only one-half of ever smokers met the USPSTF criteria for lung cancer screening, and fewer (46%) met the CMS guidelines. The NCCN recommendations captured 70% of cases, as a result of a multifaceted set of guidelines which include considerations of additional risk factors for screening eligibility, coupled with a wider age range, and less stringent criterion for history of tobacco use. These findings highlight the need to reevaluate the current clinical practice guidelines for lung cancer screening, which are based primarily on the eligibility criteria provided by CMS.

A total of 13.3% of lung cancer cases in the present investigation never smoked. This proportion of never smokers in LCEC is consistent with a recent report based on a similar cohort of 1351 pathologically confirmed incident lung cancer cases identified between 1984 and 2011 in Olmsted County, Minnesota, of which 9.3% never smoked.¹⁴ Although a recent study suggested that never smokers with a high enough risk for lung cancer would benefit from screening, a very small percentage would likely ever reach this degree of risk.¹⁵

Among ever smokers, approximately one-fifth of the lung cancer cases in the LCEC smoked <30 pack years, and almost one-third quit smoking ≥15 years ago, thereby making them ineligible for screening based on the current USPSTF and CMS guidelines. These findings are consistent with the Olmsted County data, in which 19.5% of lung cancer cases smoked <30 pack years, and 38.3% reported cessation of tobacco use more than 15 years prior.¹⁴ Among LCEC lung cancer cases, 15% were diagnosed before age 55, and almost 9% after age 80, implying that almost one-quarter of the cases would not meet the current USPSTF screening criteria. These findings are again comparable with those from Olmsted County, where 13% of cases were younger than 55, and 12% older than 80.¹⁴ Due to increased life expectancy in the United States, screening older individuals (aged 74–84) might be beneficial, particularly among females, as their survival benefits after aggressive therapy have been shown to be similar to those of younger women (aged 55–74).¹⁵

Although data are limited on how well the current “clinical practice” inclusion criteria capture the population at risk for lung cancer, several studies have retrospectively evaluated the performance of the NLST inclusion protocols on the detection of lung cancer. Pinsky and Berg reviewed data from the 2010 US census; the National Health Interview Survey; and the Surveillance, Epidemiology and End Results investigation, and found that only 27% of lung cancer patients from these registries met the criteria for screening outlined by the NLST.⁶ Using protocols based on the NLST criteria, the Pamplona International Early Lung Action Detection Program and the Pittsburgh Lung Screening Study both yielded suboptimal sensitivity.⁸ The Pamplona study screened 3061 participants aged 40 or older, who were current or former smokers with at least a 10 pack year smoking history, of whom only 1112 (36%) met the NLST screening criteria. Of the 54 lung cancer cases detected, 21 (39%) would have been missed using the more restrictive NLST criteria.^8,16 The Pittsburgh investigation included 3638 current or former smokers aged between 50 and 79 with at least a 12.5 pack year smoking history, of whom 2161 (59%) met the NLST criteria^8,17,18; of the 96 incident lung cancers identified, 19 (20%) would have been missed using the NLST criteria.

The Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial compared chest radiography with standard care among smokers (regardless of tobacco use duration and quantity) aged 55–74. Among a subset of 37,332 participants in the intervention group of the trial, 14,144 (37.9%) met the NLST guidelines.⁷ In a further investigation comparing models incorporating additional risk factors to the NLST criteria, the risk prediction models for the more inclusive PLCO trial protocols were significantly more sensitive than the NLST criteria for detecting lung cancer. This study, however, lacked cost analysis, so it remains unclear whether a screening program using this risk modeling would maintain the high cost-effectiveness achieved in the NLST. Nonetheless, of the 678 known lung cancers, the PLCO protocols successfully identified 81 (11.9%) more cases than did the NLST criteria over the six-year follow-up period.⁷ Consistent with the PLCO results, 41% of LCEC lung cancer cases would have met the NLST criteria. These findings add to a growing evidence base indicating that the current eligibility criteria for lung cancer screening may lack the sensitivity required to adequately capture the population at risk.

Several studies have reported on using emphysema as a factor to refine the eligibility protocols for lung cancer screening.^8,17,19–21 One investigation suggests narrowing the selection criteria,¹⁹ while others support adding radiologic emphysema to screening protocol guidelines to improve the detection of incident lung cancer.⁸ In our study, where 19% of ever smokers reported a history of emphysema, adding this factor to the USPSTF guidelines as a complementary option for eligibility would have captured 67 (6.4%) previously ineligible cases for screening, and 77 (7.4%) additional cases for screening eligibility using the CMS criteria. Given the high prevalence of emphysema/COPD among smokers, these findings suggest that including this risk factor in the screening protocols for lung cancer should be considered.

The most recent National Health Interview Survey indicates that cigarette smoking among US adults has declined from 20.9% (45 million persons) in 2005 to 16.8% (40 million persons) in 2014.²² This trend will directly influence characterization of the lung cancer screening population in the future. Given the limitations of the currently approved screening protocols recommended by CMS and others, which failed to capture approximately one-half of lung cancer cases in this investigation, as well as the noted shift in smoking patterns over the past decade, there is an even more pressing need to refine the risk profiles used to define the at-risk population for lung cancer screening. Additionally, the current guidelines do not take into consideration ethnicity. A very small percentage of NLST participants were considered minority, and recent studies suggest that the CMS screening eligibility criteria do not correspond with lung cancer risk among Black and Hispanic subgroups.²³ An ethnic-specific model for African-Americans has been developed and validated to provide more precise estimates for predicting risk in this population.²⁴

Although this investigation is strengthened by the inclusion of a large, well-defined patient population with biopsy-proven lung cancer, it has several limitations. First, LCEC cases likely represent a cohort with more advanced disease than individuals who would be targeted for screening, so the percentage of cases meeting the given eligibility criteria in the present study may overestimate the sensitivity of the guidelines. Second, the study was limited by the use of self-reported data on smoking status and other known risk factors. Cases who reported no prior occurrence of disease, or for which the information was not available, were both defined as having a null history, which could result in some unknown level of underreporting. Additionally, the LCEC database only collected exposure histories related to asbestos and silica, thereby potentially missing information on other exposures, such as radon and cadmium. These missing data would not influence the findings related to the USPSTF and CMS guidelines, as neither includes such risk factors among the eligibility criteria; however, the absence of such data may underestimate the sensitivity of the NCCN criteria. It is expected, however, that these exposures would be less common. Another consideration with regard to interpretation of the study findings is the potential effect of a temporal downward trend in tobacco use over the past two decades.¹⁴ Lastly, the LCEC cohort originated from a single institution in a suburban geographic location in the Northeast, and the findings may therefore not be generalizable to the population at large.

The Society of Thoracic Surgeons and American Association of Thoracic Surgery have strongly advocated for the need to expand the current lung cancer screening criteria, which are primarily based on the NLST guidelines.^25–27 The present investigation adds to a growing evidence base that highlights the need to reevaluate these restrictive protocols. Our findings suggest that implementing a weighted risk profile approach for defining individuals at high risk may improve sensitivity and capture a higher proportion of at-risk subjects who would otherwise be missed by the current guidelines. The SBLCS risk profile schema provided in this investigation is based primarily on the NCCN cutoffs for low, moderate, and high-risk groups and was not designed to provide absolute risk estimates for the development of lung cancer. The SBLCS scoring structure is presented as a simple tool that could be easily implemented in a clinical setting without necessitating regression-based risk-prediction calculations. To evaluate the performance of the suggested risk profile tool, we applied the LLP risk prediction model²⁸ to the LCEC data. The results indicated that 96% of cases had >3% risk, a cutoff that has been suggested as sufficiently high for inclusion in a screening program.²⁸ This finding is consistent with data reported by the LLP, in which 93% of LLP participants had a >3% risk of developing lung cancer.²⁸ Additional analyses also indicated that the SBLCS scoring system produced significantly well-correlated results with the LLP model (Pearson r = 0.5, p < 0.01). These data suggest the potential usefulness of the SBLCS scoring tool; however, additional research is required to validate the findings in other populations and to optimize the weighted risk assignments for each of the eligibility criteria.

Conclusions

Screening for lung cancer requires well-defined criteria that effectively identify all subgroups of the at-risk population. Lung cancer screening is in its infancy and, although it is beyond the scope of this investigation to recommend policy changes, consideration should be given to redefining the population at risk, particularly as the current screening protocols recommended by the USPSTF and CMS identified fewer than 50% of lung cancer cases in the present investigation.

Footnotes

Acknowledgements

We thank John Reagan and Sean Reagan, student volunteers of the LCEC, for their assistance in the data abstraction.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

American Cancer Society. Cancer facts and figures 2015, http://www.cancer.org/acs/groups/content/@editorial/documents/document/acspc-044552.pdf (accessed 1 September 2016).

Henschke

Yankelevitz

et al.

International Early Lung Cancer Action Program I. Survival of patients with stage I lung cancer detected on CT screening. N Engl J Med 2006; 355: 1763–1771.

Aberle

Adams

et al.

National Lung Screening Trial Research T. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011; 365: 395–409.

Moyer

Force

USPST

. Screening for lung cancer: U.S. Preventive Services Task Force recommendation statement. Ann Internal Med 2014; 160: 330–338.

Farjah

Wood

Zadworny

et al.

Resected lung cancer patients who would and would not have met screening criteria. Ann Thorac Surg 2016; 101: 274–279.

Pinsky

Berg

. Applying the national lung screening trial eligibility criteria to the US population: what percent of the population and of incident lung cancers would be covered? J Med Screen 2012; 19: 154–156.

Tammemagi

Katki

Hocking

et al.

Selection criteria for lung-cancer screening. N Engl J Med 2013; 368: 728–736.

Sanchez-Salcedo

Wilson

de-Torres

et al.

Improving selection criteria for lung cancer screening. The potential role of emphysema. Am J Respir Crit Care Med 2015; 191: 924–931.

NCCN. NCCN Guidelines (2016), https://www.nccn.org/patients/guidelines/lung_screening/index.html#162016 (accessed 1 September 2016).

10.

Spitz

Hong

Amos

et al.

A risk model for prediction of lung cancer. J Natl Cancer Inst 2007; 99: 715–726.

11.

Cassidy

Myles

van Tongeren

et al.

The LLP risk model: an individual risk prediction model for lung cancer. Br J Cancer 2008; 98: 270–276.

12.

Tammemagi

Pinsky

Caporaso

et al.

Lung cancer risk prediction: prostate, lung, colorectal and ovarian cancer screening trial models and validation. J Natl Cancer Inst 2011; 103: 1058–1068.

13.

Marcus

Chen

Raji

et al.

LLPi: Liverpool lung project risk prediction model for lung cancer incidence. Cancer Prev Res 2015; 8: 570–575.

14.

Wang

Midthun

Wampfler

et al.

Trends in the proportion of patients with lung cancer meeting screening criteria. JAMA 2015; 313: 853–855.

15.

Varlotto

Decamp

Flickinger

et al.

Would screening for lung cancer benefit 75- to 84-year-old residents of the United States?

Front Oncol 2014; 4: 37–37.

16.

Bastarrika

Garcia-Velloso

Lozano

et al.

Early lung cancer detection using spiral computed tomography and positron emission tomography. Am J Respir Crit Care Med 2005; 171: 1378–1383.

17.

Wilson

Weissfeld

Balkan

et al.

Association of radiographic emphysema and airflow obstruction with lung cancer. Am J Respir Crit Care Med 2008; 178: 738–744.

18.

Wilson

Weissfeld

Fuhrman

et al.

The Pittsburgh lung screening study (PLuSS): outcomes within 3 years of a first computed tomography scan. Am J Respir Crit Care Med 2008; 178: 956–961.

19.

Kovalchik

Tammemagi

Berg

et al.

Targeting of low-dose CT screening according to the risk of lung-cancer death. N Engl J Med 2013; 369: 245–254.

20.

de Torres

Bastarrika

Wisnivesky

et al.

Assessing the relationship between lung cancer risk and emphysema detected on low-dose CT of the chest. Chest 2007; 132: 1932–1938.

21.

Maldonado

Bartholmai

Swensen

et al.

Are airflow obstruction and radiographic evidence of emphysema risk factors for lung cancer? A nested case-control study using quantitative emphysema analysis. Chest 2010; 138: 1295–1302.

22.

Jamal

Homa

O’Connor

et al.

Current cigarette smoking among adults—United States, 2005–2014. MMWR Morb Mortal Wkly Rep 2015; 64: 1233–1240.

23.

Fiscella

Winters

Farah

et al.

Do lung cancer eligibility criteria align with risk among Blacks and Hispanics?

PLoS One 2015; 10: e0143789–e0143789.

24.

Etzel

Kachroo

Liu

et al.

Development and validation of a lung cancer risk prediction model for African-Americans. Cancer Prev Res 2008; 1: 255–265.

25.

National Comprehensive Cancer Network Screening Guidelines, 2012, http://www.rrmginc.com/docs/NCCN_GuidelinesLungCancerScreening.pdf (accessed 1 September 2016).

26.

Jaklitsch

Jacobson

Austin

et al.

The American Association for Thoracic Surgery guidelines for lung cancer screening using low-dose computed tomography scans for lung cancer survivors and other high-risk groups. J Thorac Cardiovasc Surg 2012; 144: 33–38.

27.

Comment on USPSTF Statement on Lung Cancer Screening, https://www.sts.org/sites/default/files/documents/pdf/advocacy/USPSTF%20Comments.pdf (accessed 1 September 2016).

28.

Field

Duffy

Baldwin

et al.

The UK Lung Cancer Screening Trial: a pilot randomised controlled trial of low-dose computed tomography screening for the early detection of lung cancer. Health Technol Assess 2016; 20: 1–146.

Evaluating efficacy of current lung cancer screening guidelines

Abstract

Objective

Methods

Results

Conclusions

Keywords

Introduction

Methods

Statistical analyses

Results

Discussion

Conclusions

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

References