Draft Research Plan
Lung Cancer: Screening
May 03, 2018
Recommendations made by the USPSTF are independent of the U.S. government. They should not be construed as an official position of the Agency for Healthcare Research and Quality or the U.S. Department of Health and Human Services.
* The search for evidence on treatment in the systematic review will be limited to studies of surgical resection of stage I non-small cell lung cancer.
1. a. Does screening for lung cancer with computed tomography (CT) change the incidence of lung cancer and the distribution of lung cancer types and stages?
b. Does screening for lung cancer with CT change all-cause mortality, lung cancer mortality, or quality of life?
c. Does the effectiveness of screening for lung cancer with CT differ for subgroups defined by age, sex, race/ethnicity, presence of comorbid conditions, or risk for lung cancer?
d. Does the effectiveness of screening for lung cancer with CT differ by the number or frequency of CT scans (e.g., annual screening for 3 years, as used in the National Lung Screening Trial [NLST], vs. other approaches)?
2. Does the use of risk prediction models for identifying adults at higher risk of lung cancer mortality improve the balance of benefits and harms of screening compared with use of trial eligibility criteria (e.g., NLST criteria) or USPSTF recommendations?
3. a. What is the accuracy of screening for lung cancer with CT?
b. Does the accuracy of screening for lung cancer with CT differ for subgroups defined by age, sex, race/ethnicity, presence of comorbid conditions, or risk for lung cancer?
c. Does the accuracy of screening for lung cancer with CT differ for various approaches to nodule classification (i.e., those based on nodule size and characteristics)?
4. a. What are the harms associated with screening for lung cancer with CT?
b. Do the harms of screening for lung cancer differ with the use of Lung-RADS™ (e.g., to reduce false-positive results) or similar approaches?
c. Do the harms of screening for lung cancer differ for subgroups defined by age, sex, race/ethnicity, presence of comorbid conditions, or risk for lung cancer?
5. a. What are the harms associated with workup or surveillance of nodules detected by screening?
b. Do the harms of workup or surveillance of nodules differ with the use of Lung-RADS (e.g., to reduce false-positive results) or similar approaches?
c. Do the harms of workup or surveillance of nodules differ for subgroups defined by age, sex, race/ethnicity, presence of comorbid conditions, or risk for lung cancer?
6. How effective is surgical resection for the treatment of early (stage I) non-small cell lung cancer?
7. What are the harms associated with surgical resection of early (stage I) non-small cell lung cancer?
8. What is the magnitude of change in all-cause and lung cancer mortality that results from a specified change in lung cancer incidence (and change in distribution of lung cancer stages) after screening?
Contextual questions will not be systematically reviewed and are not shown in the Analytic Framework.
1. What are the barriers to implementation of screening for lung cancer and surveillance in U.S. clinical practice (e.g., barriers to shared decisionmaking, systematically eliciting and documenting a detailed smoking history, systems for tracking nodules and followup, and availability of appropriate low-dose CT protocols)?
2. a. Are the participants of lung cancer screening trials (e.g., NLST) who reported a reduction in all-cause or lung cancer mortality representative of persons eligible for lung cancer screening in the United States (based on USPSTF recommendations or NLST criteria)?
b. How do the 5-year survival rate and life expectancy of persons eligible for lung cancer screening in the United States (based on USPSTF recommendations or NLST criteria) compare with those of NLST participants?
c. Are the settings and health care providers in lung cancer screening trials (e.g., NLST) that reported a reduction in all-cause or lung cancer mortality representative of U.S. health care settings and health care providers?
3. Does screening for lung cancer with low-dose CT have unintended benefits from detection of incidental findings (e.g., coronary artery calcium score, chronic obstructive pulmonary disease, or extrapulmonary nodules) leading to interventions that improve health outcomes?
The Proposed Research Approach identifies the study characteristics and criteria that the Evidence-based Practice Center will use to search for publications and to determine whether identified studies should be included or excluded from the Evidence Review. Criteria are overarching as well as specific to each of the key questions (KQs).
|Populations||KQs 1–5, 8: Asymptomatic adults (age ≥18 years)
KQs 6, 7: Adults (age ≥18 years) with early (stage I) non-small cell lung cancer
|KQs 1–5, 8: Children; persons with symptoms or prior diagnosis of lung cancer
KQs 6, 7: Children; persons with nonprimary lung cancer or other than stage I lung cancer
|Risk prediction||KQ 2: Externally validated models including demographic variables, clinical variables, or biomarkers intended for identifying persons at increased risk who are more likely to benefit from screening||KQ 2: Models including a single variable or biomarker; models not considering smoking and age (i.e., known risk factors for lung cancer)|
|Screening||KQs 1, 3, 4, 8: CT*||KQs 1, 3, 4, 8: No screening, chest x-ray, sputum cytology, and other screening methods|
|Workup or surveillance||KQ 5: CT, biopsy, positron emission tomography, or other tests used after screening|
|Interventions||KQs 6, 7: Surgical resection||KQs 6, 7: Chemotherapy, radiation therapy, and natural therapies; immunotherapy; targeted molecular therapy|
|Comparisons||KQs 1, 3, 8: Chest x-ray, no screening, usual care
KQ 2: USPSTF recommendations or criteria used by trials showing benefit (e.g., NLST)
KQs 4, 5: Chest x-ray, no screening, usual care, or no comparison group
KQs 6, 7: No treatment, usual care, or no comparison
|KQs 1–3, 8: Studies without a comparison group
KQs 6, 7: Comparative effectiveness studies (i.e., head-to-head studies comparing treatments)
|Outcomes||KQ 1a: Incidence of lung cancer (all stages), distribution of lung cancer types and stages
KQ 1b: All-cause mortality, lung cancer mortality, and quality of life
KQ 2: Estimated preventable lung cancer deaths or all-cause mortality; estimated screening effectiveness (e.g., number needed to screen); estimated screening harms
KQ 3: Sensitivity, specificity, and predictive value
KQ 4: Radiation exposure, false-positive results, overdiagnosis,† smoking cessation rates, psychosocial harms, and incidental findings leading to additional tests and subsequent harms
KQ 5: Radiation exposure, false-positive results, overdiagnosis,† smoking cessation rates, psychosocial harms, incidental findings leading to additional tests and subsequent harms, and harms of workup (e.g., adverse events from biopsy)
KQ 6: 5- and 10-year incidence of advanced disease and mortality (i.e., survival rates)
KQ 7: Harms of treatment, including mortality, infection, and bleeding
KQ 8: All-cause and lung cancer mortality
|Study designs||All KQs: Controlled trials
KQ 2: Modeling studies; clinical prediction tools are required to include multiple factors
KQ 3: Studies evaluating accuracy
KQs 4, 5, 7 (harms): Prospective cohort studies and case-control studies
KQ 6: Prospective cohort studies
|All KQs: All other study designs‡
KQ 2: Models including a single variable or biomarker; models not considering smoking and age (i.e., known risk factors for lung cancer)
KQs 1–5, 8: Studies with a sample size less than 1,000
KQs 6, 7: Studies with a sample size less than 500
|Study duration||KQs 1–5, 7, 8: Studies of any length
KQ 6: Studies with at least 5 years of followup
|KQ 6: Studies with less than 5 years of followup|
|Settings||Studies published in or after 2001|
|Countries||Studies conducted in countries categorized as “Very High” on the 2016 Human Development Index (as defined by the United Nations Development Programme)||Studies conducted in countries that are not categorized as “Very High” on the 2016 Human Development Index|
|Language||English||Languages other than English|
|Study quality||Good- or fair-quality studies||Poor-quality studies (according to design-specific USPSTF criteria)|
* The evidence review will focus on CT but will also search for and include new trials (published since the search cutoff dates of the last review) of other screening methods. Older studies (from before 2013) of other screening methods will not be carried forward to this update.
† Defined as detection of disease that would never progress to produce symptoms or death.
‡ Systematic reviews are excluded from the evidence review. However, separate searches will be conducted to identify relevant systematic reviews and the citations of all studies included in those systematic reviews will be reviewed to ensure that all relevant primary studies have been captured.
Abbreviations: CT=computed tomography; NLST=National Lung Screening Trial.
The USPSTF has commissioned a decision model to supplement the systematic evidence review on screening for lung cancer, as it did for the previous topic update in 2014. The decision model is a mathematical simulation that projects the health outcomes that result from alternative interventions for screening, diagnosis, and treatment. In conjunction with the evidence review, the decision model will help the USPSTF examine the benefits and harms of screening for lung cancer at the population level, reflecting current nodule management and followup guidelines, while accounting for the variability in implementation patterns across the United States. The decision model will also evaluate different starting and stopping ages, screening frequencies, and screening strategies based on individual lung cancer risk, particularly for population subgroups at higher risk for lung cancer.
For more information about how the USPSTF uses decision models in formulating its recommendations, please refer to the USPSTF procedure manual, available at https://uspreventiveservicestaskforce.org/Page/Name/procedure-manual.