eFigure 1: Screenshots of an Imbalanced Embedded Video From a Lung Cancer Screening Website
eFigure 2: Screenshot of Imbalanced Information on a Lung Cancer Screening Website
eFigure 3: Screenshots of Balanced Components of Lung Cancer Screening Websites
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Clark SD, Reuland DS, Enyioha C, Jonas DE. Assessment of Lung Cancer Screening Program Websites. JAMA Intern Med. 2020;180(6):824–830. doi:10.1001/jamainternmed.2020.0111
How do lung cancer screening program websites portray benefits and harms, and what next steps do they recommend for individuals considering screening?
In this cross-sectional study of 162 lung cancer screening program websites, potential benefits of screening were presented significantly more often than potential harms (98% vs 48%). In addition, only 22% of the websites guided individuals to a shared decision-making visit with a health care professional.
There appears to be an imbalance in the presentation of benefits and harms on lung cancer screening program websites and a lack of guideline-driven recommended next steps for shared decision-making.
The US Preventive Services Task Force recommends that individuals at high risk for lung cancer consider benefits and harms before pursuing lung cancer screening. Medical centers develop websites for their lung cancer screening programs, but to date little is known about the websites’ portrayal of benefits and harms or what next steps they recommend for individuals considering screening.
To assess the presentation of potential benefits and harms and recommended next steps on lung cancer screening program websites.
Design, Setting, and Participants
Cross-sectional content analysis of 162 lung cancer screening program websites of academic medical centers (n = 81) and state-matched community medical centers (n = 81) that were randomly selected from American College of Radiology lung cancer screening–designated centers was conducted. The study was performed from December 1, 2018, to January 31, 2019.
Main Outcomes and Measures
Website presentation of screening-associated benefits and harms was the primary outcome. Benefit was defined as any description related to the potential reduction in lung cancer mortality. Harms were based on the US Preventive Services Task Force recommendations and included false positives, false negatives, overdiagnosis, radiation exposure, and incidental findings. The secondary outcome was next steps that are recommended by websites.
Overall, the 162 lung cancer screening program websites described the potential benefits more frequently than they described any potential harms (159 [98%] vs 78 [48%], P < .01). False-positive findings were the most frequently reported (72 [44%]) potential harm. Community centers were less likely than academic centers to report any potential harm (32 [40%] vs 46 [57%], P = .03), potential harm from radiation (20 [25%] vs 35 [43%], P = .01), and overdiagnosis (0% vs 11 [14%], P < .01). One hundred nineteen websites (73%) did not explicitly recommend that individuals personally consider the potential benefits and harms of screening; community centers were less likely than academic centers to give this recommendation (15 [19%] vs 28 [35%], P = .02). Most institutions (157 [97%]) listed follow-up steps for screening, but few institutions (35 [22%]) recommended that individuals discuss benefits and harms with a health care professional.
Conclusions and Relevance
Information on public-facing websites of US lung cancer screening programs appears to lack balance with respect to portrayal of potential benefits and harms of screening. Important harms, such as overdiagnosis, were commonly ignored in the sites evaluated, and most of the centers did not explicitly guide individuals toward a guideline-recommended, shared decision-making discussion of harms and benefits.
The National Lung Screening Trial (NLST) reported a reduction in mortality from lung cancer screening with use of low-dose computed tomographic (CT) scans in high-risk individuals (those aged 55-74 years who were current or former smokers who quit within the past 15 years, and had at least 30 pack-years of smoking).1 Based on this study, screening with low-dose CT scans is now widely recommended in guidelines, including those of the United States Preventive Services Task Force (USPSTF).2-5 However, screening recommendations have been controversial, in part because screening also frequently leads to false-positive findings (resulting in additional radiologic and procedural workup and associated complications), false-negative findings, overdiagnosis, radiation exposure, and incidental findings.6,7 Because of these possible risks, guidelines also emphasize the need for individuals to understand the potential harms of screening. For eligible individuals to attain an understanding of benefits and harms, guidelines recommend that health care professionals engage these individuals in shared decision-making. For example, in their guidelines, the American College of Chest Physicians has described the balance between benefits and harms as tenuous,4 and the USPSTF states that screening “should be the result of a thorough discussion of the possible benefits, limitations, and known and uncertain harms.”2[p333] The American College of Chest Physicians and American Thoracic Society further highlight that the purpose of shared decision-making is to allow individuals to “make informed choices about whether low-dose CT scan screening is right for them,”8[p886] and the Centers for Medicare & Medicaid Services underscored the importance of shared decision-making by requiring a dedicated shared decision-making visit for screening to be covered for Medicare beneficiaries.9
With almost all individuals living in the US using the internet, including 88% of those aged 50 to 64 years and 73% of those older than 65 years, medical centers have created websites with information on health and health programs.10 These websites include information about lung cancer screening and lung cancer screening programs. Concerning lung cancer screening content on such websites, the American Thoracic Society and American College of Chest Physicians state that “any direct-to-consumer advertising…should not be coercive or overly optimistic about potential benefits.”8[p884] These lung cancer screening websites may be the first lung cancer screening information that eligible individuals encounter and may influence their impression of the benefits and harms of screening, screening decision, and how they proceed to the next steps in the process of considering screening. However, to our knowledge, no studies have examined how or whether screening program websites portray the benefits and harms of lung cancer screening or whether the websites explicitly encourage individuals to consider and discuss the harms and benefits of screening as recommended by guidelines. To date, this lack of literature is true for both academic and community medical centers, which may have different institutional objectives or processes in publishing websites, potentially leading to differences in the reporting of information on lung cancer screening. Therefore, we sought to evaluate the portrayal of harms and benefits of lung cancer screening on academic and community screening program websites and their recommended next steps for individuals. We hypothesized that academic centers would be more likely to present balanced information, including description of potential harms.
We performed a standardized web search between December 1, 2018, and January 31, 2019, for lung cancer screening program information on websites of academic and community medical centers. We evaluated the websites for the reporting of benefits, harms, and recommended next steps for eligible individuals. The institutional review board of the University of North Carolina at Chapel Hill evaluated the proposal for this study and deemed it exempt from further review because it only uses publicly available information and does not involve any protected health information.
We used the Google search engine to search for lung cancer screening information available on academic medical center websites affiliated with each of the 152 allopathic medical schools in the US that we identified from the Association of American Medical Colleges.11 For each institution, we entered “lung cancer screening [medical school name]” and evaluated the first page of search results for a webpage containing lung cancer screening program information from that institution. We excluded websites with information presented in languages other than English as well as news sites and blog posts.
We then categorized the list of American College of Radiology lung cancer screening–designated centers (n = 1779) by state. Using a random number generator, we selected a state-matched community medical center for each academic center with a screening program website. State matching was done to control for potential geographic differences. During the matching process, screening sites with the terms medical center or hospital in the name of the site from the American College of Radiology list were identified and searched for screening information through Google in the same manner as the academic centers. Oncology centers, identified as those having cancer or oncology in their name, and radiology imaging centers were excluded because these were distinct entities from the community medical centers. If we found no community medical center with screening information in the same state as an academic center with a screening website, we randomly selected the matched center from the American College of Radiology list of neighboring states.
Using a standardized data extraction form, we recorded whether websites described any benefits of screening overall or a mortality benefit. Both were coded as present or absent. A benefit was defined as any positive language about screening related to preventing lung cancer mortality or all-cause mortality, including phrases such as chance to be cured. Mortality benefit included any language that indicated a decreased chance of death or increased survival. We additionally noted whether websites portrayed the approximate magnitude of benefit based on current evidence and, if so, whether it was presented as a relative risk reduction (eg, 20% reduction in lung cancer mortality) or an absolute risk reduction (eg, 3 lung cancer deaths prevented per 1000 screened, or similar evidence). The presence of the following potential harms, each listed in the USPSTF guidelines, was also recorded as present or absent: any harm overall, false-positives, false-negatives, overdiagnosis, radiation exposure, and incidental findings.2 Radiation was recorded as a harm if it was described with harms language (ie, discussed with other harms or explicitly stating that it could cause cancer) and not if it was only mentioned that low-dose CT scans use lower levels of radiation than regular CT scans. We also recorded whether websites explicitly recommended that individuals consider both the benefits and harms of screening themselves. We further noted categories of next steps that the websites recommended for individuals: generally call the screening center with further questions, call the screening center to schedule a screening appointment, discuss screening with a primary health care professional generally, obtain a screening referral from a primary health care professional, discuss harms and benefits of screening with a primary health care professional, and discuss harms and benefits with a health care professional at the medical center’s lung cancer screening program.
We included data from any information presented in text, audio, or video format within the website. Websites were dually reviewed with conflicts resolved by discussion or a third reviewer if needed. Uncertainties about the presence of information were also discussed and resolved between reviewers. We did not consider the information linked to external websites as being present on the evaluated website. When multiple next steps were listed on a website, we counted the one with the highest ranking on the basis of what we considered to be most appropriate and aligned with national guidelines as follows, favoring explicit recommendations to discuss benefits and harms: (1) discuss harms and benefits with your primary health care professional, (2) discuss harms and benefits with a health care professional at the medical center’s lung cancer screening program, (3) discuss generally with your primary health care professional, (4) generally call the screening center to schedule a screening appointment, (5) call the screening center with further questions, and (6) obtain a referral from your primary health care professional.
We used descriptive statistics to evaluate the reporting of benefits, harms, and next steps. We used bivariate analysis with Pearson χ2 and Fisher exact tests as appropriate to compare reporting on academic and community centers, with an unpaired P value cutoff level of .05 considered statistically significant. Statistical analysis was carried out with Stata, release 15 (StataCorp LLC).
Of the 152 US allopathic medical schools, we identified 81 affiliated centers with lung cancer screening websites. After state-matching with 81 community centers, we included a total of 162 medical centers with lung cancer screening website information. Three states (Minnesota, Nebraska, and New Mexico) and Washington, DC, did not have American College of Radiology lung cancer screening–designated community centers with online lung cancer screening information, so 5 academic centers were matched to community centers in neighboring states.
On the 162 center websites, any benefit of screening was presented more frequently than any harm (159 [98%] vs 78 [48%], P < .01) (Table 1). Academic centers more frequently presented quantitative benefits (relative risk reduction or absolute risk reduction) of the NLST compared with community centers (44 [54%] vs 26 [32%], P = .01). Both academic and community centers were more likely to present the NLST relative risk reduction of 20% than the NLST absolute risk reduction of 3 lung cancer deaths prevented per 1000 people screened (66 [41%] vs 1 [1%], P < .01). False-positives (72 [44%]) were the most frequently reported harm by both types of centers and overdiagnosis (11 [7%]) was the least frequently presented. Academic centers vs community centers more frequently reported any potential harm (46 [57%] vs 32 [40%], P = .03), harms of radiation (35 [43%] vs 20 [25%], P = .01), and overdiagnosis (11 [14%] vs 0%, P < .01). No websites listed all 5 of the harms that we recorded.
One hundred nineteen websites (73%) did not explicitly recommend that individuals consider both the benefits and harms of screening, and community centers were less likely than academic centers to give this recommendation (15 [19%] vs 28 [35%], P = .02) (Table 2). Most websites (157 [97%]) listed follow-up steps for screening, but few (29 [18%]) specifically recommended discussing benefits and harms with an individual’s primary health care professional. Eighty-five of the websites (52%) guided individuals to their primary health care professional in some manner, although 25 of these 85 websites (29%) recommended only that individuals obtain a referral for screening from their primary health care professional. There was no statistically significant difference between academic and community centers in recommended next steps for individuals.
The lung cancer screening program websites of academic and community medical centers evaluated in this study rarely included a balanced message of potential benefits and harms that aligns with current recommendations. Less than half of all websites presented any harm of screening and reporting of individual harms was variable. Websites listed overdiagnosis and incidental findings at a rate of less than 10%. Nearly all websites, however, presented a benefit of lung cancer screening.
These findings reveal a possible lack of balance in presenting benefits and harms of lung cancer screening. The imbalance is consistent with historical approaches to cancer screening advertising, which ignore or minimize potential harms and costs. Experts and organizations, such as the National Cancer Institute, seek to move communication about screening from persuasion to objective information.12 This shift is true in the official policy statement of the American Thoracic Society and American College of Chest Physicians, which states that direct-to-consumer material not be “overly optimistic” about potential benefits.8 The imbalance that we observed also does not seem to align with the language and intent of national guidelines, which recommend that individuals eligible for screening be thoroughly informed of the potential benefits and harms of screening.2-4,13 Excerpts from several websites showed both balanced and imbalanced aspects from both community and academic centers, such as next steps and balanced patient stories of lung cancer–screening decisions. Table 3 provides examples of imbalanced content as well as balanced components.14-21 eFigure 1, eFigure 2, and eFigure 3 in the Supplement depict screenshots.
The lung cancer screening information presented on websites is not necessarily developed and published to serve the same purposes as true decision aids. From the Centers for Medicare & Medicaid Services perspective, responsibility for conducting and documenting shared decision-making ultimately falls to the physician who orders the initial screening test.9 However, this regulatory requirement for a single shared decision-making visit does not capture the true processes of decision-making. In reality, individuals make judgments and construct preferences as soon as they receive information, and they are susceptible to anchoring bias, in which initial information received anchors and overly influences subsequent judgments.22,23 Thus, because these websites may be the only detailed information that patients receive before attending an in-person visit at a specialized screening program, it is reasonable to think that the websites may influence patient assumptions about lung cancer screening. In addition, even with the Centers for Medicare & Medicaid Services requirement for shared decision-making before lung cancer screening, several studies indicate a low frequency of such discussions.13,24
We found that screening websites more often portrayed relative risk reduction (20% relative reduction in lung cancer mortality from the NLST) than absolute risk reduction (0.3% from the NLST). Thus, it may be difficult for individuals to develop an understanding of the low likelihood that they, as an individual, will benefit from screening.25 A systematic review of the presentation of statistical evidence found that, compared with absolute risk reduction, relative risk reduction was perceived to be greater as well as more persuasive.26 Decision-making experts echo this finding, cautioning that the use of relative risk may make the magnitude of potential benefit appear “more impressive—and potentially misleading—than absolute risk reduction.”27[p25] As with the imbalance of potential benefits and harm, any distortion in the presentation of benefits alone is unlikely to be corrected in the clinical context. Shared decision-making discussions are challenging to implement in practice, and clinicians, in general, are more likely to overestimate the benefits of screening.13,24,28,29
As noted, although screening program websites are not responsible for full delivery of guideline-recommended shared decision-making, they can create expectations in screening-eligible individuals and guide their considerations about next steps. We found that only about a quarter of websites encouraged eligible individuals to consider the benefits and harms of screening and even fewer specifically guided individuals to a visit at which the tradeoffs involved with screening would explicitly be considered. Some websites indicated that eligible individuals should simply ask for a referral from their primary health care professional to have the screening done—a recommendation that appears to bypass guideline-recommended shared decision-making. Without being primed to consider the tradeoffs and formulate their own preferences, individuals may assume the next steps are to visit a screening center for the sole purpose of completing low-dose CT scan screening. This concern is heightened by evidence suggesting that, even when tertiary screening centers provide robust, shared decision-making services and use high-quality materials at the time of the in-person visit, virtually all patients will proceed to complete screening.30 This finding contrasts with studies reporting that screening preferences among informed, screening-eligible individuals who view decision aids in nonreferral settings are heterogeneous and that substantial proportions prefer not to be screened.31,32
Our findings suggest the need for policy efforts aimed at helping patients receive timely, balanced information about preference-sensitive cancer screening decisions, including lung cancer screening. In the near term, development of expert consensus and specific standards about what information to publish on websites may help encourage the provision of balanced information across screening centers. Just as standards were created by the International Patient Decision Aid Standards collaboration for decision aids, similar standards could be developed to enhance the quality of websites by establishing a framework for improving their content and evaluation. Like the International Patient Decision Aid Standards, the standard would benefit from including categories such as content, including benefits and harms qualitatively and quantitatively described, and balance in reporting of information, including patient stories.33 Until that time, we believe academic and community medical centers should evaluate the material published on their websites and consider how it can be improved to reflect and facilitate practices recommended in lung cancer screening guidelines.
There are several limitations of this study. First, we did not include oncology centers or radiology-specific imaging centers in our sample, although in a limited search of these types of centers, we noted that websites with lung cancer screening information were sparse. In addition, we did not evaluate material that was available only on external links from our sample websites. Second, we do not know how often these websites are accessed by screening-eligible individuals or at what stage of the screening process. Third, we did not systematically assess whether websites mentioned potential psychological harms of screening, such as distress associated with prolonged follow-up of screening-detected nodules. Fourth, it is possible that some existing websites were missed with our search methods.
Overall, there appears to be an imbalance of harms and benefits reporting on both academic and community medical center lung cancer screening websites. Potential harms of screening seem to be presented much less frequently than benefits, and academic programs appeared to be more likely than community programs to report screening-associated harms.
Accepted for Publication: January 11, 2020.
Corresponding Author: Stephen D. Clark, MD, Division of General Medicine and Clinical Epidemiology, Department of Medicine, University of North Carolina at Chapel Hill, 5034 Old Clinic Building, CB #7110, Chapel Hill, NC 27599 (firstname.lastname@example.org).
Published Online: April 13, 2020. doi:10.1001/jamainternmed.2020.0111
Author Contributions: Drs Reuland and Jonas contributed equally as senior authors. Dr Clark had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Clark, Reuland, Jonas.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Clark.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Clark, Reuland, Enyioha.
Administrative, technical, or material support: Jonas.
Supervision: Reuland, Jonas.
Conflict of Interest Disclosures: None reported.
Funding/Support: Dr Clark’s and Dr Enyioha’s time was supported by the Health Resources & Services Administration–funded primary care research fellowship T32-HP14001 at the University of North Carolina at Chapel Hill
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Meeting Presentation: Data from this paper were presented at the 2019 Society of General Internal Medicine Annual Meeting; May 9, 2019; Washington, DC; at the 2019 International Shared Decision Making Conference; July 9, 2019; Quebec City, Quebec, Canada; and at the 2019 Preventing Overdiagnosis Conference; December 5, 2019; Sydney, Australia.
Additional Contributions: Donald Pathman, MD, MPH, Colin Orr, MD, MPH, Jonathan Yun, MD, MPH, Skyler McLaurin MD, and Andrea Dotson, MD, MSPH through the National Research Service Award primary care research fellowship at the University of North Carolina at Chapel Hill reviewed an earlier draft of the manuscript. None of the contributors received compensation for their assistance.