[Skip to Navigation]
Special Communication
Less Is More
February 2014

The Harms of Screening: A Proposed Taxonomy and Application to Lung Cancer Screening

Author Affiliations
  • 1Research Center for Excellence in Clinical Preventive Services, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill
  • 2Department of Medicine, School of Medicine, University of North Carolina at Chapel Hill
  • 3Department of Family Medicine, School of Medicine, University of North Carolina at Chapel Hill
  • 4Department of Health Behavior, Gillings School of Global Public Health, University of North Carolina at Chapel Hill
JAMA Intern Med. 2014;174(2):281-286. doi:10.1001/jamainternmed.2013.12745

Importance  Making rational decisions about screening requires information about its harms, but high-quality evidence is often either not available or not used. One reason may be that we lack a coherent framework, a taxonomy, for conceptualizing and studying these harms.

Objective  To create a taxonomy, we categorized harms from several sources: systematic reviews of screening, other published literature, and informal discussions with clinicians and patients. We used this information to develop an initial taxonomy and vetted it with local and national experts, making revisions as needed.

Results  We propose a taxonomy with 4 domains of harm from screening: physical effects, psychological effects, financial strain, and opportunity costs. Harms can occur at any step of the screening cascade. We provide definitions for each harm domain and illustrate the taxonomy using the example of screening for lung cancer.

Conclusions and Relevance  The taxonomy provides a systematic way to conceptualize harms as experienced by patients. As shown in the lung cancer screening example, the taxonomy also makes clear where (which domains of harms and which parts of the screening cascade) we have useful information and where there are gaps in our knowledge. The taxonomy needs further testing and validation across a broad range of screening programs. We hope that further development of this taxonomy can improve our thinking about the harms of screening, thus informing our research, policy making, and decision making with patients about the wisdom of screening.

Add or change institution
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

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.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    3 Comments for this article
    Comments on the harms of lung cancer screening
    David L. Keller, MD, MS | none
    In their discussion of the harms of screening for lung cancer, Harris and colleagues discuss harms associated with earlier initiation of treatments which would have been started later or not at all without screening. The authors state that screening and earlier treatment delays death for only 20% of patients destined to die from lung cancer. This raises the question of why the other 80% of patients are being exposed to ineffective treatments and the associated harms. Is the problem that we are not able to identify which patients will be among the fortunate 20% to postpone death by being treated? If so, then studies should be done to learn which characteristics differentiate patients who benefit from these treatments (e.g. tumor stage, grade, patient functional status, genetic profile or other factors). Perhaps these treatments are given to the unfortunate 80% for the purpose of palliation of symptoms, with the knowledge that there will be no mortality benefit. If so, then the treatments are beneficial to the patient despite failing to prolong life. The authors state that the 80% of patients who are identified as having fatal cancers which cannot be treated to prolong life suffer harms from screening by virtue of their early diagnosis, in that they will have longer to live with the knowledge of their fate, causing anxiety and other psychological harms. However, other patients might strongly prefer to be informed as early as possible of their fatal diagnosis, in order to make better use of their remaining time alive and pain-free; for them, the earlier diagnosis afforded by screening is not a harm, but a benefit.
    Contribution to the taxonomy on the harms of screening
    Bruno Heleno MD *, Karsten Juhl Jørgensen MD DrSciMed #, John Brodersen MD PhD * | * Research Unit for General Practice and Section of General Practice - University of Copenhagen, # The Nordic Cochrane Centre
    We commend Harris and colleagues for starting a timely discussion on how to classify harms associated with screening. They stress that harms happen at all levels of the screening cascade, and include often downplayed outcomes, such as physical discomfort and condition-specific distress, in their four domains: physical harms, psychological harms, financial strain and opportunity cost.(1) We have performed a systematic review on harm reporting in cancer screening trials,(2) and we struggled with the lack of an accepted taxonomy.In retrospect, we wish we had distinguished between mechanisms of harm and harmful outcomes in the protocol of our review. The three main reasons for this are also relevant forthe taxonomy proposed by Harris and colleagues. First, a single mechanism can lead to harmful outcomes in all harm domains of the taxonomy. For example, a false-positive (mechanism) can cause scarring (physical outcome), feeling less attractive (psychosocial outcome), sick leave (financial outcome) and missing appointments with friends and family (opportunity cost outcome). Second, a single harmful outcome can have several mechanisms; e.g. feeling less attractive (harmful outcome) can be a consequence of either a false-positive finding, an incidental finding, or overdiagnosis (mechanisms). For the individual, it is more relevant to know whether the harm exists, what is its likelihood and its severity than the mechanism that led to it. Third, recognising the mechanisms of harm may be useful when researchers are exploring unintended consequences of screening, or as surrogate markers of harmful outcomes. However, when researchers want to estimate the effects (beneficial or harmful) of screening they need very specific, person-oriented outcomes that can be measured in the intervention and control groups. It is difficult to speak of false-positives in an unscreened population (mechanism), while it is possible to estimate the average ‘unattractiveness feeling’ in both groups (harmful outcome). In summary, we find the four domains in the proposed taxonomy appealing. To bring their characteristics closer to biological taxonomy, we think that each harm should belong only to one domain. This could be more easily achieved if harm mechanisms are separated from harm outcomes. 1. Harris RP, Sheridan SL, Lewis CL, Barclay C, Vu MB, Kistler CE, et al. The Harms of Screening: A Proposed Taxonomy and Application to Lung Cancer Screening. JAMA Intern Med. 2014 Feb 1;174(2):281. 2. Heleno B, Thomsen MF, Rodrigues DS, Jorgensen KJ, Brodersen J. Quantification of harms in cancer screening trials: literature review. BMJ. 2013 Sep 16;347(sep16 1):f5334–f5334.
    Illustration of harms in the proposed taxonomy
    Jakob Fraes Rasmussen MD, Bruno Heleno MD, John Brodersen MD PhD | Research Unit for General Practice and Section of General Practice - University of Copenhagen
    In their recent paper about the taxonomy for harms of screening Harris and colleagues stress that the financial strain and opportunity costs not only come from the screening test but also from downstream diagnostic and therapeutic procedures.(1) In the Danish Lung Cancer Screening Trial (DLCST), we have shown that participants with positive screens (both true positives and false positives) had higher healthcare utilisation and higher costs, not only in the screening clinic and subsequent hospital care, but also in the primary care sector.(2) This provides further evidence to the opportunity costs and financial costs category in the taxonomy. Using national registries, we are now planning to assess the impact of the screening programme in terms of sick leave and early retirement on the trial participants.Moreover, Harris and colleagues state that they did not find any research exploring the harm for the participants (and their significant others) with abnormal results who are in an uncertain state for a prolonged period between the screening scan and the final diagnosis (system of surveillance). We did focus group interviews with screening participants having abnormal and false-positive CT-screening results.(3) Therefore, psychosocial consequences of being in the critical period from abnormal screening results until final diagnosis has been studied and published.(3) However, to which extent and for how long these psychosocial consequences may last have not yet been published. We did collect such data in the DLCST and are planning to publish our results in the near future. Finally, Harris and colleagues reported that no studies examining the psychological (or physical) effects of workup of incidental findings and discovery of incidental disease were identified. We are also aware of the lack of evidence in this area and we plan to publish papers about these topics from data collected via the DLCST.