Quantifying the Harms of Overdiagnosis?
17 September 2013
A central tenant of the Hippocratic Oath is “Do No Harm” and a recent article brought to my attention how few clinical trials aimed at early cancer detection measure the harms to patients in the screening process (1). (downloaded for free here: http://www.bmj.com/content/347/bmj.f5334)
There is a huge push for early detection in cancer. Groups like the Canary Foundation and the American Cancer Society have early detection squarely in their mission ethos and have aggressively emphasized self-detection and funded research for early detection trials and technologies.
A late 20th century view of cancer progression is that of a long road with distant metastasis (the life-threatening threshold in disease progression) at the end (2), offering a hypothetical window of intervention… if that window was opened by screening. These views are still dominant in many circles, but (at least in the instance of breast cancer) have recently been challenged by a rising tide of clinical trials, epidemiological data, and experimental models of metastasis. Late metastasis is far from universal in solid tumor progression.
Nonetheless, the hypothesis that early detection might lead to a decrease in cancer-associated deaths was rigorously pursued in the last decade, and has resulted in some major changes in screening policy. The culmination of these data resulted in the UK Panel on Breast Cancer Screening recommending a decrease in screening (3). While early stage diagnosis of breast cancer increased with increased early detection measures, the overall reduction in cancer related deaths were so low that it was judged as unethical to screen so many women.
Yesterday, a paper came out in the British Journal of Medicine that took this a step further: the authors attempted to quantify the harms to patients from screening. Harms can come in a variety of forms, and they include (but are not limited to):
-Invasive procedures for early detection (body fluid sampling, biopsies, etc). -Overdiagnosis (and unnecessary chemical or surgical procedures). -Unnecessary psychological harm from misdiagnosis of a condition that is not physically harmful.
Results Out of 4590 articles assessed, 198 (57 trials, 10 screening technologies) matched the inclusion criteria. False positive findings were quantified in two of 57 trials (4%, 95% confidence interval 0% to 12%), overdiagnosis in four (7%, 2% to 18%), negative psychosocial consequences in five (9%, 3% to 20%), somatic complications in 11 (19%, 10% to 32%), use of invasive follow-up procedures in 27 (47%, 34% to 61%), all cause mortality in 34 (60%, 46% to 72%), and withdrawals because of adverse effects in one trial (2%, 0% to 11%). The median percentage of space in the results section that reported harms was 12% (interquartile range 2-19%). Conclusions Cancer screening trials seldom quantify the harms of screening. Of the 57 cancer screening trials examined, the most important harms of screening—overdiagnosis and false positive findings—were quantified in only 7% and 4%, respectively.
If the numbers were not poignant enough, the authors offered a few comments on the matter:
While we acknowledge that collecting data on harms will complicate cancer screening trials, this is not a sound argument against the strong ethical obligation to collect such data. If trialists do not report certain outcomes because they consider that the harms will be either rare or irrelevant when compared with the potential decrease in mortality, such information will not be available for people who judge these outcomes differently. We think that future screening trials should collect and report the expected harms of screening (false positives, overdiagnosis and overtreatment, psychosocial consequences, somatic complications, and all cause mortality). Adequate reporting of harm requires data from the control group as these provide a reference level and help to interpret harms data from the screened group.
Not addressed by the authors is the argument that perhaps increased screening could offer a peace of mind for the concerned. Another recent review article (4) examined the psychological outcomes of patients in response to screening for rare diseases in otherwise healthy patients and found that:
Diagnostic tests for symptoms with a low risk of serious illness do little to reassure patients, decrease their anxiety, or resolve their symptoms…
The take-home message that I get out of this is that we must not assume that any intervention will not be harmful, and that we need to design trials to both quantify this effect and ethically weigh this against data obtained from the results of trials of early detection and beyond. Without such reporting there is no way to know if we are doing more harm than good.
Ryon 1. Heleno B, Thomsen MF, Rodrigues DS, Jørgensen KJ, Brodersen J. Quantification of harms in cancer screening trials: literature review. BMJ. 2013 2013-09-16 11:34:45;347. 2. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000 Jan 7;100(1):57-70. PubMed PMID: 10647931. Epub 2000/01/27. eng. 3. The benefits and harms of breast cancer screening: an independent review. Lancet. 2012 Nov 17;380(9855):1778-86. PubMed PMID: 23117178. Epub 2012/11/03. eng. 4. Rolfe A, Burton C. Reassurance after diagnostic testing with a low pretest probability of serious disease: systematic review and meta-analysis. JAMA internal medicine. 2013 Mar 25;173(6):407-16. PubMed PMID: 23440131. Epub 2013/02/27. eng.