Quantification of harms in cancer screening trials: literature reviewBMJ 2013; 347 doi: https://doi.org/10.1136/bmj.f5334 (Published 16 September 2013) Cite this as: BMJ 2013;347:f5334
- Bruno Heleno, PhD fellow1,
- Maria F Thomsen, registrar1,
- David S Rodrigues, consultant general practitioner2,
- Karsten J Jørgensen, senior researcher3,
- John Brodersen, associate professor1
- 1Research Unit for General Practice and Section of General Practice, Department of Public Health, University of Copenhagen, Øster Farimagsgade 5, PO Box 2099, 1014 Copenhagen K, Denmark
- 2Family Medicine Department, Nova Medical School, Campo dos Mártires da Pátria 130, 1169-056 Lisbon, Portugal
- 3Nordic Cochrane Centre, Rigshospitalet, Department 7811, Blegdamsvej 9, 2100 Copenhagen, Denmark
- Correspondence to: B Heleno
- Accepted 16 August 2013
Objectives To assess how often harm is quantified in randomised trials of cancer screening.
Design Two authors independently extracted data on harms from randomised cancer screening trials. Binary outcomes were described as proportions and continuous outcomes with medians and interquartile ranges.
Data sources For cancer screening previously assessed in a Cochrane review, we identified trials from their reference lists and updated the search in CENTRAL. For cancer screening not assessed in a Cochrane review, we searched CENTRAL, Medline, and Embase.
Eligibility criteria for selecting studies Randomised trials that assessed the efficacy of cancer screening for reducing incidence of cancer, cancer specific mortality, and/or all cause mortality.
Data extraction Two reviewers independently assessed articles for eligibility. Two reviewers, who were blinded to the identity of the study’s authors, assessed whether absolute numbers or incidence rates of outcomes related to harm were provided separately for the screening and control groups. The outcomes were false positive findings, overdiagnosis, negative psychosocial consequences, somatic complications, invasive follow-up procedures, all cause mortality, and withdrawals because of adverse events.
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.
We thank Rajeswari Aghoram, James Dickinson, and Yuk Tsan Wun for providing details about the search strategy used in the Cochrane Systematic Review of liver cancer screening. We thank Michael Nixon for helpful comments on an earlier draft.
Contributors: BH drafted the protocol, and KJJ and JB provided comments. BH and DSR assessed references for eligibility. BH and MFT extracted data. BH analysed data and drafted the manuscript. MFT, DSR, KJJ, and JB contributed to revisions with important intellectual content. All authors had full access to all data (including statistical reports and tables) in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. BH is guarantor.
Funding: BH is supported by Fundação para a Ciência e Tecnologia (governmental agency) grant SFRH/BD/74640/2010. The funder had no role in study design or data collection, analysis, or interpretation.
Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.
Ethical approval: Not required.
Data sharing: Comma separated files (csv) with the extracted data and the R script with the analyses are available from the authors.
This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 3.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/3.0/.