Authors Van den Bruel A, Jones C, Yang Y, et al.
Review Date March 2015
Citation BMJ 2015;350:h980 doi: 10.1136/bmj.h980 (Published 3 March 2015)
Cancer screening in general, and PSA-based prostate cancer screening in particular, has the problem of over-detection, defined as the detection of a cancerous lesion through screening that would otherwise not have caused any symptoms or early death. Due to the nature of prostate cancer, that is often slowly progressing, and the fact that treatments have serious side-effects, the harms of over-detection followed by over-treatment are significant, making the benefit/harm balance of PSA testing controversial. Most guidelines suggest PSA testing should only take place in the context of informed decisions based on information on harms and benefits. Many people do not know about, nor understand the issue of over-detection when undergoing cancer screening and it is not clear what level of over-detection they might find acceptable.
To investigate the level of over-detection people might find acceptable in screening for breast, prostate, and bowel cancer and how acceptability is influenced by the magnitude of the benefit from screening and the cancer-specific harms from over-detection.
An online survey was developed for people aged 18 years or over in the UK who were part of a panel of volunteers for online research. The sample (n=1000) was chosen to be representative of age/sex distribution of the UK population and invited by email or online marketing. Women were presented with scenarios on breast and bowel cancer, men with scenarios on prostate and bowel cancer. For each cancer, epidemiological information, descriptions of treatment and the consequences of treatments were provided. Two different scenarios of benefit were given: a 10% reduction in cancer specific mortality as a result of screening and a 50% reduction.
The main outcome measure was number of cases of over-detection people were willing to accept, ranging from 0 to 1000 (complete screened population) for each cancer type and each scenario of benefit.
Data were analysed to produce medians and 95% confidence intervals and logistic regressions were used to explore whether participant characteristics were associated with either very low or very high acceptance of over-detection. Thematic analysis was done on free text comments for those with very high or very low acceptance of over-detection.
Mean age of the respondents was 46.9 years and 49% of the sample were men. 54% had participated in cancer screening at least once in their lifetime (only 34% of men). 29.2% had heard of over-detection before, mostly from the media (44%) or the doctor (28%). Only 2 reported having read about over-detection in a screening leaflet.
There was large variability between respondents in the level of over-detection they would find acceptable, with medians ranging from 113 to 313 cases of over-detection per 1000 people screened. Acceptability of over-detection for a 50% reduction in mortality was significantly higher than for 10% reduction. Across all scenarios, 4-7% of respondents indicated they would accept no over-detection at all compared with 7-14% who thought that it would be acceptable for the entire screened population to be over-detected.
Acceptability in screening for bowel cancer was significantly lower than for breast (women) or prostate cancer (men). People aged 50 or over accepted significantly less over-detection, whereas people with higher education levels accepted more.
Qualitative analysis showed the main theme from those who would accept high over-detection was that it’s better to save one life even if there are negative consequences; however they were surprised that over-detection in screening exists. The main theme arising from those with a low acceptance for over-detection was that the negative consequences outweighed the benefit of saving a small number of lives; they were also surprised that over-detection was a problem and thought more should be done to inform the public, and education about symptoms of cancer could be an alternative to screening.
There are highly variable views of the level of over-detection in cancer screening that is considered acceptable. Lack of knowledge on over-detection seriously impairs informed decision-making. Invitations for screening should include clear information on the likelihood and consequences of over-detection to facilitate informed choice. Work is needed on how best to inform people of the benefits and risks of cancer screening.
Points to Note
- Cancer screening is associated with over-detection and over-treatment (PSA testing in particular)
- Informed decision making is impaired by lack of knowledge or understanding of over-detection and its consequences.
- Patient information on screening tests often neglects to discuss the issue of over-detection making informed decision making difficult.
- The level of over-detection acceptable to participants in this survey was highly variable suggesting a discussion about over-detection is important so that screening is done in the context of a individual’s values and priorities.
- The limitations of the study included: the sample was better educated and less ethnically diverse than the total UK population; the researchers found it was difficult to accurately quantify the harms of over-detection; the questions may not have been fully understood by the respondents, although free-text comments suggested a reasonable level of understanding.