Should we start population screening for prostate cancer? Randomised trials are still needed
Bibliographic record
Abstract
A decrease in prostate cancer mortality has been reported in the United States, following a rapid increase in the early 1990s. 1 There has also been a report of a 42% reduction in prostate cancer mortality in Tyrol, Austria, where a large-scale prostate cancer screening programme was introduced in 1993, 2 and a report by Chirpaz et al. in this issue of a reduction in prostate cancer mortality in 5 areas of France. 3 In both the United States and France, the reductions in prostate cancer mortality followed large increases in prostate cancer incidence, almost certainly largely due to the impact of PSA screening from lead time and perhaps overdiagnosis.Is such evidence sufficient to warrant adoption of prostate cancer screening as part of routine health care?Decisions regarding prostate cancer screening require very firm evidence of its effects since the costs, both economic and in terms of quality of life, of adoption of screening in the general population would be substantial.Once started as a health care policy, it would be extremely difficult to discontinue.In order to recommend population screening for prostate cancer, convincing evidence is required for both effectiveness in terms of prostate cancer mortality reduction and beneficial net effect, i.e., demonstration that benefits exceed adverse effects such as complications of treatment and overdiagnosis.Currently, there is not sufficient evidence, as we argue below.The reports so far have been ecologic analyses based on aggregate data, which show a decrease in prostate cancer mortality following introduction of PSA testing in a population.In the United States and France, the comparisons were temporal, whilst those in Austria were geographic.We argue that PSA testing may not have caused the benefit.Individual-level data were not available, and only some men in the "screened" population were in fact screened.It remains unclear if the benefit was experienced by men with screen-detected cancers rather than all men with prostate cancer, such as might be caused by the increased interest in and better treatment of prostate cancer that may have accompanied the introduction of PSA testing.In all 3 countries, prostate cancer mortality declined within a few years of introduction of PSA screening.This means that any effect attributable to screening must be due to early deaths avoided.This is puzzling because the mean lead time for PSA screening (the time screening advances diagnosis in time) is at least 5 years, 4,5 while median survival of localised prostate cancer is more than 10 years compared to about 2.5 years for advanced prostate cancer.6 Therefore, no effect would be anticipated for at least 5 years from commencement of screening, apart from that due to earlier detection and treatment of advanced prostate cancers.An early effect could result from more effective treatment of screen-detected advanced cancers, but a substantial improvement is unlikely, as most treatment trials have found only modest survival benefits for early endocrine treatment of clinically detected advanced prostate cancer.[7][8][9] As ecologic comparisons cannot exclude deaths among cases diagnosed before screening commenced, the observed reduction could also be due to improved survival among these cases, resulting from more aggressive treatment.This is illustrated by the fact that a similar decrease in prostate cancer mortality has been observed in the United King-
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.052 | 0.116 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.012 | 0.012 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.006 | 0.009 |
| Open science | 0.004 | 0.001 |
| Research integrity | 0.014 | 0.012 |
| Insufficient payload (model declined to judge) | 0.039 | 0.006 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".