Effect of Colonoscopy Screening on Risks of Colorectal Cancer and Related Death
2022 · paper · Clinical trials
A pragmatic randomized trial of invitation to a single screening colonoscopy in 84,585 people aged 55–64 in Poland, Norway and Sweden. Ten-year colorectal cancer risk was 18% lower in the invited group; colorectal cancer death and all-cause death did not differ significantly.
Checked: PDF from pure.amsterdamumc.nl, fetched 5 Oct 2026. Page numbers are those of that PDF.
3 findings
Substance Table 1 header "Participants in Poland (N=54,258)"; the column rows sum to 54,528 and only 54,528 reconciles with the overall N = 84,585.
- p. 6 Table 1, column header 'Participants in Poland'
(N = 54,258)
- p. 6 Table 1, Poland column, Group rows
18,184
- p. 6 Table 1, column header
Participants in Poland (N=54,258)
Judge's note: On the page 6 image the header reads 54,258. Invited plus usual care, female plus male, and the two age groups each sum to 54,528. With 54,528 the three country Ns add up to 84,585, and 54,927 eligible minus 54,528 gives 399 exclusions, which fits the stated total of 594. The header digits are transposed.
Second check: another model agreed
Report a problem with this finding
Substance Table 1, 'Good or very good bowel preparation', All Participants: "110,610/11,635 (91.2)"; the numerator should be 10,610.
- p. 6 Table 1, Good or very good bowel preparation, All Participants
110,610/11,635 (91.2)
Judge's note: The page 6 image shows 110,610, which is larger than its denominator. The country numerators sum to 10,610, and 10,610/11,635 gives the printed 91.2%.
Second check: another model agreed
Report a problem with this finding
Substance Figure 1 numbers at risk at year 8 ("24,000" invited, "47,769" usual care) are far below Figure 3's year 8 values (25,273; 50,356), by more than the total colorectal cancer diagnoses (259; 622).
- p. 7 Figure 1 No. at Risk, year 8
24,000
- p. 7 Figure 1 No. at Risk, year 8, usual care
47,769
- p. 8 Figure 3 No. at Risk, year 8
25,273
Only the authors can say what is right here.
Judge's note: I read both risk tables on pages 7 and 8. Figure 1 censors additionally at colorectal cancer diagnosis, so its counts at risk can fall below Figure 3's by at most the cumulative number of diagnoses (259 invited, 622 usual care). At years 2, 4, 6 and 10 the differences are 84-130 (invited) and 94-291 (usual care). At year 8 they are 1273 and 2587, which is impossible under that rule. The round figure 24,000 is also suspect. Only the authors can supply the correct year 8 values.
Second check: another model agreed
Report a problem with this finding
6 minor (typos and the like; most checked by the judge only)
Wording Death from any cause: 10-year risks "11.03%" vs "11.04%" printed with "risk ratio, 0.99" (abstract, Results, Table 2); the ratio of these risks rounds to 1.00.
- p. 8 Table 2, From any cause row
0.99 (0.96 to 1.04)
- p. 2 Abstract, Results
11.03% in the invited group and 11.04% in the usual-care group (risk ratio, 0.99;
- p. 7 Results, Colorectal Cancer–Related Death and Death from Any Cause
(risk ratio, 0.99; 95% CI, 0.96 to 1.04)
Only the authors can say what is right here.
Judge's note: I checked the abstract (page 2), the Results (page 7) and Table 2 (page 8). The other Table 2 rows equal the ratio of the printed risks (0.98/1.20 gives 0.82; 0.28/0.31 gives 0.90). For all-cause death, even the most extreme unrounded values behind 11.03% and 11.04% give at least 0.998. The printed risk difference of -0.01 matches the risks. The authors should say how the 0.99 was derived (for example, truncation or a different estimator).
Second check: another model agreed
Report a problem with this finding
Wording Cecum intubation "11,470/11,843 (96.8)" and text 'intubated in 96.8%': 11,470/11,843 = 96.85%, which rounds to 96.9%.
- p. 6 Table 1, Cecum intubation, All Participants
11,470/11,843 (96.8)
- p. 6 Results, Colonoscopy Screening
intubated in 96.8% of the colonoscopies
Judge's note: The page 6 image shows 11,470/11,843 (96.8), and the text says 96.8%. The quotient is 96.8505%, which is not a tie, and other cells in the table are rounded rather than truncated (2830/8585 is shown as 33.0). The difference is 0.1 percentage point.
Second check: another model agreed
Report a problem with this finding
Typo Repository cover page reads "Citation for pulished version (APA):" (should be 'published').
- p. 1
Citation for pulished version (APA):
Judge's note: The crop of the page 1 cover sheet shows 'pulished'. This is a misspelling on the repository cover page, which is within the in-scope pages 1-11.
Second check: none (judge only)
Report a problem with this finding
Typo Discussion: "the absolute risks of the risk of colorectal cancer" contains redundant wording.
- p. 9 Discussion, second column
the absolute risks of the risk of
- p. 9 Discussion, paragraph 4
the absolute risks of the risk of colorectal cancer
Judge's note: The page 9 crop shows the phrase as printed. It is a copy-editing slip and the meaning is recoverable.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliations: "County Council of Värmland, Karlsdad (L.E.)"; the city is Karlstad.
- p. 10 Appendix, authors' affiliations
County Council of Värmland, Karlsdad (L.E.)
- p. 10 Appendix, affiliations
Karlsdad (L.E.)
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the seat of Värmland County Council is the city of Karlstad. The page 10 crop shows 'Karlsdad'.
Second check: none (judge only)
Report a problem with this finding
Typo Reference 8 "JAMA Intern Med 2016;176:894." gives only a start page, whereas the other journal references give page ranges.
- p. 11 Reference 8
JAMA Intern Med 2016;176:894.
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the article spans pages 894-902. The page 11 reference list shows ref 8 ending at '176:894.', while the other print-journal references give full ranges.
Second check: none (judge only)