Endovascular versus Open Repair of Abdominal Aortic Aneurysm
2010 · paper · Clinical trials
A UK randomized trial of 1252 patients with large abdominal aortic aneurysms, comparing endovascular with open repair. Endovascular repair had lower 30-day mortality (1.8% vs 4.3%), but long-term death rates were similar, with more complications, reinterventions and cost.
1 finding
Reference or label Table 1 labels the cholesterol unit as "Serum cholesterol (608 and 601 patients) — μmol/liter" with values 5.1±1.2 and 5.1±1.1. The values and the footnote factor (divide by 0.02586) are for mmol/liter.
- p. 3 Table 1, serum cholesterol row
Serum cholesterol (608 and 601 patients) — μmol/liter
- p. 3 Table 1 footnote *
To convert the values for cholesterol to milligrams per deciliter, divide by 0.02586.
Judge's note: The page image shows μmol/liter (not an extraction artefact). The footnote says to divide by 0.02586 to get mg/dl. That gives about 197 mg/dl from 5.1 in mmol/liter, which is plausible; 5.1 μmol/liter would be about 0.2 mg/dl. The unit label is wrong, and it was probably carried over from the creatinine row above.
Second check: another model agreed
Report a problem with this finding
5 minor (typos and the like; most checked by the judge only)
Substance Table 1 history of cardiac disease, open-repair group: "261 (41.8)". 261/626 = 41.7%.
- p. 3 Table 1, history of cardiac disease, open repair column
261 (41.8)
Only the authors can say what is right here.
Judge's note: The page image shows 261 (41.8). This row gives no reduced denominator, and footnote * says data were available for all patients unless numbers are shown, so the denominator is 626. 261/626 = 41.69%, which rounds to 41.7, not 41.8. The adjacent endovascular value, 269/626 = 43.0%, checks out. Either the count or the percentage is off by one; only the authors can say which.
Second check: another model agreed
Report a problem with this finding
Wording Abstract Methods give the end of follow-up as "until the end of 2009", but Results (p.4) say "Patients were followed until September 1, 2009 (minimum, 5 years; maximum, 10 years)".
- p. 1 Abstract, Methods
reinterventions, and resource use until the end of 2009
- p. 4 Results, Patients
Patients were followed until September 1, 2009
- p. 1 Abstract, Methods
until the end of 2009
Only the authors can say what is right here.
Judge's note: Checked both pages in the text layer and on the page images (p.1 abstract; the p.4 crop). Recruitment ran from Sept 1, 1999 to Aug 31, 2004, so 5- and 10-year limits both end on Sept 1, 2009, which fits the Results date. The abstract applies 'end of 2009' to all outcomes, including death. Methods (p.2-3) and the rest of the article state no separate cutoff for complications or resource use. Separate cutoffs may exist but are not stated, so only the authors can say which date is right.
Second check: another model agreed
Report a problem with this finding
Wording Table 3, reintervention >4 yr, unadjusted HR "3.12 (1.47–6.80)": the point estimate is not at the log-scale midpoint of the printed interval (about 3.16).
- p. 6 Table 3, Reintervention >4 yr, unadjusted HR
3.12 (1.47–6.80)
Only the authors can say what is right here.
Judge's note: The page image confirms 3.12 (1.47-6.80). The geometric midpoint is sqrt(1.47*6.80) ≈ 3.16. Even at the rounding extremes of the bounds (1.465, 6.795) the midpoint is about 3.155, so 3.12 cannot be reached. The neighbouring entries are consistent: 3.24 (1.48-7.11) gives about 3.24, 9.97 (4.29-23.15) about 9.97, and 9.12 (3.90-21.3) about 9.11. The Methods do not say what kind of interval was used, so this is a query to the authors rather than a correction.
Second check: another model agreed
Report a problem with this finding
Typo Table 3, reintervention >6 mo–4 yr adjusted HR "9.12 (3.90–21.3)": the upper bound has one decimal, while the other bounds in the table have two.
- p. 6 Table 3, Reintervention, >6 mo–4 yr, adjusted HR
9.12 (3.90–21.3)
Judge's note: The page image confirms 21.3. Every other bound in Tables 2 and 3 has two decimals (for example 23.15, 22.72, 13.09). The value fits the estimate (sqrt(3.90*21.3) ≈ 9.11), so this is only a formatting inconsistency.
Second check: none (judge only)
Report a problem with this finding
Typo Appendix: "James Cook University Hospital, Middlesborough (19)". The town is spelled Middlesbrough.
- p. 8 Appendix, Regional Trial Investigators Committee
James Cook University Hospital, Middlesborough (19)
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the town is spelled 'Middlesbrough'. The page image (p.8) shows 'Middlesborough' as quoted.
Second check: none (judge only)