Effects of Extended-Release Metoprolol Succinate in Patients Undergoing Non-Cardiac Surgery (POISE Trial)

• p. 3 Table 1, Undergoing major vascular surgery, metoprolol

  1500 (36·0%)

Only the authors can say what is right here.

Judge's note: Checked on the page 3 image. 1500/4174 = 35·94%, which rounds to 35·9%. Nearby cells in the same column match N=4174 exactly (for example 1549 gives 37·1% and 260 gives 6·2%). A denominator of 4172 or less, i.e. a few patients with missing data, would give 36·0%, but no missing-data denominator is stated anywhere. Only the authors can say which applies.

Second check: another model agreed

• p. 3 Table 1, Coronary artery disease, metoprolol

  1805 (43·3%)

Only the authors can say what is right here.

Judge's note: Checked on the page 3 image. 1805/4174 = 43·24%. A denominator of 4173 or less would give 43·3%. The paper states no missing-data denominator.

Second check: another model agreed

• p. 3 Table 1, History of hypertension, metoprolol

  2635 (63·2%)

Only the authors can say what is right here.

Judge's note: Checked on the page 3 image. 2635/4174 = 63·13%. A denominator of 4172 or less would give 63·2%. The paper states no missing-data denominator.

Second check: another model agreed

• p. 3 Table 1, Aspirin, metoprolol

  1517 (36·4%)

Only the authors can say what is right here.

Judge's note: Checked on the page 3 image. 1517/4174 = 36·34%. A denominator of 4173 or less would give 36·4%. The paper states no missing-data denominator.

Second check: another model agreed

• p. 7 Table 5, Death, History of congestive heart failure

  535 (6·4%)

• p. 3 Table 1, Any history of congestive heart failure

  260 (6·2%)

Only the authors can say what is right here.

Judge's note: Checked on the page 3 and page 7 images. The other preoperative predictors match Table 1: creatinine 401 = 207+194, emergent surgery 878 = 440+438. No broader CHF definition is stated in the paper. Webappendix 2 is not among the files.

Second check: another model agreed

• p. 7 Table 5

  Age ≥70 years

• p. 3 Table 1

  Age >70 years

Only the authors can say what is right here.

Judge's note: Checked on the images. Table 1 (p3) has '>70', and the methods text has 'age >70 years'. Table 5 (p7) has '≥70'. I found no statement that Table 5 uses a different age cut-off, so the authors should say which threshold is correct.

Second check: another model agreed

• p. 7 Discussion, left column, per-1000 estimates

  three from undergoing cardiac revascularisation

• p. 4 Table 3, Cardiac revascularisation

  11 (0·3%)

Judge's note: Recomputed from the Table 3 counts on the page 4 image: 3·8 per 1000, which rounds to 4. Every per-1000 figure in the paragraph, including 3, is reproduced by subtracting the rounded Table 3 percentages: 0·6-0·3, 5·7-4·2, 2·9-2·2, 3·1-2·3, 1·0-0·5, 15·0-9·7, 6·6-2·4. So 'three' follows from a consistent, if imprecise, method. Minor.

Second check: another model agreed

• p. 7 Table 5, Death, Clinically significant bradycardia

  351 (4·2%)

• p. 4 Table 3, Clinically significant bradycardia

  277 (6·6%)

Only the authors can say what is right here.

Judge's note: Checked on the page 4 and page 7 images. The other intraoperative predictors in Table 5 reconcile with Table 3: hypotension 1029 = 625+404, stroke 60 = 41+19, MI 271+144 = 176+239. The paper gives no definition or analysis set that would explain 351. Webappendix 2, which describes the model, is not among the files.

Second check: another model agreed

• p. 7 Table 5, Stroke, New clinically significant atrial fibrillation

  200 (2·4%)

• p. 4 Table 3, New clinically significant atrial fibrillation

  91 (2·2%)

Only the authors can say what is right here.

Judge's note: Checked on the page 4 and page 7 images. Hypotension in the same stroke section (1029) does match Table 3. The paper gives no restriction that would explain 200. Webappendix 2 is not among the files.

Second check: another model agreed

• p. 7 Discussion, left column

  but there was substantial heterogeneity

• p. 7 Figure 4, non-fatal MI total

  p<0·0001, I2=31%

• p. 6 Discussion, right column bottom

  was moderate heterogeneity

• p. 6 Discussion, last paragraph

  was moderate heterogeneity that was explained by one

Only the authors can say what is right here.

Judge's note: Checked Figure 4 on the page 7 image: total mortality p=0·10, I²=37%; non-fatal MI p<0·0001, I²=31%. The page 6 text calls the mortality heterogeneity 'moderate'; the page 7 text calls the nine-trial MI heterogeneity 'substantial'. The paper does not say whether the Figure 4 MI total is the nine-trial analysis, or whether p<0·0001 is the effect test or the heterogeneity test. Webtable 3 is not among the files. Query for the authors.

Second check: another model agreed

• p. 4 Results, right column

  anesthaesia or analgesia

• p. 4 Results, second paragraph

  anesthaesia or analgesia used

• p. 7 Table 5 footnote

  multivariate logisitic regression

• p. 4

  anesthaesia or analgesia used in the two

Judge's note: Both misspellings are visible on the page 4 and page 7 images. Elsewhere the paper spells the word 'anaesthesia' (Table 1).

Second check: none (judge only)

• p. 4 Results, last paragraph

  (figure 2).Of the

Judge's note: The space is also missing on the page 4 image, at the bottom of the right column.

Second check: none (judge only)

• p. 6 Table 4 footnote †

  not collection, therefore we cannot state

• p. 6 Table 4 footnote †

  not collection, therefore we cannot state with certainty

Judge's note: Visible on the page 6 image, in the Table 4 footnote.

Second check: none (judge only)

• p. 7 Table 5 footnote

  multivariate logisitic regression

• p. 7 Table 5 footnote

  multivariate logisitic regression analysis

• p. 6 Table 4 footnote

  not collection, therefore

• p. 7 Table 5 footnote

  logisitic

Judge's note: Both are visible on the page images: the Table 5 footnote on p7 and the Table 4 footnote on p6.

Second check: none (judge only)

• p. 7 Table 5, Myocardial infarction without ischaemic symptoms

  271 (3·3%)

Judge's note: On the page image every other frequency in the death panel matches the denominator of 8351 (all randomised patients; 8351 is used in the abstract and results). 271/8351 = 3·245%, which rounds to 3·2%, so this is not a rounding tie.

Second check: none (judge only)

• p. 8 National coordinating centres, Colombia

  Autónoma de Bucamaranga

• p. 8 National coordinating centres

  Universidad Autónoma de Bucamaranga

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the city and university name is Bucaramanga. The page image (p8) shows 'Bucamaranga'.

Second check: none (judge only)

• p. 8 National coordinating centres, Brazil

  Dante Pazzanese Institue of Cardiology

Judge's note: Visible on the p8 image, in the national coordinating centres list.

Second check: none (judge only)

• p. 8 Investigators list

  Equador (16)

Judge's note: The p8 image shows 'Equador (16)' in the investigators list, and 'Ecuador' in both the national coordinating centres list and the national coordinators list.

Second check: none (judge only)

• p. 8 POISE Study Group, Steering committee

  L Lisheng

Judge's note: The p8 image shows 'L Lisheng' in the steering committee list. In the same author list the name appears as 'Lisheng Liu' (writing group) and 'L Liu' (national coordinators), so the name order is inconsistent.

Second check: none (judge only)

• p. 9 Reference 1

  Ballieres Clin Anaesthesiol

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the journal's name is spelled Baillière's Clinical Anaesthesiology. The p9 image shows 'Ballieres'.

Second check: none (judge only)