MRC/BHF Heart Protection Study of Cholesterol Lowering with Simvastatin in 20,536 High-Risk Individuals: A Randomised Placebo-Controlled Trial

• p. 15 Reference 27

  Friedewald T, Levy RI, Frederickson DS.

• p. 15 Reference 27

  Clin Chem 1972; 6:

• p. 15 Reference 27

  Clin Chem 1972; 6: 499–503.

• p. 15

  Friedewald T, Levy RI, Frederickson DS. Estimation of the concentration of low-density lip

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: that the Friedewald paper is Friedewald WT, Levy RI, Fredrickson DS, Clin Chem 1972; 18(6): 499–502. The page image confirms the entry as quoted, with volume 6, initial 'T' and the spelling 'Frederickson'. The '6' looks like the issue number printed in place of the volume.

Second check: another model agreed, code check

• p. 15 Reference 5

  J Chron Dis 1984; 7: 569–84.

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: that J Chronic Dis 1984 is volume 37 (issue 7). The page image confirms the entry as quoted, with volume printed as 7; it looks like the issue number given as the volume.

Second check: another model agreed

• p. 15

  Szatrowski TP, Peterson AV, Shimizu Y, et al. Serum cholesterol, other risk factors, and c

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the Szatrowski et al. Japanese cohort paper is in J Chronic Dis volume 37 (1984), according to the reviewer's Crossref record (doi:10.1016/0021-9681(84)90007-9). The page image confirms that the paper prints volume '7'. The files cannot confirm the true volume, so the authors should check it.

Second check: another model agreed, code check

• p. 1 Abstract, Findings

  was similar (and significant) in each subcategory of

• p. 10 Figure 8, Triglycerides ≥4·0

  189/813(23·2%)

Only the authors can say what is right here.

Judge's note: Looked at the abstract crop and Figure 8 on p10. The figure 2 legend gives the CIs as 95%, and the ≥4·0 triglyceride row's horizontal line runs to about 1·4, well past the line of no effect. A rough calculation from the counts gives z≈1·8 (p≈0·07). The results text (p9) uses milder wording ('about one-quarter in each subcategory'). The abstract claim of significance in every subcategory is therefore an overstatement as written; the authors should say whether only the listed subcategories were meant.

Second check: another model agreed

• p. 1 Abstract, Methods

  Primary outcomes were mortality (for overall analyses) and fatal or non-fatal vascular events (for subcategory analyses)

• p. 3 Statistical analysis

  Other secondary comparisons included the effects on major coronary events, and on major vascular events, in different subcategories

Only the authors can say what is right here.

Judge's note: Read both crops (p1 abstract, p3 statistical analysis) and searched the text for 'primary' and 'protocol'. Nothing in the article reconciles the two tiers. The analysis plan defers to the protocol and amendments on a website that is not among the files, so only the authors can say which tier applies.

Second check: another model agreed

• p. 5 Table 3 footnote

  –1·0/67%=–1·4 mmol/L for all patients

• p. 5 Results text, p11

  an average of about 1·5 mmol/L in this population

• p. 14 Conclusions

  1·5 mmol/L in this population

• p. 5 text, Compliance and effects on blood lipids

  average of about 1·5 mmol/L in this population

Judge's note: Checked the footnote crop and p5. 1·0/0·67 = 1·49. With unrounded inputs (the Table 2 footnote gives LDL differences of –0·95 and –0·97 by vitamin stratum, about –0·96 overall) the ratio is about 1·43, which would explain the 1·4. As printed, though, the example calculation does not reach its stated result, and the paper gives two values (1·4 vs about 1·5) for one quantity. This is a minor internal inconsistency.

Second check: another model agreed

• p. 9 Results text, p15, left column

  had cerebrovascular disease (p=0·001), or had peripheral vascular disease (p<0·0001)

• p. 9 Figure 7, No prior CHD rows

  172/922 (18·7%)

• p. 9 Figure 7, No prior CHD + Peripheral vascular

  327/1325 (24·7%)

Only the authors can say what is right here.

Judge's note: Checked Figure 7 on p9. In the no-prior-CHD cerebrovascular row the 95% CI runs nearly to 1·0. A rough calculation from the counts gives z≈2·6 (p≈0·01) for that row and p≈0·001 for the peripheral row. The printed values fit the 'CHD or no prior CHD' rows better (406/1645 vs 488/1635 gives z≈3·3). This is only an approximation from counts; no test statistic is printed, so the authors should say which subgroup and test each P value refers to.

Second check: another model agreed

• p. 13 Discussion, More appropriate high-risk treatment strategy

  ranged from about 20% to about 30% (figure 7

• p. 9 Figure 7, Prior MI or other CHD + Diabetes mellitus, placebo

  381/1009 (37·8%)

• p. 14 Conclusions

  event rates of about 20–30%)

• p. 13 Discussion, 'More appropriate high-risk treatment strategy'

  the placebo group that ranged from about 20% to about 30% (figure 7; corresponding to 10-year risks of over 40%)

Judge's note: Checked p13 and Figure 7 on p9. Placebo percentages run from 18·6% to 37·8%; the subtotals (20·8%, 27·5%) do sit within 20–30%. Because the sentence explicitly includes combinations of conditions, the quoted range understates the upper end of the figure it cites. 'About' softens this, so it is rated as wording only.

Second check: another model agreed

• p. 13 Discussion, paragraph on CARE breast cancer

  developed breast cancer in the present study or in the other main statin trials.6,7,9–11

• p. 15 Reference 7

  Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia.

Judge's note: Checked the p13 text and reference 7 on p15. By its own title, reference 7 is a men-only trial, so it cannot contribute breast-cancer cases among women. The citation group is probably shorthand for 'the other main statin trials', which makes this a minor citation-precision point.

Second check: another model agreed

• p. 14 Collaborators list

  L Donald, Bassetlaw:

Judge's note: The crop of p14 confirms 'L Donald, Bassetlaw:', while every other hospital entry ends with ';'. This is a punctuation slip.

Second check: none (judge only)

• p. 15 Reference 29

  Pathol 1995;. 104: 76–81.

• p. 15 Reference 29, Jialal I et al.

  Am J Clin Pathol 1995;. 104: 76–81.

Judge's note: The page image (crop and p15) shows ';.' between 1995 and the bold volume 104. The other references on the page use the form 'year; volume: pages'. This is cosmetic punctuation.

Second check: none (judge only)

• p. 16 Reference 49

  Cerivastatin and reports of fatal rhabdomylosis

• p. 16 Reference 49

  Cerivastatin and reports of fatal rhabdomylosis.

Judge's note: The page image on p16 shows 'rhabdomylosis'. Elsewhere the paper spells it 'rhabdomyolysis': in the table (extract line 901) and in the text (lines 1719 and 1896). This is a misspelling.

Second check: none (judge only)

• p. 16 Reference 39

  Laasko MP

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the co-author of Kivipelto et al. BMJ 2001 is Mikko P Laakso. The page image on p16 shows 'Laasko MP'. Nothing in the files establishes the correct spelling.

Second check: none (judge only)