Dapagliflozin in Patients with Heart Failure and Reduced Ejection Fraction
2019 · paper · Clinical trials
A randomized trial of dapagliflozin versus placebo in 4744 patients with heart failure and reduced ejection fraction. Worsening heart failure or cardiovascular death occurred in 16.3% versus 21.2%, with similar results in patients with and without diabetes.
0 findings
None that passed a second check.
10 minor (typos and the like; most checked by the judge only)
Wording The Discussion says that the findings in other advanced-disease subgroups, including "an increased NT-proBNP level", "were not consistent with the finding regarding the NYHA class". In Figure 3, however, the NT-proBNP >median group has HR 0.79 (0.68–0.92) against 0.63 (0.49–0.80) for ≤median, which is the same direction as NYHA III/IV (0.90) against II (0.63).
- p. 7 Discussion, paragraph on subgroups
increased NT-proBNP level) were not consistent with the finding regarding the NYHA class
- p. 11 Figure 3, NT-proBNP rows
0.79 (0.68–0.92)
Only the authors can say what is right here.
Judge's note: Checked the Discussion on p.7 (crop C002-a) and Figure 3 on p.11. In Figure 3, the LVEF split (0.70 for ≤median vs 0.81 for >median) and the eGFR split (0.72 for <60 vs 0.76 for ≥60) do support the sentence. The NT-proBNP point estimates show less benefit in the more advanced group, as NYHA does, although the >median CI excludes 1 while the NYHA III/IV CI does not. No interaction P values are printed. Only the authors can say what they meant by 'not consistent'; this is a wording query.
Second check: another model agreed
Report a problem with this finding
Wording In Table 2, the P value for the change in creatinine is printed as "<0.007". Every other P value in the table is either exact or "<0.001".
- p. 9 Table 2, Laboratory and other measures, Creatinine row, P Value column
<0.007
Only the authors can say what is right here.
Judge's note: The upright image of p.9 shows "<0.007" for the creatinine row (difference 0.02, 95% CI 0.01 to 0.03). No other P value in the paper uses a nonstandard '<' threshold. Only the authors can say whether P=0.007 or P<0.001 was intended. The P=0.007 in footnote ¶ refers to acute kidney injury, which is a different outcome.
Second check: another model agreed
Report a problem with this finding
Wording In Figure 3, the dapagliflozin BMI subgroup denominators ("259/1537" and "127/834") sum to 2371, not the column total of N=2373. No missing BMI data are noted.
- p. 11 Figure 3, Body-mass index
259/1537
Judge's note: Checked Figure 3 on p.11. In the dapagliflozin arm, 1537+834=2371; the placebo arm sums to 2371, and the events sum to 386 and 502. Table 1 shows that some baseline data are missing (eGFR 962/2372), so 2 patients with missing BMI is the likely explanation, but neither the figure nor the text says so.
Second check: another model agreed
Report a problem with this finding
Wording In Figure 3, the NT-proBNP subgroup denominators ("100/1193" and "286/1179"; placebo "155/1179" and "347/1191") sum to 2372 and 2370, not to the arm totals of 2373 and 2371. No missing NT-proBNP data are noted.
- p. 11 Figure 3, NT-proBNP
100/1193
Judge's note: Checked Figure 3 on p.11. The events sum correctly to 386 and 502. Missing baseline NT-proBNP is the likely reason, and Table 1 shows missing eGFR in a similar way (962/2372), but the figure does not say so. The eGFR subgroups in Figure 3 also sum to 2372 and 2370.
Second check: another model agreed
Report a problem with this finding
Typo The sacubitril–valsartan subgroup interval is printed as "(95% CI, 0.50, 1.13)" with a comma, while the rest of the paper, including the next interval in the same sentence, uses "(95% CI, 0.65 to 0.86)".
- p. 5 Results, Outcomes, sacubitril–valsartan subgroup
(95% CI, 0.50, 1.13)
Judge's note: Checked in crop C001-a on p.5. The comma form appears only here; the house style elsewhere is 'to'. This is cosmetic and the meaning is still clear.
Second check: none (judge only)
Report a problem with this finding
Typo In Table 2, the Fracture row ("49/2368 (2.1)") has a blank dapagliflozin events/100 patient-yr cell, where every other safety row prints "—".
- p. 8 Table 2, Safety outcomes, Fracture row
49/2368 (2.1)
Judge's note: Checked the upright image of p.8. The cell is empty in the Fracture row, while the placebo events/100 patient-yr cell and every other safety row show a dash. This is cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Table 2 footnote § has a subject–verb disagreement: "The total symptom score on the Kansas City Cardiomyopathy Questionnaire (KCCQ) range from 0 to 100". It should read "ranges".
- p. 9 Table 2 footnote §
score on the Kansas City Cardiomyopathy Questionnaire (KCCQ) range from 0 to 100
Judge's note: Checked the upright image of p.9. The footnote reads 'score ... range', which is a grammatical slip.
Second check: none (judge only)
Report a problem with this finding
Typo Dr. Sabatine's disclosure reads "(paid to Brigham and Woman’s Hospital)". In the same disclosure the name is spelled "Brigham and Women’s Hospital" twice.
- p. 13 Disclosures, Dr. Sabatine
(paid to Brigham and Woman’s Hospital)
- p. 13 Disclosures, Dr. Sabatine
Brigham and Woman’s Hospital
Judge's note: Checked crop C009-a on p.13. 'Woman’s' appears once, and the two other mentions in the same paragraph use 'Women’s'. This is an internal inconsistency and a misspelling.
Second check: none (judge only)
Report a problem with this finding
Typo Dr. Bělohlávek's disclosure reads "lecture fees from Getginge", which is probably a misspelling of Getinge.
- p. 13 Disclosures, Dr. Bělohlávek
lecture fees from Getginge
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the company is named Getinge. The page image (crop C010-a, p.13) does show 'Getginge'.
Second check: none (judge only)
Report a problem with this finding
Typo The appendix affiliation reads "Universidade de São Paolo, São Paolo (J.C.N.)"; the expected spelling is São Paulo.
- p. 14 Appendix, affiliations (J.C.N.)
São Paolo, São Paolo (J.C.N.)
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the Brazilian city and university are spelled 'São Paulo'. The page image (crop C011-a, p.14) does show 'São Paolo' twice.
Second check: none (judge only)