Clonidine in Patients Undergoing Noncardiac Surgery
2014 · paper · Clinical trials
A randomized trial of low-dose clonidine versus placebo in 10,010 patients having noncardiac surgery. Clonidine did not reduce death or nonfatal heart attack but increased clinically important low blood pressure and nonfatal cardiac arrest.
1 finding
Wording The Table 3 footnote says "We performed a multivariable logistic-regression analysis", but the table reports "Adjusted Hazard Ratio (95% CI)" and the footnote describes "time-dependent variables in the model".
- p. 9 Table 3 footnote *
We performed a multivariable logistic-regression analysis to determine the independent predictors of myocardial infarction.
- p. 9 Table 3 column header
Adjusted Hazard Ratio (95% CI)
- p. 3 Statistical Analysis
the only exceptions were the outcome of acute kidney injury with receipt of dialysis, for which we used a logistic-regression
- p. 9 Table 3 footnote *
We performed a multivariable logistic-regression analysis
Judge's note: Page 9 image confirms both the footnote and the column header. Page 4 text calls the estimate a hazard ratio (1.37). Page 3 Methods say Cox models were used for all outcomes, with logistic regression only for dialysis-requiring AKI. Time-dependent covariates and hazard ratios belong to a Cox model, so the footnote's name for the model contradicts the table.
Second check: another model agreed
Report a problem with this finding
4 minor (typos and the like; most checked by the judge only)
Wording Discussion, quoting ref 16: "36 events and 26 events, respectively; relative risk, 1.25; 95% CI, 0.83 to 2.21". The point estimate is not near the geometric midpoint of its interval (about 1.35), unlike the other ratios quoted from the same review.
- p. 5 Discussion, first column, top
relative risk, 1.25; 95% CI, 0.83 to 2.21
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the values actually reported in ref 16 (Wijeysundera, Am J Med 2003). Page 5 image confirms the printed text. The bounds multiply to 1.834, while 1.25 squared is 1.5625. The other quoted pairs are close to symmetric on the log scale: 0.25*0.90=0.225 vs 0.47^2=0.221, and 0.52*2.09=1.087 vs 1.05^2=1.1025. Pooling methods can produce some asymmetry, so the authors should check the transcription of the estimate or the bounds against ref 16.
Second check: another model agreed
Report a problem with this finding
Wording Table 1 (continued), clonidine column: the intraoperative anesthesia categories ("2353/4972 (47.3)" ... "392/4972 (7.9)") add up to 4973, one more than the denominator of 4972. The placebo categories add up exactly to 4960.
- p. 6 Table 1 (continued), Intraoperative anesthesia, clonidine column
2353/4972 (47.3)
Only the authors can say what is right here.
Judge's note: Page 6 image: I re-added all ten categories. The table has no footnote saying that anesthesia categories may overlap; the footnote about overlap covers surgery type only. Which count is off by one cannot be told from the files.
Second check: another model agreed
Report a problem with this finding
Typo "Results of the six trials in the systemic review" should read "systematic review".
- p. 5 Discussion, first column, second paragraph
Results of the six trials in the systemic review
Judge's note: Page 5 image shows 'systemic review'. The same review (ref 16) is called 'A systematic review' on page 4.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation reads "Research Institute HCor (Hospital do Coracão)"; the Portuguese spelling is "Coração" (cedilla missing).
- p. 10 Affiliations
Hospital do Coracão
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the hospital's name is 'Hospital do Coração'. The page 10 crop shows 'Coracão' with a tilde but no cedilla. The reviewer's 'rearranged letters' is wrong; only the cedilla is missing.
Second check: none (judge only)