Question

bias CRecall bias D. Detection bias E. Insensitive measure bias 17. A study seeks to determine whether or not carvedilol reduces risk of mortality compared to standard tx with diuretics and ACEI. 1511 patients with CHF are randomly assigned to treatment with carvedilol plus ACEI and diuretic therapy and 1518 to standard treatment with a combination of ACEI and diuretic All-cause mortality (hazard ratio 0.83 [95 % Cl : 0.74-0.93], p=0.00 17). Which of the following represents the number of patients needed to treat to save 1 life with addition of beta-blockers? A.1/(0.40-0.34) B. 1/(0.34-0.40) C. 1/(40-34) D. 1/(34-40) E. 1/0.83 was 34% (512 of 1511) for carvedilol and 40% (600 of 1518) for the control group

Answer 1

This is a calculation of the number needed to treat or NNT which means that the number of patients to be treated with the trial intervention to save one life. The formula for the same is:

NNT = 1/(%risk in control group - %risk in treatment group)

NNT = 1/(0.40 - 0.34)
Hence, A) is correct.

In B) The calculation incorrectly reverses the numbers.
In C and D) The calculations utilise the integral numbers and not the percentages.
E) False as the hazard ratio is unrelated to NNT