Author + information
- Luis C.L. Correia, MD, PhD∗ ( and )
- Laudenor P. Lemos Jr., MD
- ↵∗Medical and Public Health School of Bahia, Av. Princesa Leopoldina, 19/402, 40.150-080 Salvador, BA, Brazil
In a recent issue of the Journal, Nasir et al. (1) used the concept of number needed to treat (NNT) to rationalize the statin therapy decision. The study indicated that 40% of subjects eligible for statin therapy had no coronary artery calcification (CAC), which was associated with a lower incidence of atherosclerotic cardiovascular events (ASCVEs), yielding a higher NNT and theoretically preventing the need for statin therapy.
Although the authors appropriately suggest that “among candidates for statin therapy, clinicians should consider the role of CAC,” we suggest a further appraisal of the study's results to avoid overestimation of CAC testing utility.
In the study, candidates for statin therapy with no CAC had a NNT of 64 to prevent 1 ASCVE compared with an NNT of 28 in those with any CAC. Although it is a significant NNT gradient, a traditional risk score would certainly identify subgroups of particular baseline risks and different NNTs. Therefore, within this wide risk range, comparing CAC score stratification with no clinical stratification may have overestimated CAC testing utility.
Second, the need for testing should be evaluated by comparing post-test probability with pre-test probability of events, representing performing versus not performing testing. In those at intermediate risk, the incidence of events was 8.2%, leading to a pre-test NNT of 41. Thus, a better description of the value of CAC testing would be the comparison of an NNT of 64 (no CAC) with the pre-test NNT of 41, a smaller gradient.
Finally, reclassifying individuals is not a guarantee of improving prediction. The analysis must take into account the net of correct versus incorrect reclassifications (2). As opposed to contemporary studies, the authors did not provide this analysis.
A century ago, William Osler (3) stated that “medicine is a science of uncertainty and an art of probability.” CAC testing is definitively useful for cases in which statin therapy is a decision dilemma. However, routine consideration of CAC testing would be a demonstration that the medical mind normally seeks for a platonic certainty.
Please note: Both authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Nasir K.,
- Bittencourt M.S.,
- Blaha M.J.,
- et al.
- ↵Brainyquotes. William Osler quotes. Available at: http://www.brainyquote.com/quotes/quotes/w/williamosl159326.html. Accessed January 12. 2016.