Author + information
- Alexandre C. Pereira, MD⁎ ( and )
- Whady Hueb, MD, PhD
- ↵⁎Heart Institute (InCor), University of São Paulo Medical School, Av Dr Eneas de Carvalho Aguiar, 54, 10 andar–Bloco II, São Paulo, SP 05403-000, Brazil
We appreciate Dr. Karthikeyan’s interest in our study on the predictive power of clinical judgment in chronic coronary artery disease (1). We agree with his opinion that there is in fact no real disconnect between evidence-based medicine and an individual clinician’s judgment and that evidence from clinical trials helps add objectivity, reduces bias, and refines a clinician’s ability to make decisions.
Clinical judgment, far from a mystical definition, is the result of a complex equation that takes into account objective data from biochemical tests, imaging studies, and a patient’s history. It also uses subjective information acquired by the physician over the course of the patient–physician relationship.
We disagree, however, with the view that the different prevalence of 3-vessel disease and the complexity of lesions were primarily responsible for the nonconcordance between a clinician’s treatment option and the randomization process. Furthermore, unlike what Dr. Karthikeyan affirmed, this has not been pointed out in our report. In fact, a careful examination of Table 3 from our study (1) would allow the observation that lesion morphology distribution in patients treated by percutaneous coronary intervention (PCI) was not significantly different between concordant and discordant groups, and even the concordant group treated by PCI had an almost 50% prevalence of patients with 3-vessel disease. Angiographic findings were certainly used in the decision process. However, it should be emphasized that the angiographic variables that were investigated explained a very small percentage of our model’s overall variance. This means that clinical judgment either uses other variables not investigated in our study or it is capable of deriving information from higher-order interactions using the variables available from imaging examinations (i.e., angiographic findings) and cardiovascular risk factors that a patient may present. In fact, it probably uses both and has the capability of integrating all this information into a single decision.
No simple statements can be easily made regarding what clinical, demographic, angiographic, or biochemical variables are being used (or in what way) by clinicians to make their decision in this particular scenario. An increased number of patients could potentially permit statistical power for exploratory subgroup and higher-order interaction analysis in the aim of disclosing this important issue.
During the last several years the cardiology community has been highly influenced by medical guidelines, randomized clinical trials, and “cost-effective” algorithms. All these tools are invaluable for practicing medicine and in helping the decision-making process. Nevertheless, we should not forget that a physician’s judgment is what processes and consolidates all this information. Apparently, in this particular clinical scenario it can still make a difference.
- American College of Cardiology Foundation