Absolute risk assessment for guiding cardiovascular risk management in a chest pain clinic

In reply: We thank Blazak and Greaves1 for their comments on our article.2 In their letter, the authors noted the possible influence of bias from between‐group differences in the baseline‐reported use of anti‐hypertensive and lipid‐lowering therapy. We want to highlight that we accounted for these covariables in the ANOVA models for the primary analysis. Nonetheless, we have undertaken further analysis, and this also confirms that the results are consistent regardless of whether anti‐hypertensive or lipid‐lowering therapy was reported at baseline (Box). We propose that the mechanism for the modest improvement in blood pressure and cholesterol in the intervention arm is the adoption of lifestyle changes, supported by the noted improvement in body mass index and quality of life scores for this group. It is possible that adherence to prescribed medication may have differed between the groups at follow‐up, and this remains an important focus of the approach provided by the rapid access chest pain clinic.

We agree that an answer is only provided for the two‐thirds of patients who agreed to engage, but respectfully point out that this is applicable to all randomised clinical trials. Indeed, cancellation of trial results based on proportion declining randomisation would leave us with scant guidelines. Moreover, the absolute risk of those declining recruitment was 11.7% (compared with 12.8% and 13.1% in the control and intervention groups respectively).

Regarding smoking cessation, we agree that efforts in this area are important, and hope this is one of the major messages from our study. National guidelines3 recommend that “all smokers should be offered advice about methods to aid smoking cessation”. Although clinicians are expected to offer this advice as part of usual care (consistent with practice in the control group of our study), not all smokers attending outpatient clinics are referred to smoking cessation services.