Letter by Donzelli Regarding Article, “Potential Deaths Averted and Serious Adverse Events Incurred From Adoption of the SPRINT (Systolic Blood Pressure Intervention Trial) Intensive Blood Pressure Regimen in the United States: Projections From NHANES (National Health and Nutrition Examination Survey)”

If SPRINT (Systolic Blood Pressure Intervention Trial) eligibility criteria were applied to the 1999 to 2006 NHANES (National Health and Nutrition Examination Survey), Bress and colleagues 1 calculate that lowering systolic blood pressure (BP) to an intensive goal of <120 mm Hg (versus the standard goal of <140 mm Hg) could prevent ≈107 500 deaths per year, despite a higher incidence of treatment-related serious adverse events.A sensitivity analysis has tried to take into account the different methods of BP measurements in NHANES and SPRINT, finding reduced but still significant outcomes.The SPRINT methodology is not usually used in clinical practice.BP was measured at an office visit, in participants sitting quietly alone for 5 minutes, with an automated device (model 907, Omron Healthcare) that automatically waited 5 minutes and then took 3 BP measurements and averaged them.During the measurements, patients were left unattended.This creates important problems of comparability with the measures of most cardiovascular outcome trials, which to date have used either auscultatory or automatic oscillometric methods of seated BP measurement, sometimes recorded shortly after the patient's arrival, similar to what happens in usual clinical practice.To extrapolate the findings of SPRINT to clinical practice, one should consider that automated office BP readings are on average ≈15/8 mm Hg lower than usual BP as measured in real-life doctors' offices. 2 In NHANES, the BP measurements use "an appropriately sized cuff; BP was measured with participants seated after five minutes of rest using a mercury sphygmomanometer by a trained study physician.SBP [systolic BP] and DBP [diastolic BP] were defined as the mean of three BP measurements taken oneminute apart." 1 This procedure differs in at least 2 major ways from SPRINT measurements: The device is not an electronic automated one but a mercury sphygmomanometer used by a "trained physician," and the participant is not alone.The combination of these 2 features probably makes the systolic BP readings ≈15 mm Hg higher than measures of the same subject taken according to SPRINT; therefore, a correction of 10 or 20 mm Hg, like that used in the sensitivity analysis, 1 might seem sufficient.However, it does not take into account 2 points.First, the main issue that doctors must understand is not the entry BP criteria to apply but the BP goal to pursue.If an intensive systolic BP goal is <120 mm Hg, as claimed by SPRINT, but the physicians continue to measure BP in the usual way, this can lead to dangerous overtreatment.Second, the problem can be even worse with older patients.In an interesting study, 3 an automated office BP monitoring during 30 minutes of unattended rest, with 6 readings and the mean of the last 5, the mean systolic difference versus Letter by Donzelli Regarding Article, "Potential Deaths Averted and Serious Adverse Events Incurred From Adoption of the SPRINT (Systolic Blood Pressure Intervention Trial)