Randomized controlled clinical studies have found that an on-trearment blood pressure (BP) reduction to < 140/90 mm Hg is associated with a reduction in cardiovascular (CV) events, regardless of the specific antihypertensive agent used. In most cases, this is measured as the mean achieved BP for all study visits or the mean achieved BP at specific time points such as at last study visit or at a 6-month or annual visit. While the consistency of well-controlled on-treatment BP may improve CV outcomes more than the magnitude of BP reduction over the same period, perhaps we should be evaluating the overall duration of time with well-controlled BP rather than the mean on-treatment BP or absolute BP reduction achieved. The International Verapamil SR-Trandolapril (INVEST) trial provided an opportunity to study this question: whether the proportion of study visits at goal BP provided additional prognostic information to that provided by assessing on-treatment BP over the length of the trial. The INVEST trial was a prospective, randomized, open- label, blinded end point (PROBE) study in 22,576 patients aged at least 50 years with hypertension and coronary artery disease. It was designed to determine the relative efficacy of a calcium channel blocker (CCB)–based regimen compared with a β-blocker–based regimen. The CCB-based strategy consisted of sustained-release verapamil (verapamil SR) with the addition of trandolapril, an angiotensin-converting enzyme inhibitor, while the β–blocker-based strategy consisted of atenolol with the addition of a thiazidetype diuretic. Additional antihypertensive therapy was added as needed to achieve the target BP level of < 140/90 mm Hg, or < 130/85 mm Hg in persons with diabetes or chronic kidney disease, as recommended by the Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC VI), which was in effect at the time of the study. Trandolapril was also recommended in all patients with a history of diabetes, chronic kidney disease, or heart failure, regardless of treatment allocation. After randomization, visits were scheduled at week 6, 12, 18, and 24 and every 6 months thereafter. After an average follow-up of 2.7 years, there was no difference in the incidence of the primary composite end point (death, nonfatal myocardial infarction, nonfatal stroke) between treatment arms. Mean on-treatment BP level was 135/79 mm Hg in 72% of patients on the CCB-based therapy and 71% of those on the β-blocker–based therapy achieving a mean BP of <140/90 mm Hg. In this post hoc analysis of the INVEST trial, investigators determined the effect of BP control on the incidence of the primary and secondary study end points, pooling the results of the 2 treatment strategies. The effect of overall BP control was evaluated using both the mean achieved BP and the proportion of study visits at which BP was controlled (ie, at goal). For this purpose, patients were divided into 4 groups based on the proportion of visits in which BP was controlled (goal BP level, <140/90 mm Hg): <25%, 25% to <50%, 50% to <75%, and ≥75%. Stepwise Cox proportional hazard models were used to estimate the risk of the primary outcome with the <25% group used as the reference. A stepwise model adjusted for differences in multiple baseline characteristics. A similar evaluation was also undertaken to analyze the effects of baseline BP and mean follow-up BP. Overall, 34% of patients had controlled BP at <50% of study visits. Sixteen percent of nondiabetic and 19.5% of diabetic patients had <25% of visits with controlled BP. Baseline characteristics differed between groups: compared with patients in the >75% group, patients in the <25% group were more likely to be older, female, black, and a US resident and have a higher body mass index and diabetes. The same patients were less likely to be Hispanic or a previous or current smoker, have hypercholesterolemia, or use nitrates or aspirin. Baseline BP level was highest in the <25% group and lowest in the >75% group. The rate of the primary outcome declined as the proportion of visits with BP at goal increased, and this relationship remained significant after adjusting for differences in baseline values, including baseline BP. The incidence of the primary end point was 40% lower in the >75% control group compared with the <25% group. The greatest effect was seen in the incidence of nonfatal stroke, which was 50% lower in the >75% group compared with the <25% group. The results were similar in the subgroup of patients with diabetes at baseline. There was a significant correlation between baseline BP and the proportion of visits with controlled BP, but after adjustment for other baseline characteristics, baseline systolic and diastolic BP were not associated with an increased risk of the primary outcome. There was also a strong correlation between mean follow-up systolic BP level and the proportion of visits at which BP was controlled; however, unlike baseline BP, mean follow-up systolic BP was significantly associated with an increased risk of the primary outcome. Statistical analysis indicated an independent effect on the proportion of visits with BP at goal and mean follow-up systolic BP. In this post hoc analysis of the INVEST study, the proportion of study visits at which BP was at goal was independently predictive of the risk of subsequent CV events, regardless of the treatment strategy chosen. Consistency of achieving goal BP during treatment provides additional information on the protective effect of antihypertensive treatment. Physicians need to be concerned at each clinic visit if BP is not at goal.—Mancia G, Messerli F, Bakris G, et al, Blood pressure control and improved cardiovascular outcomes in the International Verapamil SR-Trandolapril Study. Hypertension. 2007;50:299–305. While most of our landmark clinical trials of antihypertensive therapy have been designed to compare one pharmacologic treatment strategy against another, post hoc analysis of these trials has usually found that the level of BP control is more important than the individual choice of medication to prevent CV events. That is, the particular drug almost never trumps the overall effect of BP reduction. For example, in a post hoc analysis of the Valsartan Antihypertensive Long-Term Use Evaluation (VALUE) study, the incidence of fatal and nonfatal CV events was decreased by 25% in those in whom goal BP was achieved at 6 months compared with those whose BP was not controlled, an effect that was independent of treatment allocation. In addition, it has been suggested that in the Antihypertensive Lipid-Lowering Therapy to Prevent Heart Attack Trial (ALLHAT), some of the differences in the rate of secondary end points between treatment groups may have been at least partially accounted for by differences in achieved BP, especially in certain subgroups. In these end point studies, BP control is usually defined based on mean follow-up BP or BP achieved at a specific time point of the trial, which is often at study completion. The present data from the INVEST trial suggest that the consistency of BP control, defined as the proportion of study visits in which goal BP is achieved is also an important independent predictor of CV events. The overall message is that both the magnitude of BP elevation and the duration of BP elevation above goal level are independently predictive of CV risk. The potential weakness of using only mean or final achieved BP as the definition of BP control is well documented. In both the VALUE and the Anglo-Scandinavian Cardiac Outcomes Trial–Blood Pressure–Lowering Arm (ASCOT-BPLA) studies, mean BP was similar between study groups at the end of the study. A number of commentators have suggested, however, that the findings of both of these studies were influenced by the fact that goal BP was reached in one treatment arm more quickly than in the other arm. Complex statistical analyses have been performed to attempt to determine the effect of this early BP difference, but this analysis of the INVEST study data suggests that perhaps the proportion of visits at which BP is at goal is another reasonable way to determine similarity of overall BP control between study groups. Future studies should employ multiple means of comparing achieved BP between groups including the final achieved BP as well as the proportion of visits at which BP is at goal. While interesting, one must remember that the present data were based on a nonrandomized post hoc analysis of a prospective randomized trial. In addition, this analysis was limited to a specific group of hypertensive individuals (ie, those older than 50 years with established coronary artery disease) that may not be representative of the usual patient with hypertension. In addition, many patients had significant potential confounding variables (higher baseline DP level, body mass index, cholesterol level, and black race in the group with the lowest proportion of visits with BP control) that may not have been adequately accounted for in the statistical analysis. Nonetheless, this finding is consistent with our understanding of the importance of overall BP control and has important clinical and research implications. While overall BP control rates are improving in the United States, it still appears that in many instances practitioners fail to appropriately intensify antihypertensive therapy when faced with a patient who has poorly controlled BP. These data and our understanding of the pathogenesis of elevated BP suggest that cardiovascular events are related not just to the magnitude of BP elevation but also the duration of time a patient has poorly controlled BP. This concept is emphasized in JNC 7, which calls for more rapid control of BP. JNC 7 encourages practitioners to consider initiation of BP-lowering therapy with 2 medications in patients who are >20 mm Hg above their systolic BP goal and to see patients every 4 to 6 weeks, with intensification of therapy at each visit until BP is controlled. Each visit at which BP is elevated and no intensification of therapy occurs represents a missed opportunity for outcome improvement. As we move toward an era of pay-for-performance services, it is important that we determine appropriate metrics for the assessment of BP control. These data from the INVEST study suggest that achieving goal BP at the most recent visit may not be the most appropriate metric. It is certainly tempting to use the percentage of visits at which BP is controlled as an additional measure of outcome. Unlike clinical studies, however, in which the number of visits is fixed, in clinical practice more frequent visits will be required when BP is poorly controlled. Because we want to reward practitioners for seeing patients with poorly controlled BP more frequently and in a more timely fashion, it may be reasonable to evaluate the “time in therapeutic range,” whereby each BP reading is weighted according to the length of follow-up time until the next visit. Using the time in therapeutic range will increase the incentive for practitioners to appropriately see patients with poorly controlled BP more often and to intensify therapy in an effort to reach goal BP. We need to continue to shift the focus of hypertension research away from questions of best drug selection and toward evaluation of strategies to achieve aggressive, rapid, and perhaps most importantly, consistent BP control.