Let's prioritize the right care for the right patients with hypertension.

Abstract

High-qualityhealthcare is oftensummarizedasdelivering the right care to the right patient every time. In this issueof JAMA Internal Medicine, Rodriguez and colleagues1 publish an important studyusingdata from the Atherosclerosis Risk in Communities (ARIC)Study to further inform the debate about which patients should receive antihypertensive medicines every time to achieve goal blood pressure (BP) targets— but it also informsus aboutmuchmore in the treatment of hypertension (HTN). This observational study among patients with HTN finds no additional risk among the participants in the standard systolic BP (SBP) cohort (120-139 mm Hg) comparedwith those in the lowSBPcohort (<120mmHg).This finding differs from epidemiologic data that consistently suggests a linear relationship between rising SBP in a population and increasing rates of poor outcomes (eg, stroke,myocardial infarction [MI],heart failure [HF], andmortality).Basedonthis strong epidemiologic association, conventional wisdom has heldthatusingmedicationtoachieve lowerBPs inpatientswith HTN should result in lower rates of stroke, MI, and HF, similar to those seen among persons who take no antihypertensive medications. However, Rodriguez et al1 appear to provide observational evidence that lowerBP is not always better. Recent randomized controlled trials (RCTs)2-4 have also raised the concernthatnoadditionalbenefit isgainedby loweringSBPsbelow therecentlyestablishednewtreatmentgoals.5TheseRCTswere influential in justifyingan increasedSBPgoal for those60years or older in the 2014 Guidelines set forth by the panel appointed to the Eighth Joint National Committee (JNC-8),5 especially because themethodology of the evidence review requiredRCT evidence as opposed to observational data.While Rodriguez and colleagues1 do not provide a definitive answer to those troubled by raising the SBP goal for those older than 60 years without diabetes or chronic kidney disease, they do illustrate themarginal benefits of addingmoremedications to achieve lower BPs in the general population. Both observational studies and RCTs can provide important informationaboutHTN,but theyoftenhavedifferentpurposes. The methodology used by the panel members appointed toJNC-86 foranswering their specific criticalquestions focused on the evidence supporting medication management inpatientswithhighBP.Allmembersof thepanel agreed that RCTs were the appropriate source to answer these questions prior to the external methodology team’s literature review and synthesis of the evidence. However, the evidence from RCTs caused discomfort for a minority of panel members because theRCTevidencedidnot comportwith the findings from observational studies.6 All researchmethodologieshave limitations,andthis is true of the study by Rodriguez et al.1 The authors created longitudinalcohortssegregatedbytime-dependentBPmeasuresof low, standard and elevated for patients with HTN over 9 years and then observed these patients for a median of 21.8 years and evaluated important health outcomes (stroke, HF, combination of MI/death from coronary heart disease, and composite score). As the authors acknowledge, unmeasured confounders are themost difficult factors to assess in judging the validity of observational studies. For example, higher BPs indicate moreseverediseaseandthusare linkedtopooroutcomes.Does achieving a BP of 139/88 mm Hg through the use of 3 antihypertensive medications or through unmedicated diet and exercise result in comparable outcomes?Becauseof theunderlyingseverityof thedisease,manyofuswouldsaythat thepatient takingmultiplemedications isatgreater risk.Thus, it is remarkable that Rodriguez et al1 found no differences in the low and standardSBPcohorts.Among the3 cohorts, however, a signifiRelated article page 1252 BP and CV Events in Patients With Hypertension Original Investigation Research