The role of self-measured BP in the management of hypertension

The study by Den Hond and colleagues (page 13) addresses the role of self-measured blood pressure (BP) in the management of hypertension. Self-measuring BP at home with a commercial device might provide more BP readings with fewer office visits, eliminate the white-coat syndrome, eliminate observer bias by being automated, reduce costs, and improve compliance. In addition, it might improve BP control, which has been disappointingly low despite efforts to improve compliance.

This 1-year prospective study enrolled adult men and women with diastolic BPs between 95 and 114 mm Hg on multiple screening visits who were taking no medication or were taking less than two antihypertensive medications. Those with congestive heart failure, unstable angina, Stage 3 or 4 hypertensive retinopathy, a history of myocardial infarction, cerebrovascular accident within 1 year, severe noncardiovascular illness, chronic kidney disease with creatinine greater than 2.0 mg/

dL, mental illness, substance abuse, and those working at night were excluded. Of the initial 606 patients screened in 59 medical facilities, 400 eligible patients were randomized and followed up.

All patients had their BPs measured at home using a precalibrated oscillometric automated cuff and at the office by a physician with a sphygmomanometer. Follow-up visits were made monthly for the first

2 months and every 2 months thereafter. The ambulatory BP monitoring was recorded but not disclosed or used for the management of hypertension. Changes in antihypertensive medication were guided by either office or home BP readings, depending on the initial randomization. The target BP in the Office Guidance group was attained using only the office readings. The Home Guidance group had their target BP attained exclusively using the home readings. As will be discussed below, this is a major flaw of this study.

The adjustments in medications were made by a single blinded physician to achieve a predefined diastolic BP goal of 80 to 89 mm Hg. All patients were placed on monotherapy with lisinopril (Prinivil, Zestril),

10 mg per day, after randomization. Those with a known contraindication to or those who were intolerant of angiotensin-converting enzyme inhibitors were placed on atenolol (Tenormin), 50 mg daily, instead (step 1). The lisinopril or atenolol dose

was then doubled if diastolic BP

was above 89 mm Hg (step 2). Those needing greater control were treated additionally with hydrochlorothiazide (Carozide, HydroDiuril), 25 mg per day, or amlodipine (Norvasc), 5 mg daily (step 3). Finally, prazosin (Minipress), up to 6 mg daily, was added to those already on amlodi-pine; patients not already taking am-lodipine were started on amlodipine, 5 mg per day (step 4). If the diastolic BP dipped below 80 mm Hg, therapy was reduced by one step.

The respective office and home BP readings at randomization were similar between the Office Guidance and the Home Guidance groups. All patients from both groups had their BPs taken at the office and at their homes. When each patient compared the BP readings obtained at home versus at the office, their BPs at home were consistently lower, and this observation was maintained throughout the study. At baseline in both groups, home readings were lower than the office readings by ~13.5 mm Hg for systolic and ~9.5 mm Hg for diastolic BPs. The lower home BP readings observed throughout the study may be explained by the white-coat effect, but this was not clearly established in the study.

At the end of the study, there was less intensive drug treatment and higher overall BP achieved in the Home Guidance group when their BPs obtained under the same settings were compared. The final office-measured systolic and diastolic BPs in the Home Guidance group were higher than the final office-measured BPs in the Office Guid-ance group by 6.8 and 3.5 mm Hg, respectively. This outcome, however, does not reflect the inferiority of BP therapy guided by home readings. Rather, it reflects a flaw in the study design whereby medications were added or subtracted to strictly maintain diastolic BP between 80 and 89 mm Hg throughout the study by two techniques that yielded dichotomous readings in each patient.

The data and conclusions provided by the authors of this study are difficult to interpret because of two variables. One is the lack of standardization of the home and office BP readings. The authors do not provide evidence of the accuracy of the home readings by making certain that the oscillometric readings were verified by simultaneous sphygmomanometric readings. It is well known that the real-world accuracy of oscillometric devices may be less reliable than expected.1,2 Oscillometric devices may report falsely low diastolic BP, especially in the elder-ly and under various situations of high cardiac output and low peripheral resistance, such as pregnancy, anemia, thyrotoxicosis, and aortic regurgitation.1-4

The second variable is the study design to maintain diastolic BP between 80 and 89 mm Hg throughout the study by the single-blinded physician for both groups of patients. It would have been interesting to set BP control below 90 mm Hg without the lower threshold.

The rationale for using only diastolic BP readings to control the hypertension is unclear, as isolated elevations in systolic BP increase the risks for myocardial infarction, stroke, left ventricular hypertrophy, kidney damage, and cardiovascular mortality.5,6 As with many prospective hypertension trials, the ever-evolving standards for acceptable BP make the results appear less relevant and antiquated. Clinical and epidemiologic data indicate that lower BP readings are associated with lower mortality and morbidity, and larger reductions in BP lead to larger reductions in risk of total cardiovascular events.7,8 The Prospective Studies Collaboration has also shown that cardiovascular mortality rates increase steadily as BP rises above 115/75 mm Hg.8 The coexistence of other risk factors, including hyperlipidemia and diabetes mellitus, may further amplify the risks of elevated BP.9,10

Conclusion

White-coat hypertension may reflect a disease state as evidenced by the concurrent presence of microalbuminuria, a marker of endothelial dysfunction and end organ damage.11 It is also a risk factor for cardiovascular events in both the diabetes mellitus and nondiabetes mellitus populations.12 Assuming that the BPs taken at home were reliably obtained, this study demonstrated that white-coat syndrome might be more common than currently thought. Additional studies in the office and home setting would be useful to identify higher risk patients with the white-coat syndrome and establish new guidelines based on BPs taken at home or in the physician’s office. The current study raises the question of whether there should be separate guidelines for office BP- and home BP-guided therapies. n