O r i g i n a l P a p e r

A Home Blood Pressure MonitorEquipped With a Graphic FunctionFacilitates Faster Blood Pressure Controlthan the Conventional Home BloodPressure Monitor

Tomoyuki Kabutoya, MD; Joji Ishikawa, MD; Satoshi Hoshide, MD;Kazuo Eguchi, MD; Kazuyuki Shimada, MD; Kazuomi Kario, MD

In this study, we evaluated whether antihyperten-sive therapy using a home blood pressure monitor(HBPM) equipped with a graphic display ofweekly and monthly averaged blood pressure(BP) can obtain better BP control than theconventional HBPM. Sixty-five hypertensiveoutpatients who had HBP >135 ⁄ 85 mm Hg wereenrolled by 8 doctors in 2 different hospitals. Thepatients were randomly assigned either a graph-equipped HBPM (graph-equipped HBPM group;n=33) or an HBPM without the graph function(conventional HBPM group; n=32). The patientswere treated with antihypertensive medicationstargeting HBP <135 ⁄ 85 mm Hg. After 2 months,the home systolic BP level was lower in thegraph-equipped HBPM group than in theconventional HBPM group (141.3�15.4 vs

147.7�10.8 mm Hg; P<.05); its reduction wassignificantly larger in the former group (11.9 vs5.6 mm Hg; P<.05). Using an HBP device with agraphic display could accelerate the achievementof BP control. J Clin Hypertens (Greenwich).2009;11:422–425. ª2009 Wiley Periodicals, Inc.

Home blood pressure (HBP) often showsvariability between successive readings and

between morning and evening blood pressure(BP) values.1,2 Patients are often afraid of hypo-tension due to overmedication, and physiciansoften hesitate to increase the dosages of hyper-tensive medications for this reason. Therefore,we hypothesized that an additional displayfunction to visualize the change in weekly aver-age BP values might assist physicians in theirtherapeutic decisions and help improve BPcontrol.

The purpose of this study was to evaluatewhether BP control using an HBP monitor (HBPM)equipped with a weekly averaged BP graphic dis-play can lead to better BP control and a higherproportion of patients achieving the target BPcompared with a conventional monitor.

METHODSThis study was conducted by 8 physicians at 2 hospi-tals in Japan. We screened 81 hypertensive outpatientswho were taking antihypertensive medications. After

From the Department of Medicine, Division ofCardiovascular Medicine, Jichi Medical UniversitySchool of Medicine, Tochigi, JapanAddress for correspondence:Kazuomi Kario, MD, Division of CardiovascularMedicine, Department of Medicine, Jichi MedicalUniversity School of Medicine, 3311-1, Yakushiji,Shimotsuke, Tochigi 329-0498, JapanE-mail: kkario@jichi.ac.jpManuscript received December 25, 2008;revised May 1, 2009; accepted May 29, 2009

doi: 10.1111/j.1751-7176.2009.00150.x

THE JOURNAL OF CLINICAL HYPERTENSION VOL. 11 NO. 8 AUGUST 2009422

reviewing the results of the 2-week screeningmeasurement, we excluded 12 patients who hadcontrolled HBP (<135 ⁄85 mm Hg), 3 patients whorefused HBP measurement, and 1 patient whostopped visiting his physician. Sixty-five patients withuncontrolled HBP (>135 ⁄85 mm Hg) were thusenrolled in the final study group.

Each participating institute received approval forthe study from the institutional review board of theJichi Medical University. Written informed consentwas obtained from all of the patients who wereenrolled in this study.

All of the patients were randomly assigned toeither a group using an HBPM with a graphicdisplay of weekly averaged BP (graph-equippedHBPM; n=33) or a group using an HBPM withoutthis function (conventional HBPM; n=32). Thestudy coordinator in each institute allocated thestudy patients using a computerized random num-ber with the computer SPSS software, version 11.0J(SPSS Inc, Chicago, IL).

For each patient, physicians were asked to evalu-ate the HBP (average of morning and evening BP)and to attempt to reduce it to below 135 ⁄85mm Hg in 6 months using any antihypertensivedrugs they considered appropriate.

Home BP was measured using a validated upperarm cuff-oscillometric device.3 The device wasequipped with BP memory in order to recallmeasurements with displays of numeric valuesand graphs of weekly averaged BPs and pulse rates(Figure 1). In the conventional HBPM group, HBPwas measured using the same device but modifiedto deactivate the weekly averaged values andgraph-display function (a plastic plate was attachedto the console to hide these functions). Thus, physi-cians in this group could observe only each BPvalue.

HBP was measured on the nondominant upperarm in the sitting position after 2 minutes of rest.In both groups, the HBPM measurements wereautomatically repeated 3 times per occasion with15-second intervals. Morning BP was measuredwithin 1 hour after waking, after urination,and before breakfast and taking antihypertensivemedication.4 Evening BP was measured immedi-ately before going to bed. HBP was defined by anaverage of morning and evening BP over the2 weeks immediately before visiting the physician’soffice.

Clinic BP was measured 3 times per occasionafter a rest of at least 5 minutes in the sitting posi-tion using the same HBPM device, which patientsbrought with them to the clinic.

RESULTSThe mean � SD age was 67�13 years, 46% ofparticipants were male, and the mean number ofmedicine was 1.9�1.0. The 2 groups were similarin terms of age, sex, demographic characteristics,antihypertensive drug use before enrollment, andprevalence of coexisting cardiovascular conditions.

Home systolic BP was similar in both groups(the graph-equipped HBPM group, 153.2�14.9mm Hg; the conventional HBPM group, 153.1�12.3mm Hg) at baseline. Home systolic BP in the graph-equipped HBPM group was significantly lower thanthat in the conventional HBPM group after 2 monthsof treatment (141.3�15.4 vs 147.7�10.8 mm Hg,P<.05; Figure 2). The reduction of home systolic BPin the graph-equipped HBPM group was also signifi-cantly larger than that in the conventional HBPMgroup after 2 months (11.9 vs 5.6 mm Hg; P<.05).The percentage of patients with HBP below 135 ⁄85mm Hg in the graph-equipped HBPM group washigher than that in the conventional HBPM groupafter 2 months (41% vs 13%, P<.05; Figure 2),4 months (40% vs 11%, P<.05; Figure 2),and 5 months (37% vs 16%, P<.05; Figure 2) oftreatment.

Clinic systolic BP in the graph-equipped HBPMgroup was significantly lower than that in the con-ventional HBPM group after 3 months of treatment(143.3 vs 153.0 mm Hg; P<.05).

Patients in the graph-equipped HBPM groupwere more medicated than those in the conven-tional HBPM group after 5 months (3.58 vs 2.73drugs; P<.03) and 6 months (3.74 vs 2.76 drugs;P<.02) of treatment.

DISCUSSIONIn this study, we showed that antihypertensivetreatments using an HBP device with a graphic dis-play of weekly averaged BP values can lead to

Figure 1. Analytical functions on the graph-equippedhome blood pressure (BP) monitor. Left panel: atransition of weekly BP (the dotted line shows thetarget home BP [135 ⁄ 85 mm Hg]). Right panel: weeklyBP (systolic BP, diastolic BP, and pulse rate).

VOL. 11 NO. 8 AUGUST 2009 THE JOURNAL OF CLINICAL HYPERTENSION 423

faster BP control than that using the conventionaldevice.

The faster BP control in the graph-equippedHBPM group could be explained as follows. First,by referring to the screen showing the graph andthe mean BP values, both the physician and thepatient were able to determine at a glance whetherthe HBP was being controlled sufficiently. If thepatient’s BP was not being controlled below135 ⁄85 mm Hg, this fact was clearly presented onthe screen in graphic and numeric fashion. The‘‘easy-to-read’’ display helped to convince patientsof the need to dose-up their antihypertensive medi-cations.

Second, physicians and patients often are con-cerned about temporary hypotension due to over-dosage. Patients displaying a strong white-coat

effect are particularly vulnerable to hypotension,since they may be over-dosed based on theirinflated clinic BP levels.5 Even if the physicianrefers to the HBP in such cases, he ⁄ she may not beconfident about this data due to its considerablevariation, and therefore may not be able to use itin making a dosing decision.

CONCLUSIONSBP management using an HBP device with a gra-phic display of weekly averaged BP values can facil-itate faster BP control.

Acknowledgments: We thank all of the participants in thisstudy: Tsukasa Yagi, Naoki Fujiwara, and Masashi Gokita,Ogano Central Hospital, and Takako Taketomi, HisaoShimamura, Toshiyuki Setoyama, and MasatoshiTeshigawara, Chichibu Municipal Hospital.

Figure 2. Blood pressure (BP) levels in the study population. Filled circles, black bars=graph-equipped home BPmonitoring (HBPM) group; open circles, white bars=conventional HBPM group. *P<.05 vs the conventional group byan unpaired t test or the chi-square test. (A) change of home systolic BP, (B) proportion of patients achieving BP goal,(C) change of morning systolic BP, (D) change of evening systolic BP.

THE JOURNAL OF CLINICAL HYPERTENSION VOL. 11 NO. 8 AUGUST 2009424

REFERENCES

1 Stergiou GS, Skeva II, Zourbaki AS, et al. Self-monitoringof blood pressure at home: how many measurements areneeded? J Hypertens. 1998;16:725–731.

2 De Gaudemaris R, Chau NP, Mallion JM. Homeblood pressure: variability, comparison with office readingsand proposal for reference values. Groupe de la mesure,French society of hypertension. J Hypertens. 1994;12:831–838.

3 Anwar YA, Giacco S, McCabe EJ, et al. Evaluation of theefficacy of the Omron HEM-737 IntelliSense device for use

on adults according to the recommendations of the associa-tion for the advancement of medical instrumentation.Blood Press Monit. 1998;3:261–265.

4 Kario K, Matsui Y, Shibasaki S, et al. An alpha-adrenergicblocker titrated by self-measured blood pressure recordingslowered blood pressure and microalbuminuria in patientswith morning hypertension: the Japan morning surge-1study. J Hypertens. 2008;26:1257–1265.

5 Kabutoya T, Ishikawa J, Hoshide S, et al. Determinants ofnegative white-coat effect in treated hypertensive patients:the jichi morning hypertension research (J-MORE) study.Am J Hypertens. 2009;22:35–40.

VOL. 11 NO. 8 AUGUST 2009 THE JOURNAL OF CLINICAL HYPERTENSION 425