ED: raji Op: hashmath sc: micheal : lww_mbp_

Transcription

1 ED: raji Op: hashmath sc: micheal : lww_mbp_20021 Original scientific paper 1 Home-based self-measurement of blood pressure: a proposal using new reference values (the PURAS study) Juan A. Divisón-Garrote, Carlos Sanchis, Luis M. Artigao, Julio A. Carbayo, Lucinio Carrión-Valero, Enrique López de Coca, Esperanza Martínez-Navarro, Javier Massó and on behalf of the Grupo de estudio de enfermedades vasculares de Albacete (GEVA) Objectives To establish reference values for blood pressure by means of self-measurement of blood pressure (BP) conducted at home. Design Descriptive study of the distribution of self-measured BP at home and its correspondence with clinicbased measurements of BP. Methods The aim of this study is to define the home BP levels that correspond to clinic BP thresholds 1/ 90 (hypertension) and 130/85 (normality). The sample consisting of 1411 randomly selected adults stratified by age and gender. A pre-calibrated electronic device (Omron 5CP) was used for BP and heart rate (HR) measurements and a trained nurse performed clinicbased sphygmomanometer measurements. The same nurse provided tutorials for the subjects on how to obtain 12 self-measured BP values at home using the Omron device in a single day. Results Of the 1184 volunteers that showed up at the appointment, 195 were known as hypertensives and were excluded from the study. The average age of the remaining 989 subjects (.4% females) was 44.3 years. Clinic BP values were significantly higher than self-measured BP at home regardless of age and gender. Both had good correlations (systolic BP, r = 0.84 and diastolic BP, r = 0.77). Using linear regression, the self-measured BP at home hypertension threshold would be 131/82 and the limit of normality 123/78. Using corresponding percentiles, these values would be 134/85 and 124/, respectively. Conclusions The self-measured BP at home values found in this study, when defining hypertension, are lower than values currently accepted (135/85 ). Long-term studies are necessary to confirm these results. Blood Press Monit 9: c 2004 Lippincott Williams & Wilkins. Blood Pressure Monitoring 2004, 9: Keywords:,, Centro Atentión Primaria, Albacete, Spain. Correspondence and requests for reprints to Juan Antonio Divisón-Garrote, MD, Coordinador médico, Centro de Salud de Casas Ibáñez, Carretera de Villamalea s/n, Casas Ibáñez (Albacete), Spain. jdivisong@medynet.com Received Accepted AQ2 AQ1 Introduction In recent years the self-measurement of blood pressure (BP) at home (HBP) has begun to be increasingly used since it can solve some of the limitations of the clinicbased measurements of blood pressure (CBP). These limitations include failure to diagnose isolated clinical hypertension, the white coat phenomenon, BP measurement variability, observer bias, low reproducibility, and poor correlation with damage in target organs. Also, a better cost-effective relationship has been observed with self-measured HBP than with ambulatory blood pressure measurements (ABPM) [1]. Several studies have shown that with self-measured HBP the mean values [2,3] reproducibility [4,5] and correlation with damage in target organs [6 8] are similar to the ABPM with good correlations between both methods. To be more generally used in diagnosis and follow-up of hypertensive patients, it is necessary to know the values that define the threshold of normality. To achieve this goal, long-term studies are the most appropriate. Unfortunately, the costs of these studies are high and so the available data are scarce. The only reported study of this kind was performed by the Ohasama group [9,10], who suggested 137/84 as a reference value. Crosssectional studies are more accessible and, with appropriate statistical safeguards, can provide good reference values. Reference values based on the mean plus two standard deviations or on the ninetieth or ninety-fifth percentiles [11,12] have very little clinical interest, describing results without any relation to cardiovascular mortality. More interesting are the values obtained from regression lines or from the correspondence of the c 2004 Lippincott Williams & Wilkins DOI: /01.mbp d

2 2 Blood Pressure Monitoring 2004, Vol 9 No 5 different cut-off points of CBP for which the threshold of normality is known and can how it relates to cardiovascular mortality can be established. The studies with this methodology are scarce. Even more infrequent is this kind of study in general populations and based on selfmeasured HBP. One of these is the PAMELA study [13], that relied on the values of two self-measured HBP values (one in the morning and one in the evening) in a single day with 1438 subjects. Dübendorf et al., [14] used a similar methodology but over several days using 3 subjects. The study by Didima et al., [15] had four selfmeasurements (two in the morning and two in the evening) over three days with 562 subjects. Our exhaustive search of the literature indicated a lack of studies involving large and general population samples and multiple self-measured HBP s. Neither were there any studies conducted in a Spanish population that would provide data on self-measured HBP distributions. The aim of the present study is to know which values of multiple self-measured HBP correspond with those derived from the regression lines, percentiles and cutoff points of the CBP. Subjects and methods This is a descriptive study performed in a general population in semi-urban and rural areas of the Albacete Province of Spain. These areas were selected based on a previous study of the prevalence of cardiovascular disease risk factors [16] in which the population pyramid of the selected areas and the mean in blood pressure values did not differ from the overall population of the Province. Furthermore, the participation index (83.7%) in the now selected areas was the highest of all those studied [16]. The actual selection includes the following population centres: La Roda, Casas Ibáñez, Villamalea, Fuentealbilla and Herrumbalr. The study sample of 1411 subjects (randomised and stratified with respect to age and gender) was selected from the general population census. These individuals were contacted by post with a letter explaining the background and proposals of the study and in which a specific appointment time and date to attend their local Primary Health Care Centre [PHCC] was proposed. Those who did not show up at the appointment were contacted by telephone and the aims of the study were again explained, collaboration urged and a new appointment issued. All those who agreed and kept their clinical appointment, attended the PHCC between 0900 h 1 h and were attended to by a nurse who had been previously trained to correctly perform the BP measurements as defined for the study and who had successfully passed an approval test (video of the British Hypertension Society). To start, a general questionnaire was filled in, collecting name, personal and family clinical history, habits and prior treatments. This was followed by a more specific questionnaire on the diagnosis of hypertension. Subsequently, the BP measurement was performed under standard clinic consulting-room conditions. Initially, three measurements separated by 2 3 minutes each were made with an electronic device (OMRON 5CP) that had been previously validated [17,18]. The data were noted together with the heart rate (HR). After this, three more BP measurements were made using a mercury sphygmomanometer, also with 2 3 minutes intervals after which the HR was measured again. In all the measurements, a cuff size appropriate for the patient s arm was used. The nurse then instructed the subjects, or the family member who assumed responsibility for the home-based measurements, in the use of the automatic electronic device and the nurse observed the procedure to ensure that a minimum of two self-measurements were performed correctly. At home, the patients recorded a total of 12 selfmeasurements, three measurements in each of the following four time-periods: before lunch, before dinner, at bedtime and in the morning on arising from bed. The results of the BP and HR were noted on clean datacollection sheets that were provided to the subjects. At the conclusion of the 12 self-measurements, the subject filled-out a questionnaire containing five questions on the self-measurement procedure, the person who assumed responsibility for the measurement and on any difficulties encountered. All measuring devices were calibrated prior to being used in the fieldwork. Statistical methods We have calculated the mean values and their standard deviations of the different BP and HR measurements (with the electronic device and with mercury sphygmomanometer) obtained in the clinic consulting rooms and at home. Correlations between the different methods of measurement were also calculated using Pearson s correlation coefficient. Differences in the means of the BP measured in the clinic and at home were assessed using Student s t-test for paired data. The cut-off point for the diagnosis of hypertension on the clinic measurement was established as 1/90 and the limit of normality as 130/85. These values have been recommended by recent guidelines [19,20]. From these cut-off points, and using regression lines or with the corresponding percentile distributions, the corresponding values for the homebased measurements were calculated. A maximum alpha error of 5% was tolerated. Results Of the 1411 subjects initially approached, 1184 (83.9%) attended the clinic appointments and fulfilled their 12- measurement schedule at home. Among these, 195 were

3 Home versus clinic blood pressure values Divisón-Garrote et al. 3 known hypertensive subjects undergoing treatment and their data were excluded from the study. Therefore, the final sample consisted of 989 subjects with a mean age of 44.3 years (SD = 16.4) 498 of whom (.4%) were female. The measurements at home were described as being easy by 89% of the subjects and were self-performed by 77% of them. The mean BP and HR values obtained in the consulting room were significantly higher than those measured at home (see Table 1). In the consulting room, the mean BP and HR values measured by the electronic device were significantly higher than those obtained with the mercury sphygmomanometer (Table 1). The distribution curve of self-measured HBP was normal although shifted to lower values than the CBP curve obtained with the electronic device (Figure 1). The mean values of BP, not only those obtained in the clinic with the electronic device but also those based on self-measured HBP increase with age in a similar way in males and females, being always higher in males (Figure 2). The differences in systolic pressure increase the older the subject, at least up to years of age in males and up to years in females. The differences between clinic and home pressure measurements were significant except for ages over (Figure 2 and Table 2) and similar for males and females. With respect to the HR, the values were greater in females and the distribution was somewhat irregular with respect to age. Heart rates were higher in the clinic than at home although the differences are significant only for females (Figure 3) but not clinically relevant. The mean value of the difference was 1.4 beats/minute for males (P > 0.05) and 2.4 beats/ min for females (P < 0.05). Table 1 Mean blood pressure and heart rate frequency values segregated with respect to the different measuring methods used Measurement BPS (SD) BPD (SD) HR (SD) beats/min Clinic OMRON 1st (21.7) 76.5 (11.7) 74.8 (11.5) 2nd 3rd (19.9) 74.2 (11) 74 (10.7) 1st 3rd (20.2) 75 (10.8) 74.3 (10.7) Clinic Sphigmo 1st (19) 75 (10.5) 73.8 (10) 2nd 3rd.7 (18.5) 73.8 (10.4) 73.1 (9.8) 1st 3rd (18.6) 74.2 (10.4) 73.3 (9.8) HBP (18.8) 71.1 (9.7) 71.4 (8.9) 8 (exclude 1st) (16.5) 62.6 (8.7) 71.3 (8.9) 4 (1st only).1 (19.4) 72.2 (10.2) 71.5 (9.1) Midday (x3) (23.3) 72.1 (10.7) 72.6 (10.6) Evening (19.3) 71.9 (10.5) 72.1 (10.4) Bedtime (19.1) 69.2 (10.5) 72.2 (9.8) Awakening (21.5) 71.3 (11.3) 68.8 (9.7) BP S, Systolic blood pressure; BP D, Diastolic blood pressure; SD, Standard deviation; HR, Heart Rate; HBP, self-measurement of blood pressure at home. Fig. 1 % % BP S frequency distribution BP D frequency distribution Distribution curves of the blood pressure measurements conducted in the clinic (dotted line) and at home (solid line). Panel A: Systole; Panel B: Diastole. The CBP and self-measured HBP show a high correlation. A higher correlation was obtained if we use the three clinic measurements performed with the electronic device (BPS; r = 0.84 and BPD; r = 0.77) instead of the three measurements with the mercury sphygmomanometer (BPS; r = 0.83 and BPD r = 0.73). The correlation was slightly lower if only the first clinic-based measurement with the electronic device was used (BPS r = 0.81 and BPD r = 0.71). With regard to the HR, a high correlation also existed between the clinic and home measurements with a higher correlation being observed between the three HR measurements made with the electronic device (r = 0.73) by the nurse in the interval of the three measurements of BP made with the mercury sphygmomanometer (r = 0.72). Figures 4 and 5 show, for each individual, the blood pressure data (clinic against home measurements, both using the electronic device). The differences between CBP and self-measured HBP distributions indicate that the values defining normality for self-measured HBP have to be lower than in the case of CBP. The excellent

4 4 Blood Pressure Monitoring 2004, Vol 9 No 5 Fig. 2 Fig. 3 (A) Systolic blood pressure age Beats/min (A) Heart rate (Males) > Age (B) 85 Diastolic blood pressure. (B) 78 Heart rate (Females) Beats/min age Blood pressure values measured in the clinic (dotted line) and at home (solid line) segregated with respect to age. Panel A: Systole; Panel B: Diastole > Age Heart rate measured in the clinic (dotted line) and at home (solid line). Panel A: Heart rate males; Panel B: Heart rate females. Table 2 Differences in the clinic and home-based measurements of blood pressure Age BP S (SD) BP D (SD) (9.1) P < (6.7) P < (9) P < (6.3) P < (9.4) P < (6.7) P < (10.1) P < (6.9) P < (14.3) P < (6.9) P < (13.8) P = (7.5) P = (13.8) P = NS 2 (8.4) P = NS Differences in ; BP S, Systolic blood pressure; BP D, Diastolic blood pressure; SD, standard deviation. correlation allows one to deduce the limits of normality from the corresponding values in the regression line or from the correspondence of percentiles. The values of the self-measured HBP in the selfmeasurement HBP-CBP regression line that match the cut-off point for the diagnosis of hypertension using CBP (1/90 ) are 130.7/81.6 in the overall population sample (130.7/81.8 in men and 130.5/ 81.2 in women). Using the same definition, the limit of normality was established at 122.8/78.1 in the overall population (123.8/78.5 in men and 122.2/77.6 in women). Table 3 summarizes these values, and their corresponding confidence intervals (95% CI) and probabilities calculated using the regression equation that defines the diagnosis of hypertension, with respect to age and gender. The values of self-measured HBP corresponding to the same percentiles of the cut-off points of the CBP obtained with the electronic device, used to establish the reference values, provide the diagnosis of hypertension values for the overall population as 134.2/84.5 (132.3/85 for males and 135.9/84.5 for females), and the corresponding limit of normality as 123.8/.4 for the overall population (123.8/ 79.9 for males and as125/.6 for females). If we compare this with the reference values from Tsuji et al., [10] (percentile for the BPS and percentile 87 for BPD) our 12 home-based measurements provide a value for the diagnosis of hypertension of 133.2/81.7 for the overall population.

5 Home versus clinic blood pressure values Divisón-Garrote et al. 5 Fig. 4 BPD SMBPH BPS MBPH BP Dispersion Systolic; Clinic-SMBPH BP D -Clinic Correlation = 0.77; P < BP Dispersion Systolic; Clinic-SMBPH 1 BP S -Clinic 1 1 Correlation = 0.84; P < Correlations between clinic and home-based blood pressure self-measurements. SMPBH: Self-measurements blood pressure at home. Fig. 5 CF Home CF Dispersion Clinic versus Home measurement CF Clinic beats/min Correlation = 0.73; P < Correlation between clinic and home-based measurement of heart rate. Table 4 summarizes the values that, in the selfmeasurement HBP, correspond to the mean values plus standard deviations and with the ninetieth and ninetyfifth percentiles. Of the 12 self-measurements, eight values were used (the first measurement of each measuring period has been excluded). Discussion The population sample in this study is representative of the general population of the Albacete Province with a very close distribution in age and gender. Certain limitations of other published studies have been avoided. For example, in the PAMELA study [13] with one of the largest sample sizes to date, only two home-based measurements were performed (one in the morning and one in the evening). The study by Dübendorf et al., [14] had measurements over 14 days but still with only one in the morning and one in the evening. The validity of the first measurement of a given session has been seriously questioned by several investigators [6] and therefore raises questions on the interpretations of the results of these two studies. Also, the measurements on the first day should be discarded and only the values on the subsequent days be used according to several authors

6 6 Blood Pressure Monitoring 2004, Vol 9 No 5 Table 3 Prediction of home PB values that correspond to clinic BP S = 1 and BP D =90 Group BP S () BP D () Mean CI Mean CI Total Males Females Males: years Females: years BP S, Systolic blood pressure; BP D, Diastolic blood pressure; CI, Confidence interval. Table 4 Self-measurement BP at home values according to the different statistical measurements, expressed in Measurements X + 2 SD P 90 P 95 BP S BP D BP S BP D BP S BP D n = n = 8 (exclude 1 st ) SD, Standard deviation; P 90, value of the 90th percentile; P 95, value of the 95th percentile; BP S, Systolic blood pressure; BP D, Diastolic blood pressure. [2,21]. However, it would appear that the important issue is the number of self-measurements conducted. One single day could be sufficient provided that a certain minimum of measurements were performed [5,6,22]. Significant differences between the mean values of the clinic BP measurements developed with the electronic device and mercury sphygmomanometer exist. These differences are probably due to the sequence of measurements and also to the number of measurements, as well as the white-coat effect of the subjects decreasing. Another important issue is that the same electronic device has to be used for the clinic and home measurements to avoid bias due to the device. As in other studies [5,13,14,23,24] this study confirms the fact that CBP gives higher values than self-measured HBP. The differences between these two kinds of measurement were similar in both genders and in all age groups, indicating independence with respect to these two variables. As in the PAMELA study [13], we did not observe any increase in the differences with the age of the subjects and we agree with Mancia et al., [13] in that to consider home-based values of 1/90 as the normal blood pressure upper limit would be wrong. The possible alarm reaction (or white coat syndrome ) in the clinic, characterized by a rise in blood pressure response and by tachycardia, does not appear to be the cause of the observed differences. The heart rate differences between the clinic and home measurements were significant in the overall population but, being of the order of 2 3 beats/min (clinically irrelevant), it is risky to totally assign them as an alarm reaction. In addition, if we exclude heart rate measurements of the subjects on waking-up, the differences are only of 1 2 beats/min and therefore not statistically significant. On the other hand, heart rates have shown an irregular distribution with respect to age in our sample. It is worth noting that the correlation between the CBP and self-measured HBP in the present study has been the highest among the known studies that have performed similar comparisons. The correlations were 0.81/0.75, 0.76/0.77, 0.76/0.75, and 0.73/0.64 in the references 15,14,25 and 13, respectively. This fact stresses the importance of our study to establish correspondence between the CBP and self measured HBP reference values. When comparing the data from the present study with other studies that have used similar criteria in trying to establish reference values for self-measured HBP, we observe that, using regression criteria, the values that we obtain are lower than those from the study of Didima et al., [15] and similar to the upper limit of the range proposed by the PAMELA study [13]. Using the criteria of percentile correspondence, the values obtained in our study are also lower than those of Didima et al., [15]. Using the same methodology as in the Dübendorf study [14] with only the first clinic measurement being considered, our study also provides somewhat lower values by comparison. However, in explaining these small differences it has to be noted that the population studied in the Dübendorf study comprised a slightly older age range and contained a lower percentage of females. In Didima et al., [15] using both regression criteria and percentile correspondence, the proposed values are higher than those of all the other studies. In addition, it is the only study in which the self-measured HBP values exceed those from the CBP. Table 5 summarizes the characteristics and the reference values proposed by the principal studies conducted to date. One aspect not yet clear, and which continues to generate controversy, is the number of measurements to

7 Home versus clinic blood pressure values Divisón-Garrote et al. 7 Table 5 Reference values for self-measurement BP at home from the cross-sectional studies published to date Study n Age Mean; SD Female BP Clinic BP Home Percentiles (1/ 90) Regression (1/ 90) PAMELA [13] ; 12 % 127/82 119/ /75 81 Didima [15] 562 > ; 17 58% 118/73 / / /82.7 Dübendorf [14] 3 > ; 13 47% 130/82 123/ /85.8 French HS [25] % 147/86 PURAS 989 > ; 16 % 126/76 118/ / /81.6 AQ3 AQ1 be taken into account in establishing reference values. In the different studies described above, the criteria have been quite different. For home-based measurements, consensus documents [19,20,26] do not provide clear guidelines. Our conclusion is that there is a need to perform repeated measurements in order to achieve good reproducibility and correlation with organic illness and with the morbi-mortality [10,27]. Therefore, we designed our study with 12 self-measurements performed in a single day, which we considered would be reproducible and representative of the mean blood pressure values of an individual over the time-course of a single day. Our previous experience [5,6] provided evidence that repeated measurements performed in a single day, as proposed for this study, show the same reproducibility than ambulatory monitoring and have a similar relationship to organic illness [5,6]. It appears that the measuring device, as well, can generate a certain alarm reaction when first encountered. Hence, some investigators [6] propose excluding the first measurement when the device is being used for the first time. In the present study we attempted to get around this problem by having a trained nurse guide the subject through the process. Even so, excluding the first measurement in each one of the self-measurement sessions would result in even lower proposed reference values (Table 4). With regard to the clinic measurements, the alarm reaction of the subject appears to be towards the environment of the consulting rooms and not towards the use of the electronic device, because ignoring the first measurement did not result in any differences. Our exclusion of the known hypertensives on treatment from the analysis implies a high probability of not including the white coat hypertensives. We had some doubts about using the first clinic measurement alone which previously had been related to cardiovascular morbi-mortality in long-term studies, or using the three measurements, because they correlate better with the home measurements in the end, we decided to use the mean values from the three measurements developed in the clinic with electronic device. One limitation of this study could be that the measurements were all made over a single day. However, epidemiological studies that relate clinical BP with cardiovascular risk are usually based on single day measurements. Other statistical strategies, such as the mean value plus two standard deviations or the ninetieth and the ninety-fifth percentiles that some authors have proposed, appear to over-estimate the values (Table 4). The reference value of 135/85 as the limit of normality for self-measured HBP could be an overestimate and, as stated by Myers et al., [29], appears to be based more on received opinion rather than on real data. From the data obtained in the present cross-sectional study, and while awaiting results from long-term studies that may or may not confirm our findings [28], our proposal is to consider values above 134/85 for the diagnosis of hypertension when looking at self-measured HBP and values of 124/ as the limits of normality, with only slight differences when the age and the gender of the subject are taken into account. The Ohasama study [9] proposes a higher threshold to define hypertension (137/84 ), and the reasons for the difference with our study could be that the sample of this study was older than years (mean age.9 years) and the subjects made only one home BP measurement each day. Acknowledgements We dedicate this study to the late Dr. Angel Puras- Tellaeche in warm appreciation of his role as visionary leader, guide and mentor of the GEVA group. We thank the Consejería de Sanidad de la Junta de Comunidades de Castilla-La Mancha for the funding (Health Council of the Community of Castilla la Mancha, Spain; DOCM n Mayo 1998) to conduct this study. References 1 Appel LJ, Stason WB. Ambulatory blood pressure monitoring and blood pressure self-measurement in the diagnosis and management of hypertension. Ann Intern Med 1993; 118: Chatellier G, Dutrey-Dupagne C, Vaur L, Zannad F, Genes N, Elkik F, et al. Home self blood pressure measurement in general practice. The SMART study. Am J Hypertens 1996; 9: Stewart MJ, Gough K, Reid M. White-coat hypertension: a comparison of detection using ambulatory blood pressure monitoring or home monitoring of blood pressure. J Hypertens 1996; 14:17 18 (Abstract). 4 Mancia G, Ulian L, Parati G, Trazzi S. Increase blood pressure reproducibility by repeated semiautomatic blood pressure measurements in the clinic environment. J Hypert 1994; 12: Divisón JA, Puras A, Sanchis C, Artigao LM, López Abril J, López de Coca E, et al. Exactitud y precisión en la medida de la presión arterial. Estudio comparativo de las automedidas domiciliarias con la medida de la presión

8 8 Blood Pressure Monitoring 2004, Vol 9 No 5 arterial en la consulta y la monitorización ambulatoria de la presión arterial. Aten Primaria 2001; 27: Divisón JA, Puras A, Aguilera M, Sanchis C, Artigao LM, Carrión L,et al. Automedidas domiciliarias de presión arterial y su relación con el diagnóstico de la hipertensión arterial y con la afectación orgánica: estudio comparativo con monitorización ambulatoria. Med Clin (Barc) 2000; 115: Abe H, Yokouchi M, Nagata S, Ashida T, Yoshimi H, Kawano Y, et al. Relation of office and home blood pressure to left ventricular hypertrophy and performance in patients with hypertension. High Blood Press 1992; 1: Haley WE, Harris TM, Tucker CT, Zachariah PK. Microalbumin excretion may relate better to home BP than office BP in patients with stage II-III hypertension. J Am Soc Nephrol 1996; 7:15. 9 Tsuji I, Imai I, Nagai K, Ohkubo T, Watanabe N, Minami N, et al. Proposal of reference values for home blood pressure measurement. Prognostic criteria based on a prospective observation of the general population in Ohasama Japan. Am J Hypertens 1997; 10: Ohkubo T, Imai Y, Tsuji I, Nagai K, Kato J, Kikuchi N, et al. Home blood pressure measurement has a stronger predictive power for mortality than does screening blood pressure: a population-based observation study in Ohasama Japan. J Hypertens 1998; 6: Thijs L, Staessen J, Celis H, De Gaudemaris R, Imai Y, Julius S, et al. Reference values for self-recorded blood pressure. A meta-analysis of summary data. Arch Intern Med 1998; 158: Staessen J, Fagard R, Lijnen P, Thijs L, van Hulle S, Vyncke G, et al. Ambulatory blood pressure and blood pressure measured at home: progress report on a population study. J Cardiovasc Pharmacol 1994; 23(Suppl 5): Mancia G, Sega R, Bravi C, de Vito G, Valagussa F, Cesana G, et al. Ambulatory blood pressure normality: results from the PAMELA study. J Hypertens 1995; 13: Weisser B, Grüne S, Burger R, Blickenstorfer H, Iseli J, Michelsen SH, et al. The Dübendorf Study: a population based investigation on normal values of blood pressure self-measurement. J Hum Hypertens 1994; 8: Stergiou G, Thomopoulou G, Skeva I, Mountokalakis T. Home blood pressure normalcy: the Didima study. Am J Hypertens 2000; 13: Puras A, Sanchis C, Artigao LM, Divisón JA. Prevalence, awareness, treatment and control of hypertension in a Spanish population. Eur J Epidemiol 1998; 14: Artigao LM, Llavador JJ, Puras A, López Abril J, Rubio MM, Torres C, et al. Evaluación y validación de los monitores Omron 5CP y 6/711 para automedidas de presión arterial. Aten Primaria 2000; 25: O Brien E, Mee F, Atkins N, Thomas M. Evaluation of three devices for self measurement of blood pressure according to the revised British Hypertension Society protocol: the Omron HEM 5CP, Philips HP5332 and Nissei DS-175. Blood Press Monit 1996; 1: The Seventh Report of the Joint National Committee on Prevention. Detection, Evaluation, and Treatment of High Blood Pressure. The JNC 7 Report. JAMA 2003; 289: European Society of Hypertension-European Society of Cardiology. Guidelines for the management of arterial hypertension. J Hypertens 2003; 21: Stergiou GS, Skeva II, Zourbaki AS, Mountokalakis TD. Self-monitoring of blood pressure at home. How many measurements are needed? J Hypertens 1998; 16: Mancia G, Ulian L, Parati G, Trazzi S. Increase blood pressure reproducibility by repeated semiautomatic blood pressure measurements in the clinic environment. J Hypert 1994; 12: Stergiou GS, Voutsa AV, Achimastos AD, Mountokalakis TD. Home selfmonitoring of blood pressure. Is fully automated oscillometric technique as good as conventional stethoscopic technique? Am J Hypertens 1997; 10: Schettini C, Bianchi M, Nieto F, Sandoya E, Senra H, for the Hypertension Working Group. Ambulatory blood pressure. Normality and comparison with other measurements. Hypertension 1999; 34: DeGaudemaris R, Chau NP, Mallion JM, for the groupe de la mesure, French Society of Hypertension. Home blood pressure: variability, comparison with office reading and proposal for reference values. J Hypertens 1994; 12: Asmar R, Zanchetti A. Guidelines for the use of self blood pressure monitoring: a summary report of the first international consensus conference. J Hypertens 2000; 18: Jula A, Puukka P, Karanko H. Multiple clinic and home blood pressure measurements versus ambulatory blood pressure monitoring. Hypertension 1999; 34: Genes N, Bobrie G, Vaur L, Chatellier G, Vaisse B, Mallion JM. Current aspects of arterial hypertension. Prognostic value of self-monitoring blood pressure measurements in aged hypertensive patients: a SHEAF study protocol. Press Med 1999; 28: Myers MG. Blood pressure self-measurement: where do we go from here? Am J Hypertens 2000; 13:

9 JOURNAL NAME MBP 7/27/04 ARTICLE NO: QUERIES AND / OR REMARKS AUTHOR QUERY FORM LIPPINCOTT WILLIAMS AND WILKINS Query No AQ1 AQ2 AQ3 Details Required Please provide received and accepted dates Please provide keywords Morbi-mortality? Please check the language ~ do they mean morbidity? Authors Response 1