Twenty-four hour blood pressure monitoring: in ambulatory conditions or in-hospital?

Twenty-four hour blood pressure monitoring (BPM) has been shown to carry important advantages in the diagnostic and therapeutic management of hypertensive patients compared to casual office readings [1–4]. The possibility offered by this technique to obtain information on a subject's blood pressure levels and variations in daily life conditions (i.e. in those conditions where blood pressure does exert its actual impact on the cardiovascular system over the years) is usually regarded as one of its most important and unique properties. In this issue of the journal, Conen et al. [5] appear to challenge the relevance of these findings, and provide evidence of the possible usefulness of 24-h BPM in a hospital setting rather than in ambulatory real life conditions. Twenty-four hour BPM was performed in 314 hospitalized patients in a stable clinical state. Among 139 of them with previously unknown hypertension, 53 (38%) were identified as hypertensive based on a mean 24-h blood pressure ≥ 125/80 mmHg. Most of these patients (n = 37) exhibited normal routine blood pressure as measured by the conventional technique during the period spent in hospital. Based on these observations, Conen et al. [5] conclude that 24-h BPM during hospitalization may be useful to detect patients with undiagnosed hypertension. The suggestion is undoubtedly of some interest. It would need to be considered with caution, however, not only because of specific problems of this study, but also due to more general considerations. Concerning specific aspects of Conen's study, a relevant finding was that conventional and 24-h in-hospital diastolic blood pressure values were not different, at variance from what is commonly observed when considering blood pressure recordings performed in daily life. This may depend on the expected lower values of conventional blood pressure measurements obtained in a hospital setting. However, it may also depend on the somewhat artificial conditions during which 24-h BPM was performed (see below), leading to 24-h in-hospital diastolic blood pressure values higher than those commonly obtained in ambulatory conditions out of hospital. The latter possibility is supported also by the authors’ finding of slightly higher in-hospital than out-of-hospital 24-h blood pressure levels, in the few (n = 42) patients who had also reliable 24-h ambulatory blood pressure monitoring (ABPM) data available. Unfortunately, the sample size of the study was not large enough to fully explore the interesting but complex issue of the relation between in- and out-of-hospital 24-h blood pressure levels. Moroever, in-hospital patients were under variable drug treatment, and some of the administered compounds had an effect on blood pressure values, as monitored in hospital over 24 h. Finally, Conen et al. [5] did not include a follow-up period, aimed at systematically evaluating how many of those patients who showed elevated in-hospital 24-h blood pressure were characterized by higher than normal 24-h blood pressure also in their daily life, after discharge from hospital. More generally, we have to acknowledge that blood pressure is a highly variable parameter [6]. It is subjected to large fluctuations according to the time of the day, being higher during morning hours and lowest during sleep [6,7]. Blood pressure is also known to physiologically increase under psychological stress and physical exercise [6,8]. In many subjects, blood pressure rises also when the measurement is performed in a clinical setting [6,9,10], and this pressor effect may be large enough to reach hypertensive levels, a phenomenon known as ‘white-coat’ or ‘isolated office’ hypertension [9–14]. Conversely, some individuals may have lower blood pressure in the doctor's office than during everyday activities, so that hypertension may pass unnoticed if blood pressure is measured only in the physician's office or in the clinic, a condition known as ‘masked hypertension’ [15,16]. This explains why efforts have been directed for many years to improving blood pressure assessment. Strong evidence was obtained supporting the concept that blood pressure determined outside the doctor's office, either by 24-h ABPM or by self-measurement at home, better reflects patient's cardiovascular risk than casual blood pressure readings taken in a medical environment. Experts in the field of hypertension now strongly recommend the use of 24-h ABPM and/or home self-BPM to confirm the diagnosis of hypertension, especially in subjects showing a low cardiovascular risk profile and exhibiting slightly and variably elevated office blood pressure values, possibly affected by a ‘white-coat effect’ [17]. However, this is done without dismissing the current view that blood pressure measured conventionally in a clinical setting should remain the corner stone for hypertension screening and diagnosis. Data obtained by 24-h ABPM has provided important information on the clinical relevance of a number of different blood pressure components. First, 24-h ABP levels and variability have a powerful prognostic value [1–4,17–20]. Second, the condition characterized by elevated office but normal ambulatory blood pressure levels, termed isolated office or ‘white-coat’ hypertension, is not only a frequent phenomenon, but also a phenomenon not immune from clinical consequences [9,11–13,15,21]. Third, the condition characterized by normal office and elevated out-of-office blood pressure, termed ‘masked’ hypertension, is also characterized by a non-negligible prevalence and, in some populations, carries a risk of cardiovascular complications similar to that of sustained arterial hypertension [15,16]. Why does blood pressure monitored during everyday activities better reflect patients’ cardiovascular risk than blood pressure readings obtained in the physician's office or in a clinic setting? Undoubtedly, the availability of multiple measurements over 24 h allows a better estimation of the ‘true’ blood pressure level of the patient. Indeed, the average of repeated blood pressure measurements taken in the usual environment where a patient lives, represents an effective way to assess the prevailing blood pressure levels characterizing a given individual. That this is really the case was indicated by the better reproducibility of blood pressure values monitored over 24 h away from the clinical setting as compared with isolated office blood pressure readings [22]. In the light of these data, the proposal made by Conen et al. [5] in their study to monitor blood pressure during 24 h in hospitalized patients, aimed at diagnosing previously unknown hypertension, represents a somewhat unexpected suggestion. Indeed, the conventional measurement of blood pressure in hospitalized hypertensive patients often shows a marked fall from outpatient levels during the course of hospitalization. Various factors may account for this phenomenon; in particular, the marked reduction in physical activity and the limited exposure to behavioural stress that usually characterize patients lying in a bed over most of the 24 h. On the other hand, the anxiety linked with hospitalization, the stress induced by modern examinations, and the pain or the discomfort associated with many pathological conditions, as well as the fear of discovering a serious disease, can all represent a very peculiar type of stress that might contribute to increase blood pressure monitored during the day and night in hospitalized patients. Therefore, these considerations do not appear to justify the proposal to use 24-h BPM for the routine screening of hypertension in this setting. All studies providing data on ‘blood pressure normalcy’ have been performed in outpatients. Being hospitalized, this means, for a given patient, that something is going wrong with his/her health, and we cannot exclude that such a feeling might deeply affect blood pressure behaviour over 24 h. This may be true when average 24-h or daytime blood pressure levels are concerned. It may be even more important when the nocturnal blood pressure levels and the degree of fall in blood pressure at night are the object of the evaluation. Given the documented clinical relevance of the latter parameters [23–26], the possibility that nocturnal blood pressure might be artificially altered in hospital represents a serious reason of concern when planning a 24-h BPM in such conditions. Moreover, we cannot exclude that a white-coat effect phenomenon might also contribute to a blood pressure increase at the time of the diagnostic and therapeutic procedures carried out during hospitilization. Overall, a number of methodological problems appear thus to interfere with the diagnostic reliability of 24-h BPM performed in patients during their stay in hospital. Last, but not least, no evidence is available that 24-h BPM in hospital conditions might provide clinical information with the same degree of prognostic relevance as the information obtained over recent decades by performing 24-h ABPM in patients’ daily life. All the above considerations appear to question the actual value of in-hospital 24-h BPM in the clinical management of patients with hypertension. In spite of this, it is not difficult to understand how tempting it might be to focus on 24-h blood pressure recordings in hospitalized patients, aiming to attenuate the impact of moment-to-moment blood pressure changes in the diagnosis of hypertension and to make patients’ behaviour more comparable through a standardization of their activity over the day and night time. However, the possibility of achieving such standardization, which includes the planning of similar meal times, and of comparable times to and off bed in all patients, may better apply to clinical research than to the routine management of patients in hospital practice. Indeed, these favourable features of in-hospital recordings, combined with the well known advantages of 24-h BPM in a research setting [27,28], might represent a highly desirable methodological condition for studies aiming to address pathophysiological issues, such as the mechanisms involved in the determination of blood pressure and heart rate variability over 24 h, as well as clinical pharmacology issues, such as the time distribution of the effects of specific antihypertensive drugs or drug combinations. Conen et al. [5] show that a large fraction of patients with unknown hypertension have high 24-h blood pressure levels during hospitalization, and that most of them are also at high cardiovascular risk. This is an important message to all of us. However, its importance should not result in the suggestion to perform 24-h BPM in all hospitalized patients. It should rather emphasize that much still needs to be done, even in countries where patients have easy access to health care, to improve the screening for cardiovascular risk factors and for high blood pressure in the community, and to motivate physicians to treat these conditions appropriately whenever present. Conen et al. [5] correctly conclude that 24-h BPM is a promising method for diagnosing hypertension more precisely, and that it should probably be used more frequently than it is at present. However, 24-h blood pressure recordings should be in most cases obtained in the ambulatory conditions of daily life rather than in an artificial clinical setting, such as during hospitalization. This is in order to avoid a number of confounding factors which might reduce the ability of 24-h BPM to identify subjects’ ‘true’ blood pressure and to quantify its physiological modulation over 24 h.