Home blood pressure and treatment decisions

Hypertension is a well recognized global public health issue and a major cause of cardiovascular events and death throughout the world [1–3]. It is responsible for about 13% of all deaths and 3.7% of total disability-adjusted life-years [1–3]. Globally, the overall prevalence of hypertension in adults was around 40% in 2008, and the number of adults with hypertension in 2025 has been projected to increase by about 60% to a total of 1.56 billion [1–5]. Hypertension is a chronic disease that requires periodic and correct measurements of blood pressure (BP) through three different but complementary alternatives: office BP, home BP, and 24-h ambulatory BP. Measurement of BP in the office is the cornerstone on which the knowledge of hypertension is based. Antihypertensive treatment targeted on office BP reduces the risk of cardiovascular disease, and the degree of benefit is associated with the magnitude of BP reduction [6–13]. However, office-based BP readings are limited in the amount of information they can provide as they represent a single snapshot in time [14–16]. Conversely, ambulatory BP monitoring (ABPM) provides a direct record of BP throughout the whole day in patients engaged in their usual activities. Frequent readings recorded at predefined intervals during wakefulness and sleep enable clinicians to obtain a more precise estimation of a patient's BP, to asses BP levels in the outpatient setting, and to study BP variability and circadian BP profile [15]. Furthermore, the evidence that ABPM provides information over and beyond conventional BP measurement has been growing steadily over the past 25 years, and the rationale for its use in clinical practice is soundly based [14,16–19]. Of note, clinical studies focused on the prognostic value of on-treatment ambulatory BP showed that ambulatory BP predicts cardiovascular events even after adjustment for classic risk factors including office measurements of BP [20–22]. In the office versus ambulatory blood pressure study, Clement et al.[21] assessed the association between ambulatory BP in treated patients and subsequent cardiovascular events with a median follow-up of 5 years. After adjustment for several confounders, including BP measured at the physician's office, higher mean values for 24-h ambulatory SBP and DBP were independent risk factors for new cardiovascular events. The adjusted relative risk (RR) of cardiovascular events associated with a 1-SD increment in BP was 1.34 [95% confidence interval (CI): 1.11–1.62] for 24-h ambulatory SBP, 1.30 (95% CI: 1.08–1.58) for ambulatory SBP during the daytime, and 1.27 (95% CI: 1.07–1.57) for ambulatory SBP during the night-time. For ambulatory DBP, the corresponding RRs of cardiovascular events associated with a 1-SD increment were 1.21 (95% CI: 1.01–1.46), 1.24 (95% CI: 1.03–1.49), and 1.18 (95% CI: 0.98–1.40). Similarly, in a prospective analysis of the Progetto Ipertensione Umbria Monitoraggio Ambulatoriale study [20], only 27% of treated hypertensive patients achieved office BP control (defined as BP < 140/90 mmHg), and 37% of patients achieved ambulatory BP control (defined as daytime BP < 135/85 mmHg). Ambulatory BP control significantly predicted a lesser risk for subsequent cardiovascular disease, whereas office BP control did not [20]. To date, ABPM appears to be an indispensable diagnostic tool in patients with established or suspected hypertension to refine cardiovascular risk stratification and guide therapeutic strategies. However, the use of ABPM for hypertension management may be more expensive than that of office or home BP. This aspect has been addressed by a recent comparative study estimating the resources consumed and subsequent costs for hypertension management, using home BP monitoring alone versus combined clinic measurements and ABPM (C/ABPM). Briefly, untreated hypertensive patients were randomized to use home BP or C/ABPM for antihypertensive treatment initiation and titration. The total cost of the first year of hypertension management was lower in home BP monitoring than C/ABPM arm (€1336.0 versus €1473.5 per patient, respectively), and there was no difference in achieved BP control and drug expenditure. The cost for subsequent years was €348.9 and €440.2 per patient, respectively, for home BP monitoring and C/ABPM arm and €2731.4 versus €3234.3 per patient, respectively (P < 0.001) for a 5-year projection. In view of the cost and limited availability of ABPM, increasing attention is being given to home monitoring with inexpensive semiautomated devices. In particular, this technique has attracted considerable attention in recent years because of the potential for better classification of hypertensive status and hypertension management compared with office BP [23,24]. A relationship between home BP values and the risk of cardiovascular disease has been documented in several observational studies [23,24]. These studies also have shown that the relationship with cardiovascular risk is steeper for home than for clinic BP [23,24]. On the contrary, there is no evidence from intervention studies that home BP is superior to office BP in guiding the treatment of hypertension. Hence, it remains uncertain whether tailoring treatment on the basis of home BP provides a better cardiovascular protection than traditional treatment based on office BP. To further investigate the hypothesis that on-treatment home BP monitoring provides information of prognostic significance, Shimada et al.[25] performed an additional analysis of the Home blood pressure measurement with Olmesartan Naive patients to Establish Standard Target blood pressure (HONEST) study. The report published in the current issue of the Journal [25] includes data from more than 20 000 hypertensive patients followed for about 2 years. Specifically, Shimada et al.[25] investigated the prognostic significance of office SBP and morning home SBP measured at several time points. Three types of BP measurements were evaluated: baseline BP, BP during follow-up (defined as the mean of all available BP measurements), and the achieved BP (defined as the BP measurement at the last time point during the follow-up period or the preceding day of the first cardiovascular event). Briefly, Shimada et al.[25] demonstrated that SBP during follow-up (as compared with SBP at baseline) and morning home SBP (as compared with clinic SBP) had a better prognostic significance in predicting cardiovascular events. When morning home and clinic SBP during follow-up were included in the same survival model, only morning home SBP during follow-up was identified as a significant predictive factor (hazard ratio per 1 mmHg increase: 1.033, 95% CI: 1.020–10.46, P < 0.0001) [25]. Despite the interesting results on the relation between home BP and outcomes in treated hypertensive patients, some methodological aspects of the analysis by Shimada et al.[25] need to be addressed for a proper interpretation of results. First, the percentage of patients who measured home BP according to current guidelines [26] ranged from 82.1% at baseline to 93.9% after 24 months, and the proportion of missing data on clinic BP measured at predefined time points is not reported. As pointed out by Agarwal [27] in a previous commentary on the HONEST study, if the missing data on clinic BP or home BP are not at random, they might selectively affect the association of home or clinic BP with outcomes. Second, when BP is taken at home to establish the diagnosis of hypertension or to assess BP control, the optimal schedule of BP measurements is still undefined. Increasing evidence suggests that at least 12–14 measurements should be obtained, with both morning and evening measurements taken over a period of 1 week. Yet in the HONEST study [25], home BP was only measured in two different days for each measurement point. Another aspect of the analysis by Shimada et al.[25] needs to be mentioned. To investigate the prognostic significance of clinic and home SBP, Shimada et al. developed several risk prediction models to estimate standardized hazard ratios (1 mmHg increase) and the 2-year absolute risk of cardiovascular events for each level of BP variable. To evaluate the model's discrimination ability and the clinical usefulness of added predictors, they also computed the concordance index and the net reclassification improvement (NRI), respectively. They computed NRI values when morning home SBP was added into the model including office SBP, and when clinic SBP was added into the model including morning home SBP. Interpretation of the results led the authors to conclude that follow-up morning home SBP had better reclassification ability [25]. The choice of a risk-category-based NRI seems reasonable in this context because it can demonstrate how many patients would be reallocated into different clinical risk categories by the addition of a specific BP variable. Unfortunately, Shimada et al.[25] did not report the components of the overall NRI to show the net percentages of patients with or without events correctly reclassified [28–30], and they reported overall NRI values well below 0.2, which should be considered weak and scarcely informative on a clinical standpoint. In conclusion, the use of on-treatment home BP, as recorded in the HONEST study [25], is difficult to translate into clinical decisions applied to drug treatment. However, the stage is set for a large, multinational, and intervention study formally testing the superiority of home versus office BP as a target for clinical decisions. According to current evidence, home BP monitoring provides complementary information to office and 24-h ambulatory BP [15]. It should be considered as a complementary, rather than a competitive, method to evaluate out-of-office BP [31,32]. ABPM remains mandatory to obtain information on components of the ambulatory BP profile which proved prognostic significance, including nocturnal BP dipping, ambulatory pulse pressure, night-time BP, and BP variability [16,33–38]. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.