Office blood pressure measurement types: Different methodology—Different clinical conclusions

Abstract

The measurement of blood pressure in the office (OBP) remains the basis for hypertension diagnosis and management for more than half a century. Despite the increasing use of out‐of‐office blood pressure measurement using home and less so ambulatory monitoring and their endorsement by hypertension societies, at present and for some time to come it is likely that in many people the diagnosis and management of hypertension will be based on OBP measurement alone. OBP measurement is a very variable method affected by multiple factors, which have major impact on the OBP level, reproducibility and prognostic ability. Thus, there are several types of OBP measurement, depending on the device type, conditions, observer’s presence and the number of readings. The 4 main OBP types are: Type I: Auscultatory OBP in clinical practice; Type II: Automated attended OBP in clinical practice; Type III: Research setting OBP; Type IV: Unattended automated OBP. These OBP types have different standardization level, different reproducibility, different clinical relevance and different thresholds for hypertension diagnosis. Unless the methodological details of OBP measurement are reported, any conclusions based on such measurements in clinical research and in practice are questionable.

The measurement of blood pressure (BP) is the most common procedure performed in the doctor's office. Moreover, the diagnosis of hypertension and the decisions for life‐time treatment are exclusively dependent on BP measurement. In 1964, Geoffrey Rose presented the categories of observer error in auscultatory BP measurement, including systematic error, terminal digit preference, and observer prejudice.1 Recent guidelines in the USA and Europe recognized that office BP (OBP) measurement alone is often unable to diagnose hypertension accurately, mainly due to the white coat and masked hypertension phenomena.2, 3

The US guidelines provide clear recommendations for using out‐of‐office BP monitoring (ambulatory [ABPM] or home [HBPM]) to confirm the need to initiate treatment in untreated or titrate in treated subjects.2 The European guidelines provide very similar recommendations for out‐of‐office BP monitoring, yet they give two options for diagnosing hypertension: based on repeated OBP measurements using an auscultatory or electronic (oscillometric) device on several visits, or on ABPM or HBPM “provided that these measurements are logistically and economically feasible.”3

The recommendation to base decisions on OBP alone is scientifically problematic because it has been shown that, even with carefully measured OBP in repeated visits in a research setting, about 30% of untreated or treated subjects are misdiagnosed due to the white coat and masked hypertension phenomena.4 On the other hand, this recommendation presents a realistic approach given the limited use of HBPM and much less ABPM, even in Europe and North America. A recent study in 2221 primary care physicians in Spain showed that only 3% of them recommended ABPM always and 27% usually, and for HBPM 17% and 50%, respectively.5 Thus, the reality is that at present and for some time to come it is likely that in many people the diagnosis and management of hypertension will be based on OBP measurement alone. Hence, both the US and European guidelines provide detailed guidance for doctors to obtain standardized OBP measurements.

There are two major methodological issues with OBP: (a) after half a century of wide application of the OBP measurement, the scientific community has failed to eliminate the observer related errors which remain very common in general practice, and (b) it is recognized that, even with standardized OBP measurement devoid of the observer errors (unattended automated OBP), the white coat and masked hypertension phenomena are still common and lead to misdiagnosis in a considerable proportion of untreated and treated subjects.6

In this issue of the Journal of Clinical Hypertension, Tang et al7 compared OBP measurements taken in a primary care clinic using a validated automated device and a standard protocol versus OBP taken in a research setting using the same automated device. The study showed that, despite the use of automated devices which eliminate the observer errors, primary care OBP was higher and more unstable than research setting OBP.7 More importantly, these differences were evident despite the implementation of an intervention program aiming to standardize OBP.7

These findings have important implications for clinical practice. It must be realized that OBP measurement is a very variable method affected by multiple factors, which have major impact on the OBP level, reproducibility, and prognostic value. Thus, there are several types of OBP measurement, depending on the device type, conditions, observer’s presence, and the number of readings, which have different standardization level, different reproducibility, different clinical relevance, and different thresholds for hypertension diagnosis. The four main OBP types and their characteristics and differences are summarized in Table 1 and are the following:

1. Auscultatory OBP in clinical practice: This is a poorly standardized and highly variable method which considerably overestimates BP resulting in over‐diagnosis.8 Moreover, the auscultatory devices might induce a systematic error with time which is not evident without calibration.

2. Automated attended OBP in clinical practice: This method has the advantage of avoiding the observer errors and that automated devices are more likely to remain accurate for long time (or stop working). In the last 20 years, this method has been the most widely used in hypertension outcome trials, most of which obtained 2‐3 readings with automated devices.9 The Systolic Blood Pressure Intervention Trial (SPRINT) and other studies showed that when such measurements are taken in standardized conditions (few minutes resting, no talking, 3 or more measurements) they give similar OBP values as “unattended automated OBP” (see below type IV).10, 11 However, in clinical practice issues with lack of resting period, inadequate body and arm position, and talking during or between the measurements are possible and might increase OBP.

3. Research setting OBP: This method taken with auscultatory or automated devices in the context of relatively small clinical trials performed in a single or few hypertension expert centers is “ideal” OBP as it is perfectly standardized. Such studies have investigated the OBP reproducibility, diagnostic value, drug effects, etc. Research setting OBP appears to have reproducibility12 and association with indices of preclinical organ damage close to that of ABPM.13 However, it is unrealistic to achieve such measurements in clinical practice. It is important to note that the research OBP of outcome hypertension trials 9 has not been so carefully standardized, as it was performed in multiple centers without high expertise in hypertension research. Giorgini et al reviewed the OBP methodology in 64 mega‐trials in hypertension published from 1990 to 2014 and concluded that numerous aspects of OBP measurement often deviated from guideline recommendations and varied considerably across trials.9 They stated that “the lack of uniform methodologies in outcome studies that form the foundation of evidence‐based guidelines may have significant clinical implications.”9

4. Unattended automated OBP 6, 8: This is a unique method to achieve standardized OBP in primary care because (a) observer errors are avoided, (b) talking of the patient during and between measurements is avoided, and (c) it ensures that a standard protocol (eg, triplicate measurement) is followed.10 Disadvantages are that (a) it may not be applicable in all primary care settings as it requires a special device, more time, and office space, and (b) the threshold for hypertension diagnosis is lower and rather uncertain with scarce outcome data.10, 11, 14

Table 1.

Types of office blood pressure measurement

Sources of error	I. Auscultatory clinical practice	II. Automated attended clinical practice	III. Auscultatory or automated research setting	IV. Automated unattended clinical practice
Standardization	−	+	+++	+++
Resting period	Inconsistent	Inconsistent	Yes	Yes
Devices maintenance and accuracy	Variable	Usually ok	Good	Usually ok
Cuff sizes	1‐2	1‐2	3+	2+
Talking during resting and between readings	Common	Common	No	No
Observer error, bias, terminal digit preference	Yes	No	No	No
Number of readings	1‐2	1‐3	2‐3	3‐6
Hypertension threshold (mmHg)	Significantly higher than 140/90	Equal or higher than 140/90	Equal or lower than 140/90	Equal or lower than 135/85
Time required	1′‐5′	1′‐5′	~10′	~10′
Applicability for general practice	Yes	Yes	Not possible	Not for all practices (requires device, office space, time)
Outcome data	No	Strong	Strong	Few

CONCLUSIONS

1. Unless the methodological details of OBP measurement are reported, any conclusions based on such measurements in clinical research and in practice are questionable.

2. By considering the (a) the OBP methodology used in outcome trials and (b) the feasibility for unbiased measurements in primary care, method II with triplicate attended automated OBP measurements appears to be the most appropriate for wide implementation. Educational initiatives are necessary to ensure that the remaining requirements (resting period, body position, cuff selection, no talking, triplicate measurements) are fulfilled. Low‐cost automated devices can be developed to facilitate the standardization of OBP measurement. Unattended automated OBP measurement can be used in those medical settings that can implement it in their routine evaluation of people with suspected or treated hypertension.

3. Even the most standardized OBP measurement (methods III or IV) is misleading in a considerable proportion of subjects. The recent US and European guidelines, recommending that most subjects with suspected hypertension and particularly those with OBP close to the intervention threshold (lower or higher) should be evaluated with ABPM or HBPM, must be followed as much as possible.2, 3

CONFLICT OF INTEREST

GS conducted validation studies for various manufacturers and advised manufacturers on device development. KGK and AK have nothing to declare.