How to Cope With Unreliable Office Blood Pressure Measurement?

The diagnosis of hypertension is traditionally based on conventional measurement of blood pressure (BP) in the office taken using a standard mercury sphygmomanometer and the auscultatory technique.1,2 Unfortunately, despite wide application of the technique for longer than a century, important problems still exist regarding the reliability of the diagnosis of hypertension that is based on conventional office BP measurements.1,2 This deficiency of office BP measurement is due to inherent limitations of the method but also because of its poor standardization and application.1,2 In this issue of the American Journal of Hypertension Campbell et al3 assessed the reliability of the classification of hypertension based on usual office BP measurements obtained in family practice, in comparison with carefully standardized office measurements and ambulatory BP monitoring. The study showed usual office BP to be significantly higher than standardized office BP or daytime ambulatory BP. The authors concluded that usual office BP measurement often leads to significant overestimation of BP and thereby overdiagnosis of hypertension.3 The findings of this study have probably been exaggerated due to an indication bias, given that only subjects referred for ambulatory BP monitoring were included.3 Because the major indication for ambulatory BP monitoring is suspected white coat phenomenon, it seems that study participants were those subjects whose physicians believed office BP was unrepresentative (eg, elevated office but low home BP). An “order” effect may also be present in this study, because participants were not randomized to the order of usual, standardized and ambulatory BP measurement. Nevertheless, this study is important because it brings into discussion the critical and unresolved issue of the poor methodology used for office BP measurement. Such data raise the question as to whether, after a century of use, the failure of the medical profession to master the Riva-Rocci/Korotkoff technique can be overcome. Measurement of BP aims to estimate the average level of an unstable variable, in order to make decisions about life-saving but at the same time potentially harmful lifetime drug treatment. It is striking how little attention is often given to the assessment of the “true” BP level and the decision to prescribe long-term antihypertensive drug treatment. The issue of the unreliable BP measurement by physicians was highlighted in a caustic manner in a review by Thomas Pickering4 in 1994, when he stated that “the measurement of blood pressure is much too serious to be left to physicians.” It seems that practicing physicians of nowadays are being weighed down with information overflow, technological advances, and many other commitments. It might therefore be argued that the average physician is unwilling to spend the time needed to perform thoroughly a simple conventional procedure, which is being widely used by medical assistants and even patients. An additional factor contributing to unreliable office BP measurement is the infrastructure of care delivery that, even in well-developed healthcare systems, allows insufficient time to be devoted to an office visit that has to include history taking, review of laboratory tests, physical examination and BP measurement, advice and instructions, and treatment prescription. Further to the limitations of the usual office BP measurement attributed to poor methodology and human error, important problems still exist. It is recognized that, even using carefully standardized office measurements, in more than 30% of subjects attending a hypertension clinic the white coat and the masked phenomena may lead to significant overestimation or underestimation of BP and thereby overtreatment or undertreatment.1,2 In 1991, James Conway clearly suggested that, in fact, the major drawbacks of office BP measurement cannot be overcome by saying: “Clinic blood pressure is poorly reproducible and unable to diagnose hypertension. It is in error approximately 50% of the time. Therefore, the clinic blood pressure is not worth taking and another kind of measurement has to be found.”5 In line with this view, Campbell et al3 suggested that the development of alternative methods for assessing BP should be considered. Alternative methods for hypertension diagnosis to be considered include either out-of-office BP measurement or alternative protocols for office BP measurements. Out-of-office BP measurement, either by the patients at home or by using ambulatory monitoring, is advantageous because it provides multiple BP measurements in the usual environment of each individual. There is considerable evidence suggesting that, compared to office BP, both home and ambulatory measurements are more reproducible,6 eliminate the white coat effect,7 and are more closely related to target organ damage and cardiovascular risk.1,2 Therefore, these methods have been endorsed by recent hypertension guidelines as supplementary sources of information to the practicing physician.1,2 Alternative protocols for office BP measurements might also be considered. Campbell et al3 showed that the conventional methodology is improved when carefully standardized measurements are taken by trained research nurses. The lower BPs obtained by the trained nurse should be attributed to the careful standardization of the method rather that the fact that measurements were taken by a nurse, because in family practices terminal digit preference and hypertension overdiagnosis were equally prevalent with physician- or nurse-taken office BPs.3 Myers et al8 showed higher prevalence of the white coat effect with routine office measurements compared to special study BPs taken by the same physicians, whereas BPs taken by research nurses induced the least white coat effect. More importantly, routine office measurements in that study provided the only BP measurements not correlated with left ventricular mass.8 Two important questions regarding alternative office measurement methods should be addressed. The first is how feasible the wide application of a novel methodology is in family practice and the maintenance of its performance at a high standard. This is clearly dependent on the acceptance of the new method by physicians but also its implementation in the infrastructure of care delivery. The second question, and a more tricky one, is whether BP values obtained by an alternative method for office measurement should be assessed on the basis of the established diagnostic thresholds for conventional office BP. It might well be that careful standardization of office BP measurement (relaxed environment, long period of relaxation before measurements, multiple measurements) provides “unusually” low BP values that underestimate BP when assessed on the basis of conventional diagnostic thresholds. This seems to be the case with the study of Campbell et al,3 where systolic office BP taken by the trained nurse was 3.1 mm Hg lower than awake ambulatory BP. Because the diagnostic threshold for office BP (140/90 mm Hg) is by 5 mm Hg higher than for awake ambulatory BP (135/85 mm Hg),1,2 it is not surprising that in this study ambulatory BP confirmed the presence of hypertension in 82% of study participants compared to only 45% with trained research nurse measurements.3 Given the superiority of ambulatory BP in terms of prognostic significance,1,2 the conclusion should be that, in fact, it was the trained nurse measurements that underestimated BP and misdiagnosed hypertension. In other words, office BP taken by the trained nurse eliminated the white coat effect but at the same time exaggerated the masked effect. This is probably because a 140/90 mm Hg diagnostic threshold might be too high for BP values obtained by this alternative office BP measurement method. Interestingly, the study by Campbell et al3 also investigated a novel approach for office BP measurement using a validated professional automatic BP monitor (BPTru) and these data have been recently presented in another article.9 The device was set to take six office BP measurements at 5-min intervals while the patient remained alone in a room. This approach has several advantages because it eliminates the observer bias and potentially the white coat effect. Although this method succeeded in eliminating the white coat phenomenon in this study, it clearly underestimated hypertension because average BP was even lower than that measured by the research nurse by 6.5/4 mm Hg.9 Quite rightly these invetigators discouraged the use of this approach in clinical practice.9 Back in 1943, Alan and Smirk10 conducted interesting experiments that provided important insights into the behavior of office BP. They showed office BP to be significantly reduced when multiple measurements are taken during 30 min in relaxed conditions. Such office measurements, which they called “basal BP” can lead to significant underestimation of BP and thereby underdiagnosis of hypertension if assessed by using the conventional 140/90 mm Hg threshold. These data suggest that unusual standardization of BP measurement and overzealous effort to eliminate the alarm effect induced by the clinic setting may lead into unusually low BP values that do not reflect conventional office BP. In conclusion, continuous effort should be made to achieve and maintain optimal standardization of office BP measurement. Supplementation of office measurements with home and ambulatory monitoring is often needed for an accurate diagnosis. Alternative protocols for office BP measurement might overcome some of the limitations of office BP and should therefore be investigated. A note of caution is, however, needed regarding alternative methods, to ensure that established diagnostic thresholds for conventional office measurements also apply to BP values obtained by the new methods.