6. Will alternative devices affect health care?
The SCENIHR opinion states:
The mercury sphygmomanometer is gradually disappearing from clinical use. Mercury- free blood pressure measuring devices (when clinically validated) are generally reliable substitutes for mercury-containing sphygmomanometers in routine clinical practice. These alternative devices include both auscultatory devices requiring a trained observer and automated oscillometric devices for which some instruction is needed. Clinically validated, auscultatory mercury-free devices are equivalent to mercury sphygmomanometers, and are thus suitable for specific groups of patients, including patients with arrhythmias, diabetes, pre-eclampsia and the elderly. The alternative devices using auscultation have similar limitations as the mercury sphygmomanometers regarding the observer technique and bias associated with auscultation itself. These may be avoided by using automated oscillometric devices, which, when properly validated, allow accurate blood pressure measurements. The oscillometric technique has mainly been clinically validated in adult populations including a wide range of blood pressures but not in a wide range of ages and clinical conditions, and should not be used in some specific clinical conditions including pre-eclampsia. There is no evidence of adverse effects on patients' health in clinical settings due to the replacement of mercury- containing sphygmomanometers by validated mercury-free alternatives. There are adequate alternatives in most clinical conditions/settings. In special conditions, such as pre-eclampsia, mercury-free auscultatory devices should be preferred until further validation of oscillometric devices. Very comprehensive research on population blood pressure exists throughout the world. These studies are essential for defining hypertension prevalence, awareness and treatment in any geographical region/country. A change in population blood pressure of 2 mmHg in systolic blood pressure translates to a change in stroke mortality of ten percent and coronary heart disease mortality of seven percent (Lewington et al. 2002). Therefore, data on progression from normotension to prehypertension and hypertension are very important in epidemiological research. The data have documented that prehypertension carries an increased risk for cardiovascular morbidity and mortality, and a high risk for progression to sustained hypertension (Hansen et al. 2007a, Julius et al. 2006). In this respect, changes from normotension to prehypertension are as important as the observation of hypertension itself. Reliable data are heavily dependent on blood pressure measurements carried out meticulously by properly trained personnel and with precise equipment. For this, adherence to a standardised technique over time is crucial. Findings of changes in population blood pressure are only meaningful if they are ascertained to be true differences and not related to a change in methods applied.
Nearly all results on population blood pressure have been obtained by the use of a standard mercury sphygmomanometer by well-trained health personnel (Cutler et al. 2008). Despite this, the readings are not without observer bias and end-digit preference. In an attempt to minimise observer bias and end-digit preference, a number of highly recognized epidemiological research institutions have used the Random Zero Mercury Sphygmomanometers, where the reader has to subtract a random chosen magnitude of mmHg (from 0 to 20 mmHg) at the very end of the measurement. Despite minimising observer bias, the equipment has been shown to slightly underestimate the "true” blood pressure level as obtained by the use of a standard mercury manometer (Yang et al. 2008). Another approach that has been employed is the "London School of Hygiene Sphygmomanometers” (Andersen and Jensen 2007) where the reader is blinded to the mercury column but has to tap a button when they hear the first and the fourth Korotkov sounds (phase 1 and phase 5).
In recent years, 24-hour ambulatory blood pressure measurements have been introduced in population studies and comprehensive databases have been constructed, e.g. the Idaco Database on population studies with contributions from many parts of the world (Hansen et al. 2007b). All these studies have convincingly shown that 24-hour ambulatory blood pressure measurements determined with oscillometric devices (at approximately 80 readings over 24 hours), are superior for prediction of cardiovascular morbidity and mortality as compared to a few measurements of blood pressure performed in clinical conditions with a standard mercury sphygmomanometer. In almost all these studies, although not exclusively, the comparator has been the standard mercury sphygmomanometer (Hansen et al. 2007b).
Research into normal values for home blood pressure and the prognostic implication is less comprehensive. This research has been almost exclusively carried out with automatic oscillometric devices, with measurements being compared to the mercury sphygmomanometer. Data are accumulating showing that the predictive prognostic value of a certain number of home blood pressure readings is superior to a single or a few blood pressure readings performed in a clinic using a mercury sphygmomanometer (Sega et al. 2005). The home readings are a reflection of more precise estimation of the actual blood pressure levels over many readings as compared to few readings in the clinical setting. So far, comparisons of measurements obtained with mercury sphygmomanometer versus oscillometric automatic devices, obtained in the same clinical setting for determination of population blood pressure and prognostic implications, are missing. However, in the Pamela Study, three clinic readings with a mercury sphygmomanometer were compared to two home blood pressure oscillometric readings (Sega et al. 2005). As expected, the clinical readings were somewhat higher, but the prognostic implication was not that much different.
In long-term outcome clinical trials, usually running for three to five years, mercury sphygmomanometers have been used as the gold standard for office blood pressure measurement. In some recent trials (the HOT Study, the ASCOT Study and the OnTarget Study) automatic oscillometric devices were used (Dahlöf et al. 2005, Hansson et al. 1998, Yusuf et al. 2008). In some of these studies it was shown that small differences in measured blood pressure already can have an impact on cardiovascular diseases. There is rapidly growing information on normal values and the prognostic implications of 24 hour ambulatory blood pressure measurements with oscillometric devices, while knowledge on self/home blood pressure measurements with oscillometric devices is less substantial. So far, a direct comparison between clinic blood pressure and prognostic implication based on measurements carried out with mercury sphygmomanometer and those with automatic oscillometric devices is lacking.
In conclusion, the vast majority of information on population blood pressure (secular trends, progression to hypertension and prognostic implications, and also the benefits from treatment-induced blood pressure reduction in terms of cardiovascular events prevention) has so far been obtained with the use of mercury sphygmomanometers. Reliable data on changes in population blood pressure level, incidence and prevalence of hypertension, awareness and treatment, derived from follow-up studies are dependent on the use of consistent and trustworthy methods. It can be expected that epidemiological/observational studies in the future will comprise repetitive blood pressure measurements at home carried out with well-calibrated, well-validated automatic oscillometric equipment. For the moment, mercury sphygmomanometers are essential for such validation of newly developed blood pressure measurement devices. Otherwise, the conclusions based on the results of long–term epidemiological studies on changes in population blood pressure may be seriously jeopardised.