Vera H. Koch, M.D.
Head, Pediatric Nephrology Unit
Instituto da Criança- Hospital das Clínicas da Universidade de São Paulo
Why do we need accurate blood pressure readings? (SLIDE 1)
The first question we have to ask ourselves is whether it is really necessary to be accurate for blood pressure measurement in children. The answer is yes, because:
the level of blood pressure in childhood is the strongest predictor of adult blood pressure level (Lauer & Clarke,1989; Berenson et al, 1991; Gillman et al, 1993), hypertension increases two to three times an individual’s risk of cardiovascular morbidity (Kannel WB, 1996; Klag et al, 1996), the relationship between hypertension and cardiovascular disease seems to be continuous and cardiovascular risk depends on blood pressure itself, coexistent risk factors and whether there is hypertensive end-organ damage .
Which conditions should be standardized for blood pressure measurement? (SLIDE 2)
If accuracy is obligatory, a standardized protocol should be considered for blood pressure measurement, which would make the comparison of results obtained by different studies in different countries possible. Observers should be trained and certified to minimize measurement biases, a homogeneous decision regarding equipment factors such as an appropriate cuff bladder size or the alternative use of mercury manometers or oscillometric devices should be taken, technique factors such as recording forth, fifth or both Korotkoff sounds as for diastolic blood pressure need to be taken into consideration as well the number of necessary measurements to estimate a child’s blood pressure and the influence of environmental factors such as the time of the day and ambient temperature on the measured value of a child’s blood pressure (Gillman & Cook, 1995)
. Some of these factors will be discussed separately in the next paragraphs.
Classically to obtain an accurate blood pressure measurement, a cuff bladder width of approximately 40% of the upper arm circumference should be chosen because it most closely approximates intra-arterial readings (Perloff et al, 1993). The bladder length should be at least 90% of arm circumference to avoid overestimation of BP especially in children (Vyse TJ, 1987). Another less known effect of the cuff size (Slide 4) occurs when according to the above mentioned instructions for cuff selection, the cuff size should be changed to a larger one, the change of cuff leads to an abrupt fall in the value of measured blood pressure, that is not arm dependent, but cuff dependent (Whincup et al, 1989)
This very inconvenient effect may be responsible for two issues: 1. any association between blood pressure and arm circumference, such as body mass, will be biased towards the null 2. in longitudinal studies as one changes the cuff to a larger one, measured blood pressure is lower than prior readings, which could lead to inadequate inverse correlations of blood pressure to chronological age or height. Arafat and Matoo, 1999 reviewed commercially available blood pressure cuffs and detected that the size pattern of available cuffs, labeled as infant, pediatric, small adult, adult and large adult is heterogeneous among the different manufacturers (Slide 5), the authors conclude that cuff sizes need to be standardized, indicate bladder size and suggest that they should be color-coded for convenience
Number of necessary measurements to estimate a child’s blood pressure (SLIDE 6)
Another important issue to consider is the number of measurements that should be repeated within a visit and between visits in order to determine a child’s blood pressure. The work of Gillman and Goodman 1995, demonstrates that it depends on the instrument and on the technique. For auscultatory equipment, using the mercury manometer or the random zero manometer, among 162 children 8 to 12 years of age, the systolic blood pressure values obtained after four weekly visits, three measurements per visit, leveled off after about 2-3 measurements per visit, but the difference between visits was large until about three or four visits. For oscillometric equipment, using the Dinamap model 845XT, among 106 children, 9 to 13 years of age, the systolic blood pressure values obtained after three weekly visits, four measurements per visit, demonstrated that for the Dinamap device the first of several measurements at a particular visit was generally higher than the following ones, the obtained values started to level off after 4-5 measurements within a visit, the "first measurement effect" repeated itself even after 3 consecutive visits .
The diastolic dilemma (SLIDE 7)
There has been an ongoing controversy over whether the muffling (Korotkoff 4- K4) or the disappearance of sounds (Korotkoff 5- K5) should be preferentially considered for the measurement of diastolic blood pressure in children. As neither value correctly defines intra-arterial diastolic blood pressure ( K5 is approximately 9mmHg higher than direct DBP) and K5 is easier for the human ear to discern than K4, current recommendations favors the use of K5.
The stetoscope diaphragm vs the bell (SLIDE 7)
The bell is preferred for blood pressure auscultation in adults. This issue is still controversial in children, as placing the bell adequately in small children may compress the artery and produce falsely low diastolic values. Thus some authors advocate the use of the diaphragm for small children (Londe S, 1987), while others suggest that the bell properly used should accomplish better auscultatory results (Prineas& Jacobs, 1983)
Time of the day and environmental temperature (SLIDE 7)
It is clear from ambulatory blood pressure studies that blood pressure varies along the 24 hours of the day, presenting lower values during sleep and higher values during vigil, with a peak in the morning and another one in late afternoon (Krull et al, 1993).There is a negative relationship between BP and temperature an increase in 10 ° C leads to a fall of approximately 5-7 mmHg in SBP and DBP (Prineas et al, 1980; Jenner DA, 1987)
SLIDE 9 shows the lack of homogeneous methodology of nine studies, which composed the Second Task Force of Blood Pressure Measurement in Children and were reviewed by Rosner et al in 1993. The Update of the Second Task Force of Blood Pressure Measurement in Children added a tenth study to these list (NHANHES III).
SLIDE 10 shows the same lack of methodology homogeneity in the six studies from which the European pediatric blood pressure normative data is presently derived (de Man et al, 1991).
It is important to emphasize that this lack of homogeneity is not a consequence of carelessness but rather of the multiple difficulties involved in performing epidemiological studies in the pediatric age group. Unfortunately, according to Nielsen et al, 1989 "Confusion concerning the most suitable cuff… is responsible for at least some of the scatter between blood pressure studies" and Arafat & Matoo, 1999, referring to the Update of the Second Task Force of Blood Pressure Measurement in Children, suggest that "A new multicenter study, using uniform criteria for cuff selection, may be necessary to establish the accuracy of the published nomogram on normal BP in children."
What blood pressure measuring device should be used in the future?
The mercury manometer is our old friend, it is simple, accurate and easy to service; standard Hg readings are the main basis for blood pressure-disease associations and although blood pressure readings with this instrument are subject to digit preference and observer bias, this issue is eliminated by observer training. Unfortunately, mercury has toxic effects on the environment and the mercury manometer will have to be gradually replaced.
The aneroid sphygmomanometer registers BP through a mechanically intricate system. Its accuracy is affected by everyday use. When calibrated against a mercury manometer a mean difference of 3mmHg is acceptable, although up to 30% have errors > 7 mmHg. Readings are also subject to digit preference and observer bias.
What about automated sphygmomanometry? The most widely used oscillometric devices are manufactured under the name "Dinamap". Several models have been developed, each with an updated algorithm. Validation data has to be obtained separately for each model. Systolic and diastolic blood pressures are calculated as a function of the mean arterial pressure, which is the point of maximal oscillation and are calibrated to be equivalent to intra-aortic pressures. The devices are easy to use and strongly correlated to intra-arterial readings (SLIDE 11). Accuracy is affected by motion of the arm and measurements are affected by a "first –reading effect", as mentioned above.
Automated oscillometry devices have to be validated before they can be recommended for clinical use. Validation protocols, based on comparative measurements between oscillometric equipment and the mercury manometer, were devised by the British Hypertension Society (BHS) and by the American Association of Medical Instruments (AAMI). Presently both protocols have been reconciled and are used in association to validate oscillometric devices. SLIDE 12 depicts instruments presently validated and recommended for hospital use. SLIDE 13 shows instruments presently validated and recommended for self-measurement (home blood pressure).
Is it possible to use auscultatory and oscillometric devices interchangeably? Unfortunately not, as Korotkoff
1 is approximately 3mmHg lower than direct systolic blood pressure and, as we mentioned earlier, K5 is approximately 9mmHg higher than direct DBP. Park et al, 2001, testing the Dinamap 8100 against the standard mercury manometer found that the Dinamap detected mean systolic and diastolic blood pressure values significantly above auscultatory readings while Barker et al, 2000 testing the Omrom M1 against the standard mercury manometer concluded that the Omron M1 overestimates higher pressures and underestimates lower pressures. There is a lack of validated and approved automated devices for use in clinical and epidemiological setting for the pediatric age group (Wells et al, 1998)
Ambulatory Blood Pressure Monitoring (ABPM) in Children (SLIDE 14)
The current general indications for ABPM are: identification of white coat hypertension,borderline hypertension, identification of nocturnal hypertension, drug resistant hypertension, guide to anti-hypertensive medication, hypertension of pregnancy and identification of hypotension. The slide shows some of the current issues for ABPM use in Pediatrics, the main problem is the lack of definite normative data. It is a promising methodology as its recordings shows good accuracy and reproducibility in children (Lurbe et al, 1999). Actual definitions of sleep/wake periods, using actigraphy or a detailed diary of daily activities, are necessary for accurate sleep BP decline determination (Eissa et al, 2001). White coat effect (white coat hypertension or white coat normotension), known to the adult literature was also confirmed in the pediatric population and similarly to data in adults, left ventricular mass index and left ventricular hypertrophy are more closely related in children to 24-hr systolic blood pressure than with casual systolic blood pressure (Belsha et al, 1998). According to Kapuku et al, 1999-left ventricular hypertrophy can be predicted by initial ambulatory systolic parameters. SLIDE 15 shows some of the studies performed in high-risk pediatric populations.
SLIDE 16and 17 show a study of our group comparing in normotensive (NT) and hypertensive (HT) adolescents, casual blood pressure and ABPM parameters. Casual blood pressure was measured in two different and separate environments (clinic and ABPM unit), by two trained observers and shows, for systolic and diastolic blood pressure, in both normotensive and hypertensive populations, an alarm reaction when exposed to an unknown environment and observer (ABPM unit). It should also be noted that, contrary to findings in the adult population, mean casual systolic/diastolic BP, measured in clinic, are lower than mean ABPM parameters while awake, in normotensive and hypertensive adolescents. SLIDE 18 shows, for the same study, a comparison between findings on casual auscultatory measurements (clinic and ABPM unit), and ABPM parameters in hypertensive adolescents, such as systolic and diastolic ABPM means (S/D ABPM means ), systolic and diastolic blood pressure descent during sleep (S/D sleep BP descent), systolic and diastolic blood pressure load (S/D BP Load). It emphasizes that although parameters of normality are still under construction for ABPM in the pediatric age range, ABPM is a promising tool for the follow-up of pediatric hypertensive patients, in which instance it seems superior to casual blood pressure evaluation as it uncovers the white coat effect.
SLIDE 19 presents the updated list of commercially available tested and recommended ABPM devices as of 2001 and SLIDE 20 presents ABPM device validation data for children, in which the Spacelabs 90207, widely used in pediatric studies and the TM 2421, used in a recent large pediatric study (O’Sullivan 1999) have not scored well enough to be recommended. As of 2001, the only device recommended for children, according to BHS and AAMI protocol is the QuietTraK, is amazingly an auscultatory equipment, this modality of device is generally not adopted in pediatric studies because the noise of children in movement interferes with the accuracy of the microphone determination of the measured blood pressure value. SLIDE 21 shows a list of some large pediatric ABPM studies and demonstrates that as for casual BP, studies are being executed without methodological homogeneity. Different devices, different measurement protocols cannot be considered together to generate norms.
In conclusion as of 2001, the main problem for the diagnosis and management of hypertension in children is the lack of sound normative data for casual and ambulatory blood pressure values. The only solution for this issue is to propose a multicentric study with a homogenous protocol to obtain normal multi-ethnic casual and ambulatory pediatric blood pressure values, only then studies to correlate blood pressure level and hypertensive end - organ damage will be possible