Journal of Human Hypertension (1999) 13, 695–700 1999 Stockton Press. All rights reserved 0950-9240/99 $15.00 http://www.stockton-press.co.uk/jhh ORIGINAL ARTICLE Differences in blood pressure values by simultaneous auscultation of Korotkoff sounds inside the cuff and in the antecubital fossa F Weber1, M Anlauf 2, H Hirche3, U Roggenbuck3 and Th Philipp1 1Department for Kidney Disease and Hypertension, Center for Internal Medicine and 3Institute of Medical Informatics, Biometry and Epidemiology, University Essen; 2Medical Clinic II, Zentralkrankenhaus Reinkenheide, Bremerhaven, Germany To elucidate whether auscultation of the Korotkoff B with differences in systolic/diastolic BP of sounds inside the cuff and in the antecubital fossa leads ؉0.8 ؎ 1.0/؊8.5 ؎ 2.2 mm Hg in the same-arm test and .to different blood pressure (BP) values we measured BP ؉0.4 ؎ 4.8/؊10.6 ؎ 5.2 mm Hg in the opposite-arm test at both sites simultaneously with identical flat stetho- Subject’s age was the main variable determining differ- scopes in a same-arm test design (part A) in 64 normo- ences in diastolic BP with significantly higher differ- tensive (N: 32 men, 32 women; mean age: 38.7 ؎ 15.1 ences in younger than in older subjects, indicating that years) and 67 hypertensive subjects (H: 36 men, 31 the elastic properties of arteries may be responsible for -women; mean age: 44.6 ؎ 12.9 years), and additionally these differences. Our results demonstrate that a modi in a same- and opposite-arm test design (part B) in 20 fication in the auscultatory technique of BP measure- normotensive young women. While in part A systolic BP ment produces significantly different diastolic BP measured inside the cuff was only slightly higher (N: values, the magnitude of which is important for our con- ؉1.6 ؎ 3.2 mm Hg; H: ؉1.0 ؎ 1.4 mm Hg), diastolic BP ceptions of threshold and target values in diagnosing .was considerably lower (N: ؊10.6 ؎ 5.6 mm Hg; H: and treating hypertension ؊8.4 ؎ 4.9 mm Hg). This result was corroborated by part Keywords: blood pressure measurement; Korotkoff sounds; inside cuff; antecubital fossa Introduction indirect measurement of BP6 nor recent reviews7,8 consider this change in methodology. Our experi- The common indirect method of measuring blood ences in testing such devices9–11 led us to the sus- pressure (BP) is based on the occluding-cuff auscul- picion of systematic influences of the microphone tatory technique introduced by Riva-Rocci1 and Kor- 2 position on the results of BP measurement. otkoff. According to the recommendations of the This study was designed to answer the follow- American Heart Association, a cuff of proper size ing questions: should be wrapped smoothly and snugly around the arm, with the lower margin 2.5 cm above the ante- (1) Does the auscultatory detection of Korotkoff cubital space, and the stethoscope should be pos- sounds inside the cuff result in BP data system- itioned over the palpated brachial artery below the atically different from those measured in the cuff at the antecubital fossa.3 All recommendations fossa cubiti? for BP measurements emphasised that the stetho- (2) If this holds true, are the differences of clinical scope should not be in contact with the cuff.4,5 In relevance? contrast to these recommendations in new devices (3) Do subject characteristics like age, gender, upper using the auscultatory method the microphone is arm circumference or height of BP have an embedded in the lower edge of the cuff. As a conse- impact on the extent of difference, if any? quence of this modification of the auscultatory tech- nique the Korotkoff sounds are not picked up in the fossa cubiti but inside the cuff’s width. Neither the Subjects and methods American Heart Association’s recommendation for Part A Sixty-four normotensive and 67 hypertensive sub- jects (Table 1) were randomly selected from our out- Correspondence: Priv.-Doz. Dr. Franz Weber, Thiemannstr. 18, 45219 Essen, Germany patient clinic. After taking the upper arm circumfer- Received 21 March 1999; revised 24 May 1999; accepted 1 June ence the BP was measured by two observers using 1999 the auscultatory technique simultaneously in the Differences in BP value using Korotkoff sounds F Weber et al 696 Table 1 Age and sex distribution of the subjects in the study. Blood pressure (SBP and DBP = systolic and diastolic blood pressure) was calculated as the mean of the measurements of both observers. 45 hypertensive subjects took antihypertensive medication Normotensives Hypertensives male female male female n 32 32 36 31 Age Ͻ35 years 14 14 9 7 35–55 years 13 12 18 16 Ͼ55 years 5 6 9 8 mean ± s.d. 38.4 ± 16.2 38.9 ± 14.1 43.6 ± 12.3 45.7 ± 13.6 age-range (years) 15–73 19–67 14–62 16–67 Systolic BP (mean ± s.d.) 122.6 ± 13.8 115.6 ± 14.4 149.8 ± 20.1 153.2 ± 17.5 Diastolic BP (mean ± s.d.) 68.8 ± 15.8 68.8 ± 14.2 89.5 ± 13.8 88.0 ± 13.9 same arm as proposed by Atkins et al.12 Both observers were checked for auditory or visual impairment. In each subject two specifically trained observers (O1 and O2) performed 10 consecutive BP measure- ments in the same arm with identical flat (thickness: 8 mm) stethoscopes (‘Lightweight’). While O1 heard the Korotkoff sound over the cubital artery (a.c.) at the antecubital fossa, O2 simultaneously measured over the brachial artery (a.b.) with the stethoscope head inserted into a small pouch at the lower edge of the cuff. For all measurements the standard cuff size was used. After five measurements the observers changed the stethoscopes and sites of aus- cultation. The sequence of tasks of both observers Figure 1 Difference (mean ± s.d.) of BP (mm Hg) in same-arm and randomly changed from subject to subject. Systolic opposite-arm tests performed in 20 young female subjects. S = pressure was noted at the onset of at least two con- systolic, D = diastolic blood pressure. The arrow denotes the secutive beats, the diastolic pressure as well at muf- change of the location of auscultation site and of observers. fling (phase IV of Korotkoff sounds, only in hypertensives) as at cessation (phase V of Korotkoff sounds) of sounds. Both observers took the readings measurements were performed in the same way but from the same mercury manometer without being with the cuff bearing the stethoscope head wrapped aware of the values of the other observer. around the left arm. In a pre-study validation test the difference (mean ± s.d.) between the two observers (72 measurements, Part B double-stethoscope, 36 normotensive subjects, cubi- One could object to a same-arm test design that dif- tal artery) was −0.2 ± 1.4 mm Hg for the systolic BP ferences of pressure readings between the brachial and +0.5 ± 2.6 mm Hg for the diastolic BP. and cubital artery might be due to a disturbance of All measurements were done in a completely the sound conduction along the artery induced by silent environment. the stethoscope head placed proximally. To test this hypothesis the same observers as in part A per- Statistical analysis formed 12 simultaneous measurements in a same- arm and opposite-arm test design (Figure 1) in 20 For all pairs of readings the BP over the cubital normotensive young women (mean age ± s.d.: artery was subtracted from that over the brachial 21.3 ± 3.9 years, range: 15–30 years). The first four artery (a.b. − a.c.). For each subject the mean value comparative measurements were performed in the for the 10 (Part A) or four (same-arm, Part B) and same way as in part A on the right arm, with auscul- eight (opposite-arm, Part B) differences, respect- tation over the a.c. by O1 for two measurements and ively, represented the basis of our analysis (mean by O2 for the following two measurements. There- reading difference = MRD). The values showed suf- after two cuffs (one with an inserted stethoscope ficient normality. For the description of the results head) for the opposite-arm test were connected via we used mean values and standard deviations as a y-tube to the mercury manometer allowing simul- well as scatterplots with least square linear taneous measurements at both arms. For the follow- regression analysis. The influence of the variables ing four comparative measurements the cuff with gender, age, upper arm circumference and hyperten- the inserted stethoscope head was wrapped around sion (yes/no), proven as potential factors in earlier the right arm. After two of these measurements O1 studies, was analysed by multifactorial analysis of and O2 changed the auscultation site. The last four variances using these factors as well as the interac- Differences in BP value using Korotkoff sounds F Weber et al 697 tions of gender with upper arm circumference and Part B age, respectively. As level for statistical significance For all 20 normotensive women in the same-arm test a P value of 0.05 (two-sided) is considered. The ± empirically resulting P-values are given descrip- design BP measured inside the cuff (mean s.d.: 115 ± 12/68 ± 8 mm Hg) were 0.8 ± 1.0 mm Hg tively without further adjustment for multiple test- ± ing. higher for the systolic and 8.5 2.2 mm Hg lower (P Ͻ 0.001) for the diastolic BP compared to the measurements in the fossa cubiti (mean ± s.d.: 114 ± 12/77 ± 8 mm Hg). Results Nearly the same result was achieved in the opposite-arm test design.