Family Practice Vol. 19, No. 5, 439-441
© Oxford University Press 2002
The differential blood pressure sign in general practice: prevalence and prognostic value
The Mid Devon Medical Practice, School Surgery, Fore Street, Witheridge, Devon EX16 8AH and
a Research & Development Support Unit, School of Postgraduate Medicine and Health Sciences, Noy Scott House, Haldon View Terrace, Exeter EX2 5EQ, UK.
Dr CE Clark; E-mail: Chris.Clark{at}gp-L83023.nhs.uk
Clark CE and Powell RJ. The differential blood pressure sign in general practice: prevalence and prognostic value. Family Practice 2002; 19: 439–441.
Received 6 September 2001; Revised 8 November 2001; Accepted 13 May 2002.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. Patients sometimes have differences of
20/10 mmHg in their blood pressure depending on which arm is measured. The prevalence and prognostic value of this finding in general practice are unknown. If these differences are due to peripheral vascular disease, these patients may be at increased risk of cardiovascular or cerebrovascular events. Objective. Our aim was to establish the frequency and prognostic value of a blood pressure difference between arms in one rural general practice.
Methods. Paired blood pressure readings were collected from patients attending the surgery. The outcome measures of myocardial infarction, new diagnosis of angina, a cerebrovascular event or death were recorded prospectively.
Results. A total of 280 patients were examined, and of these 13.6% had a systolic blood pressure difference (SBPD) of 20 mmHg, and 23.2% a diastolic blood pressure difference (DBPD) of 10 mmHg. Eighty-three patients were followed-up for 5.6 years. Patients with a DBPD of 10 mmHg showed a mean event-free survival of 3.3 years [95% confidence interval (CI) 2.2–4.4] compared with 5.0 years (95% CI 4.7–5.3) for those with a DBPD of <10 mmHg (P < 0.0001). Patients with an SBPD of 20 mmHg showed a mean event-free survival of 3.5 years (95% CI 2.3–4.7) compared with 4.9 years (95% CI 4.5–5.2) for an SBPD of <20 mmHg (P = 0.043).
Conclusions. During a single assessment of blood pressure, there will be a minority of patients with a difference of 20/10 mmHg between their right and left arms. Measurement of both arms is therefore necessary to diagnose and treat hypertension accurately. This study suggests an association between blood pressure difference and increased morbidity and mortality. Priority should be given to managing other risk factors aggressively in those patients with a reproducible blood pressure difference of 20/10 mmHg.
Keywords. Blood pressure measurement, general practice, hypertension.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Guidelines suggest that, in blood pressure measurement, if a consistent difference of at least 20 mmHg systolic or 10 mmHg diastolic exists, then the arm with the higher pressure should be used for decision making.1 Many studies have examined selected populations to establish the prevalence of such a difference, using sequential or simultaneous techniques with standard, random zero or automated sphygmomanometers.2 Sequential measurement with a standard sphygmomanometer is the simplest and most practical method of detecting a difference in general practice. The prevalence of a difference in an unselected general practice population is unknown, but potentially important, as such a finding would require repeated measurements to confirm it and might influence management by allowing better standardization of measurements using the arm with the higher pressure.
The prognostic value of such a finding is also unknown. If it is assumed that peripheral vascular disease is the cause, then a blood pressure difference could predict an increased risk of ischaemic cardiac and/or cerebrovascular events. If this were the case, then the detection of a difference might allow more appropriate targeting of limited resources in primary prevention.
This pilot study was undertaken to establish the point prevalence of a difference of 20/10 mmHg at a single consultation, and to examine whether there might be an association of a difference of 20/10 mmHg between the arms with subsequent cardiovascular and cerebrovascular events.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Patients attending for re-registration medicals had their blood pressures measured sequentially in both arms with a standard calibrated mercury sphygmomanometer. Second and third pairs of measurements were obtained opportunistically at subsequent visits. Patients with an upper limb deformity or hemiparesis were excluded. The measurements obtained were used to calculate the point prevalence of a difference of
20/10 mmHg. Subjects with two or three pairs of measurements were monitored prospectively to identify all deaths, cardiovascular-related deaths or cardiovascular and cerebrovascular events. Demographic information was obtained from patients’ records and the results were analysed using SPSS for Windows v9.0
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Prevalence
Pairs of readings were collected from 280 patients (age range 6–88, median 60 years) from May 1994 to October 1995. The prevalence of a systolic blood pressure difference (SBPD) of
20 mmHg was 13.6% [95% confidence interval (CI) 9.6–17.6%], and of a diastolic blood pressure difference (DBPD) of 10 mmHg was 23.2% (95% CI 18.3–28.1%). The frequency of a DBPD fell to 14.7% (95% CI 10.5–18.8%) in those with three pairs of readings; therefore, only patients with two or three pairs of readings were studied for the prognostic significance of a difference.
Prognosis
Eighty-three subjects had two or three pairs of readings collected during the study period. There were 45 (54%) females and 38 (46%) males. Median age was 69 years (range 19–86). Thirteen (16%; 95% CI 8.1–23.9%) were smokers; 17 (21%; 95% CI 12.2–29.7%) had cholesterol measurements recorded (mean 6.3 mmol/l); nine (11%; 95% CI 4.3–17.7%) had a history of ischaemic heart disease (IHD); and four (5%; 95% CI 0.3–9.7%) had a history of cerebrovascular events (CVE).
Fifty-three (64%; 95% CI 53.5–74.2%) of the patients were hypertensive. Mean systolic pressure was 159 mmHg (range 95–249), mean diastolic was 85 (range 59–106). Eleven (13%; 95% CI 5.9–20.5%) patients had an SBPD of 20 mmHg, and 14 (17%; 95% CI 8.8–24.9%) a DBPD of 10 mmHg. Mean absolute pressure differences were 9.1/4.8 mmHg, and no association was demonstrated between the absolute blood pressure values and SBPD or DBPD.
The maximum time from entry to follow-up was 5.6 years. During this period, 17 patients (20%; 95% CI 11.8–29.2%) had IHD events, and two (2%; 95% CI 0–5.7%) had CVEs. Six patients died (one from cancer and five from cardiovascular disease), giving a total of 20 (24%; 95% CI 14.9–33.3%) events or deaths. There was no significant association with previous history of IHD or CVE.
For patients with a DBPD of 10 mmHg, the mean event-free survival was 3.3 years (95% CI 2.2–4.4) compared with 5.0 years (95% CI 4.7–5.3; P < 0.0001) for a DBPD of <10 mmHg (Fig. 1
). For patients with an SBPD 20 mmHg, mean event-free survival was 3.5 years (95% CI 2.3–4.7) compared with 4.9 years (95% CI 4.5–5.2; P = 0.043) for <20 mmHg.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This study has established that the point prevalence of an SBPD of
20 mmHg is 13.6%, and of a DBPD of 10 mmHg is 23.2% for a single consultation in one unselected general practice population. A literature review2 of 11 previous studies showed a prevalence range of 12–18.4% for an SBPD of 20 mmHg, and 13–33.7% for a DBPD of 10 mmHg. Both of the figures for this study fall within this range, and indicate how many patients might require further assessment. Cyriax described such a difference in 1921 as the "differential blood pressure sign",3 and it has since become accepted wisdom that such a finding should be investigated further.1 No previous publication was found linking blood pressure differences to prognosis. Early workers attributed such differences to functional vascular changes,4 or the normal anatomy of the aortic arch and vessels.5 However, a higher prevalence of differences has been observed in patients with coronary heart disease or peripheral vascular disease, which suggests a pathological cause.6 Another study showed that vascular disease patients with differences displayed evidence on angiography of arterial stenosis on the side of the lower pressure,7 also suggesting that a difference could be associated with peripheral vascular disease. Thus detection of a difference might indicate existing asymptomatic peripheral vascular disease.
This would offer one explanation of our findings of reduced event-free survival times for patients with systolic or diastolic differences. However, this preliminary study lacks sufficient size to draw a firm conclusion. The sequential method of sampling adopted was convenient but took no account of the possible within-person variation, or any effects of the sequential measurements themselves on blood pressure. Other shortcomings were the unblinded method of data collection, and the use of a standard rather than a random-zero sphygmomanometer.
The majority of subjects (64%) followed-up were hypertensive, because they were more likely to have repeated measurements than normotensive patients. It is possible that the observed blood pressure differences occurred in patients with higher blood pressure, and therefore higher risks of events. In this case, it should have been possible to detect a correlation between blood pressure difference and blood pressure readings, which we could not. Two previous studies which compared normotensive and hypertensive patients showed conflicting results,8,9 thus the absolute blood pressure may not fully explain these observations.
Even if confounding variables can explain our results, detecting a blood pressure differential may still identify those patients who should be assessed further to determine their overall risk of cardiovascular or cerebrovascular events. Further work is underway to correlate these measurements with a simultaneous method of assessment, and with the ankle/brachial pressure index, which has also been proposed as a useful test to identify those who should be managed most aggressively,10 but requires more specialized equipment, training and time to measure. Until further evidence is available, we suggest that both arms should be measured when assessing every new hypertensive in order to standardize readings, and those patients with a difference should be managed aggressively on the assumption that they may already have asymptomatic peripheral vascular disease.
|
Acknowledgments |
|---|
CEC was supported by a bursary from the Somerset and North & East Devon Primary Care Research Network.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 O’Brien ET, Petrie JC, Littler WA et al. Blood Pressure Measurement: Recommendations of the British Hypertensive Society, 3rd edn. London: BMJ Publishing Group, 1997.
2 Clark CE. Differences in blood pressure between arms might reflect peripheral vascular disease. Br Med J 2001; 323: 399–400.
3 Cyriax EF. Unilateral alterations in blood pressure: the differential blood pressure sign (second communication). Q J Med 1921; 9: 309–313.
4 Kay WE, Gardner KD. Comparative blood pressures in the two arms. Calif West Med 1930; 33: 578–579.
5 Amsterdam B, Amsterdam AL. Disparity in blood pressures in both arms in normals and hypertensives and its clinical significance. NY J Med 1943; 43: 2994–2300.
6 Frank SM, Norris EJ, Christopherson R, Beattie C. Right and left arm blood pressure discrepancies in vascular surgery patients. Anesthesiology 1991; 75: 457–463.[Web of Science][Medline]
7 Moll F, Six J, Mutsaerts D. Misleading upper extremity blood pressure measurements in vascular occlusive disease. Bruit 1983; 8: 18–19.
8 Southby R. Some clinical observations on blood pressure and their practical application, with special reference to variation of blood pressure readings in the two arms. M J Aust 1935; 2: 569–580.
9 Harrison EG, Roth GM, Hines EA. Bilateral indirect and direct arterial pressures. Circulation 1960; 22: 419–436.
10 Applegate WH. Ankle/arm blood pressure index. A useful test for clinical practice? J Am Med Assoc 1993; 270: 497–498.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Agarwal Response to Prognostic Significance of Between-Arm Blood Pressure Differences: Between-Arm Blood Pressure Difference and Mortality and the Interarm Blood Pressure Difference: Between-Arm Blood Pressure Differences Are Important Hypertension, August 1, 2008; 52(2): e16 - e17. [Full Text] [PDF]
|
|
|
|
|
|
C. E. Clark and J. L. Campbell The Interarm Blood Pressure Difference Hypertension, August 1, 2008; 52(2): e15 - e15. [Full Text] [PDF]
|
|
|
|
 |
|
 C. E Clark, J. L Campbell, R. J Powell, and J. F Thompson The inter-arm blood pressure difference and peripheral vascular disease: cross-sectional study Fam. Pract., October 1, 2007; 24(5): 420 - 426. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||