Human blood pressure determination by sphygmomanometry D Perloff… · Perloff et al Human Blood Pressure Determination by Sphygmomanometry are stressed in light of the epidemiological

1524-4539 Copyright © 1993 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online ISSN:Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX 72514

1993;88;2460-2470 CirculationD Perloff, C Grim, J Flack, ED Frohlich, M Hill, M McDonald and BZ Morgenstern

Human blood pressure determination by sphygmomanometry

http://circ.ahajournals.orgthe World Wide Web at:

The online version of this article, along with updated information and services, is located on

http://www.lww.com/reprintsReprints: Information about reprints can be found online at

[email protected]. E-mail: Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax: Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters Kluwer

http://circ.ahajournals.org/subscriptions/Subscriptions: Information about subscribing to Circulation is online at

by on November 17, 2009 circ.ahajournals.orgDownloaded from

http://circ.ahajournals.org


http://circ.ahajournals.org/subscriptions/

mailto:[email protected]

http://www.lww.com/reprints

2460

AHA Medical/Scientific StatementSpecial Report

Human Blood PressureDetermination by Sphygmomanometry

Dorothee Perloff, MD; Carlene Grim, MSN, SpDN; John Flack, MD; Edward D. Frohlich, MD;Martha Hill, PhD, RN; Mary McDonald, MSPH, RN; and Bruce Z. Morgenstern, MD,

Writing Group

T his is the sixth edition of the American HeartAssociation's recommendations for indirectmeasurement of arterial blood pressure. The

purpose is to facilitate the AHA's mission to reducedisability and death from cardiovascular diseases andstroke by providing practical guidelines for measuringan important physiological variable that contributes toand is associated with these major diseases. The recom-mendations are intended to provide a uniform standardand methodology for day-to-day use by all health careworkers and for use in epidemiological, educational,and clinical studies. This edition clarifies and updatesinformation provided in previous versions. The writinggroup is indebted to the previous groups and theirchairmen for their extensive work and expresses specialappreciation to the late Dr. Walter Kirkendall for hislifelong work in the field of hypertension.

Arterial blood pressure, one of the "vital signs," is animportant indicator of a person's state of health; there-fore, its measurement is a part of every completephysical examination. Blood pressure measurement isdone to screen for hypertension, to assess a person'ssuitability for certain occupations and activities, toestimate long-term cardiovascular risk, to determineeligibility for insurance, and as a part of the manage-ment of patients with many types of medical problems.Inappropriately low blood pressure, or clinical shock, isa medical emergency, and inappropriately high bloodpressure is a marker for the chronic condition hyper-tension, which is a major risk factor for prematurecardiovascular, cerebrovascular, renovascular, andother vascular diseases.The gold standard for measurement of arterial pres-

sure is direct intra-arterial measurement with a cathe-ter. However, this technique is neither practical norappropriate for repeated measurements in nonhospital-ized patients or asymptomatic individuals, nor for large-scale public health screenings. Instead, the indirectmethod of measurement is commonly used. With thistechnique the pressure required to collapse the artery inthe upper arm or leg is determined by use of a sphyg-momanometer (an occluding cuff, stethoscope, and ma-nometer). The cuff is inflated to a level above arterialpressure (as indicated by obliteration of the pulse). As

Requests for reprints should be sent to the Office of ScientificAffairs, American Heart Association, 7272 Greenville Avenue,Dallas, TX 75231-4596.

the cuff is gradually deflated, the pressure is noted atwhich sounds produced by the arterial pulse waves(Korotkoff sounds) appear and disappear again as flowthrough the artery resumes. The direct and indirectmethods yield similar measurements, but these arerarely identical because the direct method measurespressure and the indirect method is more indicative offlow. The indirect method is generally less accurate andless reproducible. Nevertheless, it is sufficiently accu-rate for many diagnostic and therapeutic studies andwill continue to be used because it is practical, simple,low in cost, and noninvasive.Because the level of arterial pressure is the basis for

major diagnostic and therapeutic decisions in medicine,the measurement must be correct and maximally repro-ducible. Although the occluding cuff technique appearssimple and easy to learn, there are many possible causesof error and inaccuracy. Therefore, the technique formeasurement should be standardized so that informa-tion from different observers is comparable and can bereadily used in serial evaluations of an individual or forepidemiological and research studies. This report out-lines a simple, standardized, step-by-step protocol forthe indirect measurement of blood pressure. Routine,careful adherence to this protocol will facilitate accu-rate and reproducible (and reliable) quantification ofthis important physiological variable. This report isintended for everyone who measures blood pressure:health care workers, researchers, and people who mea-sure their own blood pressure.The scientific bases for these recommendations are

provided when available; otherwise, consensus recom-mendations are made, in the interest of standardizingthe procedure. The bibliography provides the sources ofour recommendations and is a basis for further reading.The special considerations of measuring blood pressurein infants, children, the elderly, pregnant women, andthe obese are reviewed. Because of the growing interestin self-measurement and ambulatory blood pressuremeasurement, these techniques and their applicabilityare also discussed. Appendix A is a list of potentialerrors in measuring blood pressure and ways to corrector avoid them. We emphasize the need for systematictraining and recertification of observers and the desir-ability of regular recalibration of equipment. The im-portance of making accurate blood pressure measure-ments for the evaluation and treatment of hypertension



Perloff et al Human Blood Pressure Determination by Sphygmomanometry

are stressed in light of the epidemiological importanceof this diagnosis.

Epidemiology of HypertensionHypertension is one of the major modifiable risk

factors for stroke, coronary heart disease, congestiveheart failure, renal failure, and peripheral vasculardisease.' It is estimated that 50 million Americans havea blood pressure level consistently at or above 140/90mm Hg, and are therefore at increased risk for devel-oping complications.2 The risk from hypertension ispresent whether only systolic pressure, only diastolicpressure, or both are elevated.3 Other potentially mod-ifiable risk factors for cardiovascular, cerebrovascular,renovascular, and other vascular diseases include dys-lipidemias, tobacco use, lack of exercise, diabetes mel-litus, and excessive alcohol intake; the concurrent pres-ence of multiple risk factors further increases the risk ofdeveloping cardiovascular disease.24'5 The diagnosis ofhypertension consequently has serious implications forlongevity and also affects insurability, employability,and suitability for certain occupations and activities. Itis therefore imperative that the diagnosis of hyperten-sion be based on accurate, representative, and repro-ducible measurements.

Hypertension is defined as a persistently elevatedblood pressure; that is, a pressure that exceeds anarbitrarily set level of normalcy.6 Pressures above thislevel are associated with an escalating risk for thedevelopment of arteriosclerosis, left ventricular hyper-trophy, nephrosclerosis, and cerebrovascular disease.Involvement of these target organs in turn leads to riskof decompensation of cardiac and renal function and ofthrombotic, embolic, aneurysmal, or hemorrhagic vas-cular disease.5 The risk increases progressively: thehigher the blood pressure and number of concurrentrisk factors, the more advanced the degree of targetorgan involvement.57 Systolic and diastolic blood pres-sure are closely correlated and are correlated with riskof cardiovascular disease independently and in combi-nation. However, high systolic pressure may contributemore to the risk of complications than high diastolicpressure.3 There is no blood pressure level below whichthere is no risk or above which cardiovascular compli-cations are inevitable. The arbitrary division between"normal" and "abnormal," despite the fact that riskincreases in a continuum, is useful for classification ofindividual patients and for facilitation of both diagnosticand therapeutic decisions. Because the risk that anelevated pressure will lead to clinical sequelae is longterm rather than short term, the diagnosis of hyperten-sion should be based on multiple determinations overseveral visits. The Joint National Committee on theDiagnosis, Evaluation, and Treatment of Hypertension(JNC V) has recommended that in the nonhospitalizedadult at least two measurements be made on two ormore separate occasions, and only if the average of tworeadings is at or above 140/90 mm Hg should theindividual be labeled as having hypertension (see Ap-pendix B).2 A staged approach is, of course, unneces-sary in the patient who has very high or rapidly risingblood pressure and is therefore at risk for developingacute cardiovascular complications.Blood pressure levels normally fluctuate considerably

from day to night or over longer periods of time.8 Blood

pressure is influenced by level of activity, exercise orrest, degree of wakefulness or sleep, environmentalfactors such as temperature, mood (friendly or hostile),and a multitude of other emotional or psychologicalfactors that reflect a person's response to the internaland external milieu.9'10 No single blood pressure mea-surement reflects the entire day's fluctuations. Yetblood pressure is frequently treated as a definitive,constant, physiological characteristic, and the individualnumerical values are used as the basis for major diag-nostic and therapeutic decisions, the most important ofwhich is the diagnosis of hypertension. Therefore, tominimize the influence of extraneous factors on bloodpressure variability, the use of a standardized protocolfor measuring blood pressure is critical.

Indirect Blood Pressure MeasurementIn 1896 Riva Rocci introduced a method for indirect

measurement of blood pressure, based on measuringthe external pressure required to compress the brachialartery so that arterial pulsations could no longer betransmitted through the artery. The artery is occludedby wrapping an inflatable bladder, which is encased in anondistensible cuff, around an extremity, and inflatingthe bladder until the pressure in the cuff exceeds that inthe artery. When the artery is occluded, transmittedpulse waves can no longer be palpated or heard distal tothe point of occlusion. As the pressure in the bladder isreduced by opening a valve on the inflation bulb,pulsatile blood flow reappears through the partiallycompressed artery, producing repetitive sounds gener-ated by the pulsatile flow. These "Korotkoff sounds" arenamed for the Russian physician who first described theauscultatory method in 1905."1 The level of the pressurein the inflatable bladder (reflected by the level on themanometer to which it is connected) at the appearanceof the first Korotkoff sound is the maximum pressuregenerated during each cardiac cycle: the systolic pres-sure. The level of pressure at which the sounds disap-pear permanently, when the artery is no longer com-pressed and blood flow is completely restored, is theresting pressure between cardiac contractions: the dia-stolic pressure."As the pressure is reduced during deflation of the

occluding cuff, the Korotkoff sounds change in qualityand intensity. The five phases of this change are char-acterized as follows:

Phase I: First appearance of clear, repetitive, tappingsounds. This coincides approximately withthe reappearance of a palpable pulse.

Phase II: Sounds are softer and longer, with the qual-ity of an intermittent murmur.

Phase III: Sounds again become crisper and louder.Phase IV: Sounds are muffled, less distinct, and softer.Phase V: Sounds disappear completely.The pressure at which the sounds first appear (onset

of Phase I) corresponds to the systolic pressure. Therehas been some debate, however, over whether thepressure at the onset of muffling (Phase IV) or disap-pearance of sound (Phase V) best corresponds withdiastolic blood pressure.' Generally, disappearance ofsound correlates better with intra-arterial pressure thandoes muffling, and the pressures at which muffling and

2461



2462 Circulation Vol 88, No 5, Part 1 November 1993

disappearance occur often differ by only a few millime-ters of mercury. Therefore, the onset of Phase V is usedto define the diastolic pressure in adults. The onset ofPhase V is defined by the level at which the last soundis heard or by the onset of silence, the level 2 mm Hgbelow that. Both definitions have been used in epide-miological studies, but the onset of silence, or absenceof sound, is difficult to define or describe accurately, andthe level of the last audible sound heard is easier toteach and to standardize. The consensus of the commit-tee is therefore that, in the interest of uniformity, PhaseV should be operationally defined by the last soundheard. In children less than 13 years old, pregnantwomen, and patients with high cardiac output or periph-eral vasodilatation, sounds are often heard at levels farbelow those at which muffling occurs, sometimes tolevels approaching 0 mm Hg. In these situations, forpractical purposes, muffling should be used to indicatediastolic pressure, but both muffling (Phase IV) anddisappearance (Phase V) should be recorded.

Occasionally the Korotkoff sounds become inaudibleduring Phase II or III, only to reappear as the pressurein the cuff is reduced further. This period of silence iscalled the auscultatory gap and is especially common inolder and hypertensive patients.13Pulsus alternans, or beat-to-beat variation in the

intensity of the Korotkoff sounds, occurs in people withfluctuating systolic pressure. When the Korotkoffsounds first appear, while the arterial lumen is stillpartially compressed, only alternate ventricular contrac-tions generate a systolic pressure sufficiently high toproduce a pulse wave that results in audible sounds. Asthe pressure in the occluding cuff is reduced and thepressure in the inflatable bladder falls below the levelgenerated by even the weaker systolic contractions,these too produce a pulse wave and the frequency of theKorotkoff sounds suddenly doubles.Pulsusparadoxus is an exaggeration of the normal fall

in systolic pressure that occurs during inspiration. Dur-ing inspiration, venous return to the right atrium and,therefore, right ventricular stroke volume increase, butvenous return to the left atrium and left ventricularstroke volume decrease. This reduction of pressure isnormally on the order of 2 to 4 mm Hg, and isconsidered abnormal when it exceeds 10 mm Hg. Pulsusparadoxus is commonly seen in patients with constric-tive pericarditis or pericardial tamponade, but may alsooccur in patients with severe pulmonary disease anddeep labored breathing or restrictive cardiomyopathy.

Procedure for Measuring Blood PressureThe systematic step-by-step procedure for measuring

blood pressure is described in this section. The errorsthat can occur at each stage, and suggestions for avoid-ing them, are summarized in Appendix A.

EquipmentThe equipment for indirect measurement of blood

pressure consists of a stethoscope, or sensing micro-phone for auscultation or detection of the Korotkoffsounds, and a sphygmomanometer. The sphygmoma-nometer comprises a manometer (mercury or aneroid),with a calibrated scale for measuring pressure, and aninflation system. The latter consists of an inflatablebladder encased in a nondistensible cuff that can be

securely wrapped around the limb, an inflation bulb formanual inflation of the bladder in the cuff, and tubingconnecting both the manometer and the inflation bulbto the bladder.

Stethoscope. The stethoscope, which consists of amicrophone or head, tubing, and ear pieces, is placedover the occluded artery to amplify the Korotkoffsounds. The bell, or low-frequency filter of the micro-phone, permits more accurate auscultation of theKorotkoff sounds than the diaphragm, especially atdiastolic pressures; its routine use is therefore recom-mended.14 However, the diaphragm is widely usedbecause it is convenient, easily placed, and moregenerally available. Some of the sphygmomanometersdesigned for self-measurement of blood pressure areequipped with a microphone or sensor built into theoccluding cuff. This obviates the inconvenience ofholding the bell over the brachial pulse with one handwhile compressing the bulb with the other hand, butmay introduce extraneous sounds. The length of thetubing from the ear pieces to the microphone shouldbe long enough that the seated observer can conve-niently place the bell over the artery and listen for thesounds while observing the manometer at eye level (12to 15 in, 30 to 38 cm).Manometer. The calibrated manometer reflects the

pressure in the occluding cuff by the height of a columnof mercury or by the location of a rotating needle on adial scale. The mercury manometer is preferred andrecommended over the aneroid manometer because it ismore accurate, easier to maintain, and less likely tobecome decalibrated.'5 However, because of the risk ofthe toxic effects of mercury spills, mercury manometersmust be handled carefully and their use has beendiscouraged in some areas of high traffic where acciden-tal spills are more prone to occur. The column ofmercury should have a height and calibration scale fromo to 300 mm Hg, marked at intervals of 2 and 10 mm Hg.Smaller scales from 0 to 260 mm Hg are also available.The mercury reservoir should be full so that the uppercurve of the meniscus is at the zero level before the cuffis inflated. The mercury should rise and fall freely in thecolumn; bouncing or hesitation of the mercury duringdeflation reflects clogging of the air vent or dirt or airbubbles in the column. The column should be keptvertical except on the specially designed floor models,which have a slanted scale. See Appendix C for safetyprecautions for the use of mercury.The aneroid manometer is also widely used and can

provide accurate measurements if properly calibrated.16However, because of its construction, it is prone tomechanical alterations that can affect its accuracy. Theaneroid manometer consists of a metal bellows, whichexpands as the pressure in the cuff increases, and amechanical amplifier that transmits this expansionthrough a lever to the indicator needle, which rotatesaround a circular, calibrated scale. The needle shouldrest at the zero point before the cuff is inflated andreturn to that point after the cycle of inflation anddeflation. Aneroid manometers require maintenanceevery 6 months and should be handled gently to avoiddecalibration. The accuracy of the calibration should bechecked regularly (Fig 1). Recalibration is requiredwhen the readings differ from the standard mercurymanometer by more than 4 mm Hg. When decalibrated,




1,9 AA

S- - -I I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

L - - -- -- -I

A: Ideal arm circumference; B: Range of acceptable arm circumferences;C: Bladder length; D: Midline of bladder; E: Bladder width; F: Cuff width

FIG 2. Dimensions of bladder and cuff in relation to armcircumference.

FIG 1. Calibrating the aneroid manometer with the mer-cury manometer using a Y tube.

aneroid manometers tend to underread the pressure,especially at higher levels, but may be inconsistent intheir variation from the mercury standard at any bloodpressure level.

Inflation system. The inflation system consists of anexpandable, rectangular rubber bladder encased in anondistensible cuff, an inflation bulb, and connectingtubing. The cuffs used for oscillometric determinationof the blood pressure do not have a removable bladderseparate from the cuff. The cuff is applied by wrappingit completely around the limb so that the noninflatableportion overlaps that containing the bladder, and issecured with a self-adhesive material such as Velcro.The bladder is connected by rubber tubing to themanometer, which reflects the pressure inside the blad-der. The tubing must be free of leaks and long enoughthat the subject can be comfortably positioned next tothe manometer without the tube's becoming kinked orcompressed, which could interfere with smooth inflationor deflation. The bladder is also connected by a shortertube to a valved rubber bulb, which is used to inflate anddeflate the bladder at a rate controlled by the releasevalve.

Considerable thought has been given to the size anddimensions of the occluding bladder and its encasingcuff.17-19 The variables considered are the length andwidth of the bladder and the ratio of one to the other,the relation of the width of the bladder to the length ofthe upper arm, and the relation of both the width andthe length of the bladder to the circumference of thearm. For any arm circumference, as the bladder size(length and width) is increased in a stepwise manner theblood pressure reading is progressively lower, until aplateau is reached at which no further increase in cuffsize results in further reduction of the blood pressuremeasurement.'9 This plateau occurs approximately

when the width of the bladder is 40% of the circumfer-ence of the upper arm at the midpoint and the length ofthe bladder is 80% of the circumference of the arm. Theuse of a bladder that is too small (too narrow or tooshort) for the circumference of the limb results inoverestimation of the blood pressure.19 The error ofunderestimation of the pressure, which results from theuse of a bladder that is too large, is less than the errorof overestimation with a bladder that is too small. Todate there is no published systematic study that relatesthe measurements made with stepwise increments incuff dimensions (length and width), in arms of varyingcircumferences and lengths, to simultaneous intra-arte-rial measurements. Hence, the committee's consensusrecommendation is based on the available literature. Itis recommended that the width of the bladder be 40% ofthe arm circumference, and the length of the bladder belong enough to encircle at least 80% of the arm inadults. In children the occluding bladder should be longenough to encircle the arm completely (100%). Over-lapping of the ends of the bladder in children does notappear to introduce an error in measurement.The cuffs generally available have been classified by

the width of the bladder rather than by the length, andare labeled "newborn," "infant," "child," "small adult,""adult," "large adult," and "thigh," but different man-ufacturers have produced cuffs of varying dimensions(both length and width) under these names because nouniversal standards have been established. A largenumber of cuffs with varying sizes of bladders arecommercially available, but not all of these are of therecommended dimensions. Ideally, every cuff should belabeled with the dimensions of the enclosed bladder; aline should mark the center of the bladder, and twolines should indicate the range of arm circumferencesfor which the bladder is suitable, ie, encircling 80% to100% of the circumference (Fig 2). Unlabeled cuffsshould be so marked by the user. For cuffs with longerbladders, a length:width ratio of more than 2:1, thecorresponding appropriate arm circumference isgreater; however, the ratio of bladder width to armcircumference should be as close as possible to 0.40.(See Appendix D for a list of some acceptable bladdersizes and the arm circumferences for which they aresuitable. A number of cuffs with intermediate bladderdimensions are available, but to simplify the selectiononly some of them are listed here.) Although it is notfeasible for every examiner to have all cuff sizes avail-able, except under research conditions, it is strongly

1-0 a *1- C a

i_-

fr

2463




recommended that the practitioner have several sizesavailable to meet the needs of the population served.However, in individuals with very wide but short arms,the appropriately sized bladder may be difficult to apply.Likewise, in individuals with very large or musculararms, even the largest cuff may be inadequate. TheBritish Hypertension Society has recommended the useof one large cuff (12.5 x 35 cm) for all adults with anarm circumference up to 42 cm, to avoid the need formultiple cuff sizes.18 20 The use of such a cuff could leadto systematic underestimation or overestimation of thepressure when the ratio of bladder width to arm circum-ference is different from 0.40.Automated devices. Many automated devices are avail-

able to measure blood pressure by both auscultatoryand oscillometric techniques. The oscillometric methodis based on detecting the oscillations on the lateral wallsof the occluded artery as the cuff is deflated. Theoscillations begin at approximately the level of systolicpressure and reach their greatest amplitude at the levelof mean arterial pressure. Diastolic pressure is a derivedvalue. Systolic blood pressure measurement by thesedevices is accurate, but diastolic pressure, which isderived from the systolic and mean pressures, may notbe. The cuffs used for oscillometric measurement areconstructed without a removable bladder. Althoughabsolute measurements made with these cuffs may varyslightly from those made with standard cuffs, overalltrends in blood pressure level can be readily tracked.However, serious doubt has been raised about theaccuracy of devices applied to the finger or wrist be-cause their extreme sensitivity to position results in widefluctuations in blood pressure readings except whenthey are used under strictly standardized and constantconditions; hence, their use is not recommended. Dopp-ler devices, which amplify the Doppler signal fromflowing blood, are also used with standard sphygmoma-nometers and obviate the need for a stethoscope. Theyare especially useful in taking infants' blood pressure orin situations where the auscultatory signal is difficult tohear.

ObserverEvery person who makes indirect blood pressure

measurements must be carefully trained and madeaware of the potential pitfalls. Several excellent pro-grams, some using videocassettes, provide standardizedinstruction, training, and testing of observers.21,22 Un-fortunately, many health care professionals do not par-ticipate in regular retraining programs to improve andreassess their skills in blood pressure measurement,despite considerable variability in their knowledge, skill,and technique. The observer must be able to concen-trate on the task and have reasonably good eyesight,hearing, and manual dexterity as well as hand-eye-earcoordination. The observer must be comfortably posi-tioned to be able to (1) inflate and deflate the bladder inthe cuff gradually, (2) see the manometer and themeniscus of the column of mercury or the indicatorneedle on the aneroid scale, (3) hear the Korotkoffsounds, differentiating them from extraneous noises, (4)make note of and remember the level of the pressure atthe first appearance, at muffling, and at the disappear-ance off the Korokof sounds, while continuing to

the systolic and diastolic pressure (Phases IV and V)accurately to the nearest 2 mm Hg.

Observer errors can also result from subconsciousbiases. Terminal digit preference is caused by thetendency to round pressure readings off to numbersending in zero instead of recording more accurately tothe nearest 2 mm Hg. A cut-off or direction bias resultsin falsely recording pressures as being above or below apredetermined level or dividing line between "normal"and "hypertensive." An observer may also be influencedby knowledge of earlier readings.

SubjectFor screening and monitoring purposes, the blood

pressure is measured in the upper arm, with the subjectseated. Serial measurements should be performed onthe same arm for consistency. Under clinical circum-stances, measurements are often performed in otherpositions as well. In the initial evaluation of hyperten-sive patients, the blood pressure should be measured inboth arms and occasionally in the legs. The subject mustbe comfortably seated with the midpoint of the upperarm at the level of the heart (approximately the fourthintercostal space, when the individual is seated).23When the subject is standing, care must be taken tosupport the raised arm at the level of the heart. Whenthe subject is lying down, the arm should be at the sideof the body, slightly raised from the bed or examinationtable, at the level of the middle of the chest. Ideally themeasurements should be made after a period of rest ina quiet, relaxed setting, not immediately after exertionor ingestion of coffee or during conversation; the legsshould be uncrossed, with the feet resting firmly on thefloor, not dangling, and the back supported, because anyform of isometric exercise during the measurement willtransiently raise the blood pressure level.24 Blood pres-sure levels are affected by environmental, emotional,and physical stimuli, so every effort should be made tostandardize the conditions of the measurement, keepingextraneous influences to a minimum. Anticipation ofpain or anxiety about the procedure and its outcomecan raise the blood pressure level and potentially lead tooverestimation of the usual blood pressure levels.

TechniqueThe intent and purpose of the measurement should

be explained to the subject in a reassuring manner andevery effort made to put the subject at ease.The sequential steps for measuring the blood pres-

sure in the upper extremity, as for routine screening andmonitoring purposes, should include the following:

1. Have paper and pen at hand for immediaterecording of the pressure.

2. Seat the subject in a quiet, calm environment withhis or her bared arm resting on a standard table or othersupport so the midpoint of the upper arm is at the levelof the heart.

3. Estimate by inspection or measure with a tape thecircumference of the bare upper arm at the midpointbetween the acromium and olecranon process (betweenthe shoulder and elbow) and select an appropriatelysized cuff. The bladder inside the cuff should encircle80% of the arm in adults and 100% of the arm in

deflate the occluding cuff, and (5) remember and record children less than 13 years old. If in doubt, use a larger




cuff. If the available cuff is too small, this should benoted.

4. Palpate the brachial artery and place the cuff sothat the midline of the bladder is over the arterialpulsation, then wrap and secure the cuff snugly aroundthe subject's bare upper arm. Avoid rolling up the sleevein such a manner that it forms a tight tourniquet aroundthe upper arm. Loose application of the cuff results inoverestimation of the pressure. The lower edge of thecuff should be 1 in (2 cm) above the antecubital fossa(bend of the elbow), where the head of the stethoscopeis to be placed.

5. Place the manometer so the center of the mercurycolumn or aneroid dial is at eye level and easily visible tothe observer and the tubing from the cuff is unobstructed.

6. Inflate the cuff rapidly to 70 mm Hg, and increaseby 10 mm Hg increments while palpating the radialpulse. Note the level of pressure at which the pulsedisappears and subsequently reappears during defla-tion.25 This procedure, the palpatory method, provides anecessary preliminary approximation of the systolicblood pressure to ensure an adequate level of inflationwhen the actual, auscultatory measurement is made.The palpatory method is particularly useful to avoidunderinflation of the cuff in patients with an ausculta-tory gap and overinflation in those with very low bloodpressure.

7. Place the earpieces of the stethoscope into the earcanals, angled forward to fit snugly. Switch the stetho-scope head to the low-frequency position (bell). Thesetting can be confirmed by listening as the stethoscopehead is tapped gently.

8. Place the head of the stethoscope over the brachialartery pulsation, just above and medial to the antecu-bital fossa but below the lower edge of the cuff, and holdit firmly in place, making sure that the head makescontact with the skin around its entire circumference.Wedging the head of the stethoscope under the edge ofthe cuff may free up one hand but results in consider-able extraneous noise.

9. Inflate the bladder rapidly and steadily to a pres-sure 20 to 30 mm Hg above the level previously deter-mined by palpation, then partially unscrew (open) thevalve and deflate the bladder at 2 mm/sec while listeningfor the appearance of the Korotkoff sounds.26

10. As the pressure in the bladder falls, note the levelof the pressure on the manometer at the first appear-ance of repetitive sounds (Phase I) and at the mufflingof these sounds (Phase IV) and when they disappear(Phase V). During the period the Korotkoff sounds areaudible, the rate of deflation should be no more than 2mm per pulse beat, thereby compensating for both rapidand slow heart rates.

11. After the last Korotkoff sound is heard, the cuffshould be deflated slowly for at least another 10 mm Hg,to ensure that no further sounds are audible, and thenrapidly and completely deflated, and the subject shouldbe allowed to rest for at least 30 seconds.

12. The systolic (Phase I) and diastolic (Phase V)pressures should be immediately recorded, rounded off(upwards) to the nearest 2 mm Hg. In children, andwhen sounds are heard nearly to a level of 0 mm Hg, thePhase IV pressure should also be recorded. All valuesshould be recorded together with the name of thesubject, the date and time of the measurement, the arm

on which the measurement was made, the subject'sposition, and the cuff size (when a nonstandard size isused).

13. The measurement should be repeated after atleast 30 seconds, and the two readings averaged. Inclinical situations additional measurements can bemade in the same or opposite arm, in the same or analternative position.25

Systematic errors that observers often make are listedin Appendix A, as are suggestions for improving thetechnique.

Blood Pressure Recording in Special SituationsInfants and Children

Measuring blood pressure in infants and childrenpresents special problems because of their frequent lackof cooperation, although the same techniques are usedas in adults.27-30 Several pediatric cuff sizes are availableand should be selected as indicated (see Appendix D)to ensure that the bladder completely encircles the limb.Because the Korotkoff sounds are often heard throughthe entire period of deflation, determining diastolicpressure as Phase IV is recommended in children lessthan 13 years old. (See "Report of the Second TaskForce on Blood Pressure Control in Children.27") Insmall children and infants, the palpatory method isoften used for approximating systolic pressure, eventhough this may be 5 to 10 mm Hg lower than the levelmeasured by auscultation. In very small infants, theblood pressure is often determined by the flush method,which involves placing a suitable cuff on the arm or leg,raising the limb, and wrapping the extremity distal tothe cuff firmly with an elastic bandage until it is drainedof blood and blanches. The limb is then lowered to heartlevel, the cuff is rapidly inflated, and the bandage isremoved. As the pressure in the cuff is gradually re-duced, flushing of the limb indicates the level at whichflow returns.28 This level corresponds to mean bloodpressure but is inaccurate in infants with anemia, hypo-thermia, or edema. The technique is rarely used now;newer automated oscillometric or Doppler equipmentcan be used instead.31

Elderly PatientsIn elderly patients who have sclerotic, calcified vessels,

it is likely that the systolic pressure is overestimated by theindirect method of measurement. A readily palpable bra-chial artery that can be felt even when the cuff has beeninflated and the blood flow is interrupted (positive Osler'ssign) provides a clue that the measurement may beinaccurate.32 Under such circumstances, an erroneousdiagnosis of hypertension, or "pseudohypertension," maybe made, although this can only be confirmed by directmeasurement. Because postural hypotension is often ob-served in elderly patients, blood pressure measurementshould routinely be made in both the sitting and standingpositions, especially in patients who are labeled hyperten-sive or who are receiving antihypertensive therapy. Be-cause of the tendency for blood pressure to be more labilein elderly patients, it is particularly important to obtainseveral baseline determinations before making diagnosticor therapeutic decisions.

2465




Pregnant PatientsIn pregnant women, a rise in blood pressure or the

diagnosis of hypertension has major significance for theoutcome of the pregnancy for both the mother and thefetus. Measuring the blood pressure of pregnant womenis more complicated because of the wide pulse pressureand the need to record both Phase IV and Phase Vbecause sounds can often be heard to 0 mm Hg. In thethird trimester, especially, the mother's position canaffect blood pressure levels; measurements made withthe woman in the left lateral decubitus position, withthe arm below the level of the heart, are often lowerthan those made in the sitting position with the arm atheart level.33

Obese PatientsA longer and wider cuff is needed for adequate

compression of the brachial artery in the obese patientwith a very large upper arm.18,19,34,35 A large cuff mayalso be required for a big, muscular arm with a promi-nent biceps over which a regular, nontapered cuff mightnot fit smoothly. In both situations it is particularlyimportant to place the center of the bladder over thebrachial artery pulse. If the upper arm is relatively shortdespite the large circumference, it may be difficult to fita standard large adult cuff over the arm. The BritishHypertension Society's recommendation to use a verylong cuff (12.5 x 35 cm) could obviate this problem. Inthe rare patient with an arm circumference greater than50 cm, when even a thigh cuff cannot be fitted over thearm, it is recommended that the health care practitionerwrap the cuff around the patient's forearm and feel forthe appearance of the radial pulse at the wrist. Theaccuracy of this method has not been validated, but itprovides at least a general estimate of the systolic bloodpressure. The error of overestimating the pressure whenmeasuring with a cuff that is too small for an obese armcan be considerable and can lead to misclassification ofan individual as hypertensive and to unnecessary con-cern and therapy.

Miscellaneous ConditionsIn patients who are clinically in shock it may be

difficult to hear the Korotkoff sounds or palpate aperipheral pulse because of generalized vasoconstric-tion. Indirect blood pressure measurements can be veryunreliable in this situation. An approximate estimate ofthe systolic pressure can be made using the palpatorymethod. The direct method, using an intra-arterial line,is preferable under these circumstances.

In patients who have a high cardiac output (eg,thyrotoxic or febrile patients or patients with aorticvalve insufficiency, an arteriovenous fistula, or periph-eral vasodilatation) and in children, the Korotkoffsounds can often be heard to a level close to 0 mm Hg.In such patients the pressure at which Phase IV (muf-fling) occurs should be used as an approximate index ofdiastolic pressure, but both Phase IV and Phase Vpressures should be recorded.

In patients who have recently undergone a mastec-tomy with extensive axillary node dissection or othersurgical procedure involving the arm or shoulder, it isrecommended that blood pressure be measured in theopposite arm. Likewise, in dialysis patients, measuring

blood pressure in the arm with the arteriovenous fistulashould be avoided.

In patients with cardiac dysrhythmias such as atrialfibrillation or frequent premature beats, the systolicpressure may vary widely from beat to beat, so thepressures at which the first and last Korotkoff soundsare heard may differ from measurement to measure-ment. In such situations the rate of deflation should bemore gradual and multiple blood pressure determina-tions should be made and averaged. In patients with aslow heart rate, it is particularly important to reduce thepressure in the cuff gradually (2 mm Hg per heart beat)to avoid underestimation of systolic pressure and over-estimation of diastolic pressure.

Disparity in pressure between the two arms may occurin patients with congenital heart disease, peripheralvascular disease, unilateral neurological and muscu-loskeletal abnormalities, and aortic dissection. Althoughthe blood pressure should routinely be measured in botharms on the initial examination of a hypertensive patient,under the above circumstances this recommendation is ofparticular importance. A consistent difference should benoted, and in hypertensive patients the arm with thehigher pressure should be used for subsequent measure-ments. In patients with stenosis of the subclavian, axil-lary, or brachial artery, the presence of a bruit mayinterfere with interpretation of the Korotkoff sounds.An auscultatory gap, often detected in patients with

high systolic pressure, is not abnormal.'3'36 However, ifthe systolic pressure is not first estimated by palpation,insufficient inflation of the cuff can lead to erroneousdesignation of the lower end of the gap as the systolicpressure.Measurements made during exercise, as during a

treadmill test, are difficult to make and often inaccuratewhen made with currently available equipment. The useof oscillometric or Doppler equipment may provide amore accurate measurement in these situations.

Measuring blood pressure in the lower extremity isindicated in subjects who are suspected of having coarc-tation of the aorta or other types of obstructive aorticdisease. The thigh cuff is used in adults and a large cuffin infants and children. With the subject lying facedown, the cuff is applied with the center of the inflatablebladder over the posterior aspect of the midthigh. Thehead of the stethoscope is placed over the artery in thepopliteal fossa (behind the knee) and the Korotkoffsounds are monitored as the pressure in the bladder islowered, in the same manner as in the arms. If thesubject is unable to lie face down, thigh pressures can betaken with the subject lying on the side or back, with theknee slightly flexed so the stethoscope can be placedover the popliteal pulse easily. The diastolic pressure inthe legs is usually similar to that in the arms, while thesystolic pressure may be 20 to 30 mm Hg higher. Systolicpressure can be approximated in the lower leg bypalpating the posterior tibial or dorsalis pedis pulseswhile inflating and deflating the cuff applied to the calf,with the lower end of the cuff above the malleolus.

Self-Measurement or Home Measurementof Blood Pressure

Self-measurement of blood pressure, usually at home,has become popular, especially among hypertensiveindividuals.37 Self-measurement facilitates patients' par-




ticipation in the health care process and can simplifytitration of antihypertensive drug treatment without theneed for frequent visits to a health care facility. Itinforms both patient and physician of trends or suddenchanges in the level of pressure, eliminates the "white-coat" effect of visiting the doctor's office (see below),and permits more frequent sampling of blood pressure.

Patients, like physicians and nurses, need to be carefullytrained and periodically retrained in the techniques ofmeasurement and recording of blood pressure. The equip-ment must be portable, simple to use, reliable, and easilycalibrated. Manual inflation of the bulb is a form ofisometric exercise that can raise the blood pressure witheach inflation, especially in elderly or debilitated individ-uals. Several automatic and semiautomatic machines havebeen developed to facilitate self-measurement.38,39 Somegive a digital printout of the pressure and pulse and arecompletely automatic and self-inflating. Many are basedon detection of Korotkoff sounds, and others use theoscillometric technique. The newer machines are easier touse than the mercury manometer and stethoscope butrequire periodic recalibration.

Ambulatory Blood Pressure MeasurementsAmbulatory blood pressure measurements refers to

repeated measurements made away from the medicalenvironment with a portable, automatic (self-inflating)recorder in patients engaging in their usual activities.39These measurements are particularly useful in patientsin whom there is a wide disparity between physician-and self-determined pressures, making it difficult toknow which blood pressure level to use for diagnosticand therapeutic decisions. Ambulatory measurementsare also useful for relating fluctuations in blood pres-sure level to symptoms reported by the patient and forassessing the effect of antihypertensive therapy. A num-

ber of miniature, lightweight, silent, fully automaticmachines are now available, but continued validation oftheir accuracy, calibration, and conformity with stan-dards is urged.40,41

In most, but not all, hypertensive patients the bloodpressure obtained in a physician's office is higher thanthe pressures measured during the remainder of theday. If the office pressures are in the range defined ashypertensive, and the pressures during the remainder ofthe day are in the "normal" range, the patient is said tohave "white-coat hypertension."42 Patients who respondto the physician's examination with an exaggerated risein blood pressure tend to have a lesser pressor responseto an encounter with a nurse and may have lower oreven normal blood pressures at other times, as deter-mined by automatic ambulatory blood pressure moni-toring.43 Patients who have white-coat hypertension maybe at risk for being diagnosed as having persistenthypertension and for being overtreated. However, be-cause all existing epidemiological studies of hyperten-sion are based on office blood pressure measurements,further longitudinal research is needed to clarify therole of ambulatory blood pressure monitoring in themanagement of patients with raised arterial pressure.

ConclusionThe above recommendations for the technique of

measuring blood pressure represent the consensus ofthe committee and are based on currently availableinformation and the work of previous committees. Therecommendations are subject to further revision asbetter validation data become available. The intent atthis time is to provide, in the interest of uniformity andconsistency, a standardized technique for measuringblood pressure.

2467




APPENDIX ACommon Problems in Measuring Blood Pressure and Recommendations for Avoiding Them

Problem Result RecommendationEquipmentStethoscope ear pieces pluggedEar pieces poorly fittingBell or diaphragm crackedTubing too long

Mercury manometerMeniscus not at 0 at restColumn not verticalBouncing of mercury withinflation/deflation

Aneroid manometerNeedle not at 0 at rest

Bladder/cuffToo narrow for armToo wide for arm

Inflation systemFaulty valves

Leaky tubing or bulbObserverDigit preference

Cut-off biasDirection biasFatigue or poor memorySubjectArm below heart level

Arm above heart level

Back unsupportedLegs danglingArrhythmiaLarge or muscular armCalcified arteriesTechniqueCuffWrapped too looselyApplied over clothing

ManometerBelow eye levelAbove eye level

Stethoscope headNot in contact with skinApplied too firmlyNot over arteryTouching tubing or cuff

Palpatory pressure omitted

Inflation level too high

Inflation level too low

Inflation rate too slow

Deflation rate too fast

Deflation rate too slow

Poor sound transmissionDistorted soundsDistorted soundsDistorted sounds

Inaccurate readingInaccurate readingInaccurate reading

Inaccurate reading

Blood pressure too highUnable to fit on arm

Inaccurate readingDifficulty inflating and deflating bladderInaccurate reading

Inaccurate reading

Inaccurate readingInaccurate readingInaccurate reading

Reading too high

Reading too low

Blood pressure too highBlood pressure too highBlood pressure level variableBlood pressure reading highBlood pressure reading high

Blood pressure reading too highInaccurate reading

Blood pressure reading too lowBlood pressure reading too high

Extraneous noiseDiastolic reading too lowSounds not well heardExtraneous noiseDanger of missing auscultatory gapUnderestimation of systolic pressurePatient discomfort

Underestimation of systolic pressure

Patient discomfortDiastolic pressure too highSystolic pressure too lowDiastolic pressure too highForearm congestionDiastolic pressure too high

Clean ear piecesAngle ear pieces forwardReplace equipmentLength from ear pieces to bell should be 12 to15 in (30 to 38 cm)

Replace or remove mercuryPlace manometer on level surfaceClean tubing and air vent, replace mercury

Recalibrate

Use cuff length 80% of circumferenceUse regular but longer cuff

Replace equipment

Replace equipment

Be aware of tendency; record blood pressure tonearest 2 mm HgRecord to nearest 2 mm HgRecord to nearest 2 mm HgWrite down reading immediately

Place patient with midpoint of upper arm at heartlevelPlace patient with midpoint of upper arm at heartlevelAvoid isometric exercise during measurementAvoid isometric exercise during measurementMake multiple measurements and averageUse appropriate cuff sizeNote presence of positive Osler sign in record

Rewrap more snuglyRemove arm from sleeve

Place manometer at eye levelPlace manometer at eye level

Place head correctlyPlace head correctlyPlace head over palpated arteryPlace below edge of cuffRoutinely check systolic pressure by palpationfirstInflate to 30 mm Hg above palpatory bloodpressureInflate to 30 mm Hg above palpatory bloodpressureInflate at even rate

Deflate at 2 mm Hg/s or 2 mm Hg per beatDeflate at 2 mm Hg/s or 2 mm Hg per beatDeflate at 2 mm Hg/s or 2 mm Hg per beatCompletely deflate cuff at end of measurement

In patients in whom the Korotkoff sounds are faint and difficult to hear, the following technique may help: Have the subject raise the armover the head and make a fist several times. Inflate the cuff, while the arm is still overhead but the hand relaxed, to a level 50 mm Hg aboveexpected systolic level, have the patient lower the arm rapidly, and measure the blood pressure in the usual manner. Draining the venousblood in this fashion often amplifies the Korotkoff sounds and makes weak sounds, particularly diastolic sounds, more audible.




APPENDIX BRecommendations of the Joint National Committee onthe Diagnosis, Evaluation, and Treatment ofHypertension for Classifying and Defining BloodPressure Levels for Adults (Aged 18 Years and Older)*

Systolic DiastolicCategory (mm Hg) (mm Hg)Normalt <130 <85High normal 130-139 85-89HypertensiontStage 1 (mild) 140-159 90-99Stage 2 (moderate) 160-179 100-109Stage 3 (severe) 180-209 110-119Stage 4 (very severe) .210 >120

*Not taking antihypertensive drugs and not acutely ill. Whensystolic and diastolic pressures fall into different categories, thehigher category should be selected to classify the individual'sblood pressure status. For instance, 160/92 mm Hg should beclassified as Stage 2, and 180/120 mm Hg should be classifiedas Stage 4. Isolated systolic hypertension is defined as systolicblood pressure >140 mm Hg and diastolic blood pressure <90mm Hg and staged appropriately (eg, 170/85 mm Hg is definedas Stage 2 isolated systolic hypertension).

tOptimal blood pressure with respect to cardiovascular risk issystolic blood pressure <120 mm Hg and diastolic blood pres-sure <80 mm Hg. However, unusually low readings should beevaluated for clinical significance.tBased on the average of two or more readings taken at each

of two or more visits following an initial screening.In addition to classifying stages of hypertension based on

average blood pressure levels, the clinican should specify pres-ence or absence of target-organ disease and additional riskfactors. For example, a patient with diabetes and a bloodpressure of 142/94 mm Hg plus left ventricular hypertrophyshould be classified as "Stage 1 hypertension with target-organdisease (left ventricular hypertrophy) and with another major riskfactor (diabetes)." This specificity is important for risk classifica-tion and management.From Reference 2.

References1. An epidemiological approach to describing risk associated with

blood pressure levels: final report of the Working Group on Riskand High Blood Pressure. Hypertension. 1985;7:641-651.

2. The fifth report of the Joint National Committee on Detection,Evaluation, and Treatment of High Blood Pressure (JNCW). ArchIntern Med. 1993;153:154-183.

3. Stamler J, Neaton JD, Wentworth DN. Blood pressure (systolicand diastolic) and risk of fatal coronary heart disease. Hyper-tension. 1989; 13(suppl):I-2-I-12.

4. Assmann G, Schulte H. The Prospective Cardiovascular MunsterStudy: prevalence and prognostic significance of hyperlipidemia inmen with systemic hypertension. Am J Cardiol. 1987;59:9G-17G.

5. MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J,Abbott R, Godwin J, Dyer A, Stamler J. Blood pressure, stroke,and coronary heart disease: Part I. Prolonged differences in bloodpressure: prospective observational studies corrected for theregression dilution bias. Lancet. 1990;335:765-774.

6. Pickering G. Normotension and hypertension: the mysterious via-bility of the false. Am J Med. 1978;65:561-563. Editorial.

7. Sokolow M, Perloff D. The prognosis of essential hypertensiontreated conservatively. Circulation. 1961;23:697-713.

8. Armitage P, Fox W, Rose GA, Tinker CM. The variability ofmeasurements of casual blood pressure: II. survey experience. ClinSci. 1966;30:337-344.

9. Fredrikson M, Matthews KA. Cardiovascular responses to behav-ioural stress and hypertension: a meta-analytic review. Ann BehavMed. 1990;12:30-39.

10. Rosner B, Polk BF. The implications of blood pressure variabilityfor clinical and screening purposes. J Chronic Dis. 1979;32:451-461.

APPENDIX CPrecautions Against Contamination With Mercury

Metallic mercury is an element that is liquid at roomtemperature and tends to break into tiny, highly mobiledroplets when spilled. These droplets vaporize and can con-taminate the atmosphere. In laboratories, offices, and clinicswhere mercury manometers are used regularly, precautionsmust be taken to limit the inhalation, ingestion, or absorptionof mercury by personnel in case of a spill or breakage. It isrecommended that health care personnel refer defective mer-cury manometers for servicing by professionals. In the event ofa mercury spill, the room should be well ventilated and thespilled mercury carefully swept up, not vacuumed, and takento the laboratory for disposal. Personnel involved with theregular use of mercury manometers should be familiar withthe available institutional facilities for handling mercury spills.

APPENDIX DAcceptable Bladder Dimensions (in cm) for Arms ofDifferent Sizes*

ArmBladder Bladder CircumferenceWidth Length Range at

Cuff (cm) (cm) Midpoint (cm)Newborn 3 6 <6Infant 5 15 6-15tChild 8 21 16-21 tSmall adult 10 24 22-26Adult 13 30 27-34Large adult 16 38 35-44Adult thigh 20 42 45-52

*There is some overlapping of the recommended range forarm circumferences in order to limit the number of cuffs; it isrecommended that the larger cuff be used when available.tTo approximate the bladder width:arm circumference ratio of

0.40 more closely in infants and children, additional cuffs areavailable.

11. McCutcheon EP, Rushmer RF. Korotkoff sounds: an experimentalcritique. Circ Res. 1967;20:149-161.

12. Frohlich ED, Grim C, Labarthe DR, Maxwell MH, Perloff D,Weidman WH. Recommendations for human blood pressuredetermination by sphygmomanometers. Circulation. 1988;77:501A-514A.

13. Askey JM. The auscultatory gap in sphygmomanometry. AnnIntern Med. 1974;80:94-97.

14. Prineas RJ, Jacobs D. Quality of Korotkoff sounds: bell vs dia-phragm, cubital fossa vs brachial artery. Prev Med. 1983;12:715-719.

15. Sloan PJM, Zezulka A, Davies P, Sangal A, Beevers M, BeeversDG. Standardized methods for comparison of sphygmoma-nometers. J Hypertens. 1984;2:547-551.

16. Bailey RH, Knaus VL, Bauer JH. Aneroid sphygmomanometers:an assessment of accuracy at a university hospital and clinics. ArchIntern Med. 1991;151:1409-1412.

17. Karvonen MJ, Telivuo LU, Jarvinen EJK Sphygmomanometer cuffsize and the accuracy of indirect measurement of blood pressure.Am J Cardiol. 1964;13:688-693.

18. van Montfrans GA, van der Hoeven GMA, Karemaker JM,Wieling W. Dunning AJ. Accuracy of auscultatory blood pressuremeasurement with a long cuff. Br Med J (Clin Res Ed). 1987;295:354-355.

19. Rastam L, Prineas RJ, Gomez-Marin 0. Ratio of cuff width/armcircumference as a determinant of arterial blood pressure mea-surements in adults. J Intern Med. 1990;227:225-232.

20. Petrie JC, O'Brien ET, Littler WA, de Swiet M. Recommendationson blood pressure measurement. BrMed J (Clin Res Ed). 1986;293:611-615.

2469




21. Curb JD, Labarthe DR, Cooper SP, Cutter GR, Hawkins CM.Training and certification of blood pressure observers. Hyper-tension. 1983;5:610-614.

22. Blood Pressure Measurement: Current Training Programs, Manualsand Instructional Materials. Bethesda, Md: National Heart, Lung,and Blood Institute, Education-Program Information Center;1991.

23. Webster J, Newnham D, Petrie JC, Lovell HG. Influence of armposition on measurement of blood pressure. Br Med J (Clin ResEd). 1984;288:1574-1575.

24. Cushman WC, Cooper KM, Home RA, Meydrech EF. Effect ofback support and stethoscope head on seated blood pressuredeterminations. Am J Hypertens. 1990;3:240-241.

25. Hill MN, Grim CM. How to take a precise blood pressure. Am JNurs. 1991;91:38-42.

26. King GE. Influence of rate of cuff inflation and deflation onobserved blood pressure by sphygmomanometry. Am Heart J.1963;65:303-306.

27. Task Force on Blood Pressure Control in Children. Report of theSecond Task Force on Blood Pressure Control in Children- 1987.Pediatrics. 1987;79:1-25.

28. O'Brien ET, O'Malley K. ABC of blood pressure measurement:infancy and childhood. Br Med J. 1979;2:1048-1049.

29. de Swiet M, Dillon MJ, Littler W, O'Brien E, Padfield PL, PetrieJC. Measurement of blood pressure in children: recommendationsof a working party of the British Hypertension Society. Br Med J(Clin Res Ed). 1989;299:497.

30. Voors AW. Cuff bladder size in a blood pressure survey ofchildren. Am J Epidemiol. 1975;101:489-494.

31. Colan SD, Fujii A, Borow KM, MacPherson D, Sanders SP. Non-invasive determination of systolic, diastolic and end-systolic bloodpressure in neonates, infants and young children: comparison withcentral aortic pressure measurements. Am J Cardiol. 1983;52:867-870.

32. Messerli FH. Osler's maneuver, pseudohypertension, and truehypertension in the elderly. Am J Med. 1986;80:906-910.

33. National High Blood Pressure Education Program Working Groupreport on high blood pressure in pregnancy. Am J Obstet Gynecol.1990;163:1691-1712.

34. Prineas RJ. Measurement of blood pressure in the obese. AnnEpidemiol. 1991;1:321-336.

35. King GE. Errors in clinical measurement of blood pressure inobesity. Clin Sci. 1967;32:223-237.

36. Blank SG, West JE, Muller FB, Pecker MS, Laragh JH, PickeringTG. Characterization of auscultatory gaps with wideband externalpulse recording. Hypertension. 1991;17:225-233.

37. Mejia AD, Julius S, Jones KA, Schork NJ, Kneisley J. TheTecumseh Blood Pressure Study: normative data on bloodpressure self-determination. Arch Intern Med. 1990;150:1209-1213.

38. Hunt JC, Frohlich ED, Moser M, Roccella EJ, Keighley EA.Devices used for self-measurement of blood pressure: revisedstatement of the National High Blood Pressure EducationProgram. Arch Intern Med. 1985;145:2231-2234.

39. Blood pressure monitors. Consumer Reports. 1992;5:295-299.40. National High Blood Pressure Education Program Working Group

report on ambulatory blood pressure monitoring. Arch Intern Med.1990;150:2270-2280.

41. O'Brien E, Petrie J, Littler W, de Swiet M, Padfield PL, O'MalleyK, Jamieson M, Altman D, Bland M, Atkins N. The British Hyper-tension Society protocol for the evaluation of automated and semi-automated blood pressure measuring devices with special ref-erence to ambulatory systems. J Hypertens. 1990;8:607-619.

42. Pickering TG, James GD, Boddie C, Harshfield GA, Blank S,Laragh JH. How common is white coat hypertension? JAMA. 1988;259:225-228.

43. Mancia G, Parati G, Pomidossi G, Grassi G, Casadei R, ZanchettiA. Alerting reaction and rise in blood pressure during mea-surement by physician and nurse. Hypertension. 1987;9:209-215.



Documents

Human blood pressure determination by sphygmomanometry D Perloff… · Perloff et al Human Blood Pressure Determination by Sphygmomanometry are stressed in light of the epidemiological