1
Dual Monitor Protocol Comparing the Accuracy of the Oscar 2 & Spacelabs 90207 24-hr Ambulatory Blood Pressure Monitors V Patteson Lombardi 1 , Patrick C Reichhold 2 , Nicholas R Dietz 2 & Donald L Pate 3 University of Oregon, Biology 1 & Human Physiology 2 & Institute of Neuroscience 3 , Eugene, OR USA 97403 3 2 AHA Council on Hypertension AHA Council on the Kidney in Cardiovascular Disease American Society of Hypertension Joint Scientific Sessions 2017, September 14-17, 2017 Hyatt Regency, San Francisco, CA USA 97411 Background We developed a Dual Monitor Protocol with lab postural challenges to study the accuracy and reliability of ambulatory blood pressure monitors (ABPMs). First, we studied the Accutracker II ABPM (SunTech Medical) and determined in normotensives (n=29), hypertensives (n=12), and alcohol-dependents (n=11) that the monitor significantly underestimated diastolic blood pressure (DBP) and had the most problems with standing DBPs. When we used differences between a mercury (Hg) column and ABPM simultaneous, same arm measurements in the lab to correct for 24-hr BPs, we found that several patients were misclassified. SunTech Medical discontinued manufacturing the auscultatory Accutracker II in 2010. Here we present our initial studies examining the accuracy and reliability of two oscillometric ABPMs, the Oscar 2 (SunTech Medical) and the Spacelabs 90207 (Spacelabs Healthcare). Hypotheses Ultimately, proprietary equations developed for oscillometric ABPMs were derived using auscultatory techniques. Constant pulse pressure assumptions or static equations cannot be applied to dynamic populations. Though mean arterial pressure (MAP) corresponds to the peak pulsatile pressure in the cuff, it is extremely difficult to project outward from MAP to estimate pressures, especially DBP . Thus, we predicted that oscillometric ABPMs also would have difficulties assessing DBPs and/or exhibit a flashlight effect with phase-shift inaccuracies at either the systolic- (SBP) or diastolic-end of the continuum . Methods We programmed Oscar 2 and Spacelabs 90207 ABPMs to record simultaneous, opposite arm BPs for 24 hr in a hypertensive male (62 yr). Clocks were synchronized and cuffs were switched every 2-3 hr during waking hr. In the lab, observers assessed simultaneous same arm BPs for both arms using an Hg column and Thinklabs digital stethoscope with 2nd-generation ambient noise rejection and headset output. ABPMs recorded simultaneous opposite arm BPs interspersed with Hg column measurements. To develop corrections factors for the 24-hr monitoring, all lab measurements were made in triplicate and differences were calculated between observers’ Hg column mean values and each ABPM’s mean values for supine, seated and standing postures. In attempt to examine extreme ends of the pressure continuum, we also used two Spacelabs ABPMs to assess simultaneous and sequential same and opposite arm pressures over 24-hr in a hypertensive male (52 yr) and a single Spacelabs ABPM to assess non-dominant arm pressures in a normotensive male (21 yr). Correction factors established in the lab were applied to 24-hr data. Summary & Conclusions 1. If patients do not match the subject sample used to develop specific nomogram-like prediction equations and/or have widely varying pulse pressures, then they may be inappropriately diagnosed by 24-hr ABPMs. A flashlight effect or blood pressure phase-shift does appear to exist especially with BPs at either end of the continuum. DBPs and SBPs may be inconsistent and unpredictably inaccurate based on the specific monitor used and patients’ pressure ranges, postures and activity levels. 2. Initial data implies that the Oscar II and Spacelabs ABPM equations were developed from subjects who were seated. This may be responsible for the significant ABPM measurement errors relative to observers using an Hg column in the lab. As with the auscultatory ABPM, standing DBPs may be problematic with hypertensives and supine DBPs with normotensives. ABPM proprietary equations developed from a single posture cannot be applied over an entire 24-hr monitoring period when patients move about and assume a variable % distribution of postures. 3. Though we have examined only a few subjects with oscillometric monitors, these limited data do provide some compelling evidence that ABPMs’ estimations may be poor and unpredictable depending upon the specific monitor and patient circumstances. 4. Prediction equations will not improve unless ABPMs can (a) detect unique postures and movements and (b) be designed and calibrated to make sensitivity adjustments based on known variations in K-sound and oscillometric-pulse magnitudes that occur with changes in posture and activity . Both the Oscar and Spacelabs ABPMs exhibit run-away inflations with simple movements like walking or stair climbing . It is likely that motion artifacts are interpreted by ABPMs as not having achieved peak systolic pressures. These continuous inflations may be startling and at times painful for subjects. 5. The Association for the Advancement of Medical Instrumentation (AAMI) and the European Society of Hypertension (ESH International) protocols must require ABPM validation for a variety of postures and activity levels. Additionally, manufacturers should be required to include exact descriptions of age ranges, genders, races and ethnicities examined in developing their proprietary equations. Until these basic steps become mandatory, ABPMs may provide inaccurate measurements that lead to misclassification, inappropriate diagnoses and mismanagement of patients. 1 1 Pre-exercise, Standing, Relaxed Arm Post-exercise, Standing, Relaxed Arm 20 mV 20 mV >/= 2.5 to 3-fold increase in K-sound intensity 2 ? ? ? 3 4 5 6 8 9 10 11 12 13 14 15 4 13 Results thru , & for further Discussion 15 14 Special thanks to Katie Foster and Richard Padgett, MD, of Oregon Heart & Vascular Institute and John Halliwill, PhD, of U of O Human Physiology for their help with ambulatory monitors. 2 13 7

Dual Monitor Protocol Comparing the Accuracy of the Oscar ...€¦ · Background We developed a Dual Monitor Protocol with lab postural challenges to study the accuracy and reliability

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Dual Monitor Protocol Comparing the Accuracy of the Oscar ...€¦ · Background We developed a Dual Monitor Protocol with lab postural challenges to study the accuracy and reliability

Dual Monitor Protocol Comparing the Accuracy of the Oscar 2 & Spacelabs 90207 24-hr Ambulatory Blood Pressure Monitors

V Patteson Lombardi1, Patrick C Reichhold2, Nicholas R Dietz2 & Donald L Pate3

University of Oregon, Biology1 & Human Physiology2 & Institute of Neuroscience3, Eugene, OR USA 97403

3

2

AHA Council on HypertensionAHA Council on the Kidney in Cardiovascular Disease

American Society of HypertensionJoint Scientific Sessions 2017, September 14-17, 2017

Hyatt Regency, San Francisco, CA USA 97411

BackgroundWe developed a Dual Monitor Protocol with labpostural challenges to study the accuracy andreliability of ambulatory blood pressure monitors(ABPMs). First, we studied the Accutracker IIABPM (SunTech Medical) and determined innormotensives (n=29), hypertensives (n=12), andalcohol-dependents (n=11) that the monitorsignificantly underestimated diastolic bloodpressure (DBP) and had the most problems withstanding DBPs. When we used differences betweena mercury (Hg) column and ABPM simultaneous,same arm measurements in the lab to correct for24-hr BPs, we found that several patients weremisclassified. SunTech Medical discontinuedmanufacturing the auscultatory Accutracker II in2010. Here we present our initial studiesexamining the accuracy and reliability of twooscillometric ABPMs, the Oscar 2 (SunTechMedical) and the Spacelabs 90207 (SpacelabsHealthcare).

HypothesesUltimately, proprietary equations developed foroscillometric ABPMs were derived usingauscultatory techniques. Constant pulse pressureassumptions or static equations cannot be appliedto dynamic populations. Though mean arterialpressure (MAP) corresponds to the peak pulsatilepressure in the cuff, it is extremely difficult toproject outward from MAP to estimate pressures,especially DBP . Thus, we predicted thatoscillometric ABPMs also would have difficultiesassessing DBPs and/or exhibit a flashlight effectwith phase-shift inaccuracies at either the systolic-(SBP) or diastolic-end of the continuum .

MethodsWe programmed Oscar 2 and Spacelabs 90207ABPMs to record simultaneous, opposite arm BPsfor 24 hr in a hypertensive male (62 yr). Clockswere synchronized and cuffs were switched every2-3 hr during waking hr. In the lab, observersassessed simultaneous same arm BPs for botharms using an Hg column and Thinklabs digitalstethoscope with 2nd-generation ambient noiserejection and headset output. ABPMs recordedsimultaneous opposite arm BPs interspersed withHg column measurements. To develop correctionsfactors for the 24-hr monitoring, all labmeasurements were made in triplicate anddifferences were calculated between observers’ Hgcolumn mean values and each ABPM’s mean valuesfor supine, seated and standing postures. Inattempt to examine extreme ends of the pressurecontinuum, we also used two Spacelabs ABPMs toassess simultaneous and sequential same andopposite arm pressures over 24-hr in ahypertensive male (52 yr) and a single SpacelabsABPM to assess non-dominant arm pressures in anormotensive male (21 yr). Correction factorsestablished in the lab were applied to 24-hr data.

Summary & Conclusions

1. If patients do not match the subject sample usedto develop specific nomogram-like predictionequations and/or have widely varying pulsepressures, then they may be inappropriatelydiagnosed by 24-hr ABPMs. A flashlight effect orblood pressure phase-shift does appear to existespecially with BPs at either end of the continuum.DBPs and SBPs may be inconsistent andunpredictably inaccurate based on the specificmonitor used and patients’ pressure ranges,postures and activity levels.

2. Initial data implies that the Oscar II andSpacelabs ABPM equations were developed fromsubjects who were seated. This may beresponsible for the significant ABPM measurementerrors relative to observers using an Hg column inthe lab. As with the auscultatory ABPM, standingDBPs may be problematic with hypertensives andsupine DBPs with normotensives. ABPMproprietary equations developed from a singleposture cannot be applied over an entire 24-hrmonitoring period when patients move about andassume a variable % distribution of postures.

3. Though we have examined only a few subjectswith oscillometric monitors, these limited data doprovide some compelling evidence that ABPMs’estimations may be poor and unpredictabledepending upon the specific monitor and patientcircumstances.

4. Prediction equations will not improve unlessABPMs can (a) detect unique postures andmovements and (b) be designed and calibrated tomake sensitivity adjustments based on knownvariations in K-sound and oscillometric-pulsemagnitudes that occur with changes in postureand activity . Both the Oscar and SpacelabsABPMs exhibit run-away inflations with simplemovements like walking or stair climbing . It islikely that motion artifacts are interpreted byABPMs as not having achieved peak systolicpressures. These continuous inflations may bestartling and at times painful for subjects.

5. The Association for the Advancement of MedicalInstrumentation (AAMI) and the European Societyof Hypertension (ESH International) protocolsmust require ABPM validation for a variety ofpostures and activity levels. Additionally,manufacturers should be required to include exactdescriptions of age ranges, genders, races andethnicities examined in developing theirproprietary equations. Until these basic stepsbecome mandatory, ABPMs may provideinaccurate measurements that lead tomisclassification, inappropriate diagnoses andmismanagement of patients.

1

1

Pre-exercise,

Standing,

Relaxed ArmPost-exercise,

Standing,

Relaxed Arm

20 mV20 mV

>/= 2.5 to 3-fold

increase in

K-sound intensity2

?

?

?

3

4

5

6

8

9

10

11

12

13

14

15

4 13Results thru , & for further Discussion1514

Special thanks to Katie Foster and Richard Padgett, MD, of Oregon Heart & Vascular Instituteand John Halliwill, PhD, of U of O Human Physiology for their help with ambulatory monitors.

2

13

7