Oscar 2 & Spacelabs 90207/90217 Ambulatory Blood Pressure Monitors Differ from Each Other & Trained Clinicians Using a Hg Column in the Lab

V Patteson Lombardi1, Patrick C Reichhold2, Jennifer L Cramer2, Hannah P Harkness3, Natalie J DeBell2,1, Nicholas R Dietz2,1 & Donald L Pate4. University of Oregon, Biology1, Human Physiology2, General Science3 &

Institute of Neuroscience4, Eugene, OR USA 97403-1210AHA Council on Hypertension

AHA Council on the Kidney in Cardiovascular DiseaseHypertension Scientific Sessions 2019

Marriott New Orleans, New Orleans, LA USA 70130September 5-8, 2019

Background: In 1990, we developed a Dual MonitorProtocol with simultaneous, same arm blood pressure(BP) measurements and postural challenges to test theaccuracy and reliability of 24-hr ambulatory bloodpressure monitors (ABPMs). ABPM differences withmercury (Hg) column BPs in the lab were used toadjust for ABPM inaccuracies during 24-hr field tests.We found in 15 normotensives, 14 hypertensives and11 primary alcohol-dependents that auscultatory,Accutracker II ABPMs had excessive inter-devicevariability, far lower levels of agreement thanobservers using a Hg column and problems assessingdiastolic (DBP) and systolic pressures (SBP) thatprovoked patient misclassifications (up to 73% DBP, upto 20% SBP incorrectly classified).

Purpose: Our goals were to determine the accuracyand reliability of popular oscillometric ABPMs, theOscar 2 (SunTech Medical, Morrisville, NC) andSpacelabs 90207/90217 (Spacelabs Healthcare,Snoqualmie, WA), by using a modification of our DualMonitor Protocol.

Hypotheses: Oscillometric ABPMs ultimately relyupon auscultation to derive sample-specific, staticequations similar to a nomogram. Like a beamprojecting out from a flashlight, mean arterial pressure(MAP), estimated from peak cuff pressure oscillations,is used to predict SBP and DBP. As with auscultatoryABPMs, we predicted that the Oscar and Spacelabsoscillometric ABPMs would have a lower level ofagreement compared to observers using a Hg column,making them susceptible to measurement errors andpatient misclassifications.

Methods:

Baseline: In the lab, simultaneous, same arm BPswere assessed in 10 subjects by two observers usingan Hg column and Thinklabs digital stethoscope withambient-noise reduction. Observers’ Hg column BPsalternated with simultaneous, opposite arm BPs byOscar and Spacelabs ABPMs. To minimize potentialarm differences, ABPM cuffs were switched andsimultaneous BPs were repeated on opposite arms. Allmeasurements were in triplicate on both arms at heartlevel with observers blinded to each other’s results andthose of the monitors. A Latin square design was usedto minimize order bias and to ensure that eachtechnique measured BPs in the 1st, 2nd and 3rd

position.

Postural: Eight of 10 subjects had non-dominant armBPs measured alternately by Hg column, Oscar andSpacelabs ABPMs in triplicate for supine, seated andstanding, relaxed arm conditions.

Summary & Conclusions:Oscar and Spacelabs oscillometric ABPMs are erratic in estimating individual BPs, have a low level of agreement and can differ clinically and statistically from observers using a Hg column in acontrolled laboratory setting. Over 70% (Spacelabs) and 90% (Oscar) of SBPs can be overestimated by 5 or more mm Hg, misclassifying patients, especially those near category thresholds. Theseinaccuracies become increasingly important with new US guidelines shifting the definition of hypertension to lower levels. Over- and underestimations vary with posture. Our results, though in a smallnumber of mostly (70%, 7/10) normotensive subjects, do not provide confidence in 24-hr measurements during a time when patients are active and assume multiple postures. At the very least, USand international protocols for validating ABPMs must develop more stringent standards and require postural testing. Until then, we have difficulty putting our trust in ambulatory BP monitors.

We are grateful to the American Foundation for Hypertension Research & Education

for giving us a travel award in 2018 and to Ashley Clark and the American Heart

Association staff for their continued help and dedication to conference participants.

For further information about protocols & results, please contact V. Pat Lombardi at lombardi@uoregon.edu

Special thanks to Katie Foster, Laura Emmons, Donna Wilcox and Richard Padgett, MD, of the Oregon Heart & Vascular Institute and John Halliwill, PhD, of the U of O

Department of Human Physiology for their help with ambulatory monitors.

Results:Despite the controlled laboratory setting, the Oscar overestimated 90% (9/10) and the Spacelabs 70% (7/10) of observers’ SBPs by > 5 mm Hg (Table 1). The ABPMs either overestimated (both Oscarand Spacelabs 30%, 3/10) or underestimated (Oscar 10%, 1/10; Spacelabs 20%, 2/10) observers’ DBPs. The Oscar misclassified 20% (4/20) of BPs based on AHA-ACC (2017) Ø and 25% (5/20)based on ESH (2018) Ø guidelines. The Spacelabs misclassified 10% (2/20) of BPs using AHA-ACC Ø and 20% (4/20) using ESH Ø guidelines. The Oscar overestimated the observers’ Hg column SBPby 8.8 mm Hg, with 95% of measurements (~ 2 SD) 24.2 mm Hg above to 6.5 mm Hg below the observers (30.7 mm Hg range) (Figure 1). The Oscar overestimated the observers’ DBP by 1.2 mmHg, but exhibited extreme variability with 95% of DBPs 17.8 mm Hg above to 15.4 mm Hg below observers (33.2 mm Hg range) (Figure 2). The Spacelabs overestimated the observers’ SBP by 5.4mm Hg, with 95% of SBPs 19.9 mm Hg above to 9.0 mm Hg below the observers (28.9 mm Hg range) (Figure 3). The Spacelabs overestimated the observers’ DBP by 1.4 mm Hg, but with widevariability as 95% of DBPs were 19.4 mm Hg above to 16.6 mm Hg below observers (36.1 mm Hg range) (Figure 4). The Oscar overestimated the Spacelabs (or Spacelabs underestimated the Oscar)SBP by an average of 3.4 mm Hg with 95% of Oscar SBP measurements 18.1 mm above to 11.3 mm Hg below to the Spacelabs (29.4 mm Hg range) (Figure 5). The Oscar underestimated theSpacelabs DBP by less than 1 mm Hg (0.2 mm Hg), but for 95% of DBPs was 10.3 mm Hg below to 9.9 mm Hg above the Spacelabs (20.2 mm Hg range) (Figure 6). The Oscar overestimated theObservers’ SBP by 3.5 mm Hg (95% CI: 0.7 to 6.4 mm Hg) for supine (P < 0.05), 4.7 mm Hg (95 CI: 1.6 to 7.8 mm Hg) for seated (P < 0.01) and 3.8 mm Hg (95% CI: 0.6 to 8.2 mm Hg) forstanding (tendency, P < 0.10) conditions (Figure 7). The Oscar overestimated the Observers’ DBP by 6.5 mm Hg (95% CI: 2.4 to 10.7 mm Hg) for supine (P < 0.01), but underestimated theObservers’ DBP by 4.9 mm Hg (95% CI: 0.6 to 9.1 mm Hg) for standing (P < 0.05) conditions. For supine measures, the Spacelabs overestimated (P < 0.05) the Observers’ DBP by 6.4 mm Hg (95%CI: 1.2 to 11.7 mm Hg) (Figure 8). For standing measures, the Spacelabs underestimated (P < 0.05) the Observers’ DBP by 5.8 mm Hg (95 CI: 1.1 to 10.4 mm Hg). Note the variability for bothABPMs based on posture and the stepwise progression of DBP differences from supine to seated to standing, indicating static, nomogram equations that do not adjust for posture.

Figure 1. Observers Hg column vs Oscar 2 SBP differences for baseline lab tests.

Figure 2. Observers Hg column vs Oscar 2 DBP differences for baseline lab tests.

Figure 3. Observers Hg column vs Space-labs SBP differences for baseline lab tests.

Figure 4. Observers Hg column vs Space-labs DBP differences for baseline lab tests.

Figure 5. Oscar 2 vs Spacelabs SBP differences for baseline tests.

Figure 7. Observers Hg column vs Oscar 2 BP differences for postural tests.

Figure 6. Oscar 2 vs Spacelabs DBP differences for baseline tests.

Figure 8. Observers Hg column vs Space-labs BP differences for postural tests.