Funding agencies: This project was supported by a grant from the Institut Universitaire de Cardiologie et de Pneumologie de Québec Foundation. Marie-Eve Leblanc is supported by the Fondation de la Recherche en Sciences Infirmières du Québec and the Ministère de l'Éducation, des Loisirs et du Sports (MELS). Annie Ferland is supported by the Institut de Recherche en Santé du Canada (IRSC). Sara Croteau is supported by the Fonds de la Recherche en Santé du Québec (FRSQ). Paul Poirier is a senior clinical scientist from the FRSQ.
Blood pressure assessment in severe obesity: Validation of a forearm approach
Article first published online: 22 JUN 2013
Copyright © 2013 The Obesity Society
Volume 21, Issue 12, pages E533–E541, December 2013
How to Cite
Leblanc, M.-É., Croteau, S., Ferland, A., Bussières, J., Cloutier, L., Hould, F.-S., Biertho, L., Moustarah, F., Marceau, S. and Poirier, P. (2013), Blood pressure assessment in severe obesity: Validation of a forearm approach. Obesity, 21: E533–E541. doi: 10.1002/oby.20458
- Issue published online: 3 DEC 2013
- Article first published online: 22 JUN 2013
- Accepted manuscript online: 20 MAR 2013 02:17AM EST
- Manuscript Accepted: 6 MAR 2013
- Manuscript Received: 6 OCT 2012
Obesity is frequently associated with systemic hypertension. Blood pressure measure is inaccurate in severely obese patients because of poor cuff size fitting. The aim of the study is to assess the degree of agreement between the intra-arterial method as the gold standard vs. noninvasive methods, i.e., forearm blood pressure and upper-arm blood pressure measures.
Design and Methods
A total of 1285 measures of intra-arterial and forearm blood pressure were taken in 51 severely obese patients in a supine position in the operating and the recovery room. A subset of 352 upper-arm measures were taken in the recovery room and compared to the intra-arterial and the forearm methods.
Correlation between the intra-arterial and the forearm measures was 0.90 (P < 0.001) for the 2570 data (systolic and diastolic). Compared to intra-arterial, the forearm method overestimated systolic (6 ± 16 mm Hg, P < 0.001) and underestimated diastolic blood pressure (2 ± 11 mm Hg, P = 0.03). Compared to intra-arterial, upper-arm underestimated systolic (8 ± 16 mm Hg, P < 0.01) and overestimated diastolic blood pressure (9 ± 7 mm Hg, P < 0.001).
The magnitude of differences between the intra-arterial and forearm method was less than differences between the intra-arterial and upper-arm method. Our results suggest that forearm method may be a more accurate alternative to upper-arm measurement to assess blood pressure in severely obese patients.
Between 1986 and 2000, the prevalence of a BMI over 30, 40, and 50 kg/m2 have doubled, quadrupled, and quintupled, respectively, in the United States. Obesity is defined as severe at a BMI of ≥40 kg/m2 or ≥35 kg/m2 with the presence of comorbidities like type 2 diabetes, dyslipidemia, or systemic hypertension [1, 2]. Hypertension is one of the most common comorbidity associated with severe obesity . In order to manage hypertension, it is mandatory to measure blood pressure correctly [4-6]. However, in a severely obese population, accurate blood pressure measurement can be problematic given that upper arms are frequently short, large and conical. This is important since severe obesity represents ∼5% of the US population . More so, given the prevalence of severely obese patients, cardiologists, surgeons, anesthesiologists, and other healthcare professionals will be confronted more often with proper cardiovascular evaluation in this population . Additionally, hypertension guidelines do not extensively discuss blood pressure measurement in obese patients [4-6, 9].
The gold standard for accurate and reliable blood pressure measurement is invasive intra-arterial blood pressure measurement [10, 11]. Intra-arterial measurement is applied almost exclusively in the operative room, in critical care units or to confirm a diagnosis of pseudo-hypertension. Few studies used intra-arterial blood pressure as the gold standard reference to validate forearm blood pressure measurement among severely obese patients. In 1956, auscultatory forearm blood pressure measurement was compared to the intra-arterial method in six severely obese subjects . Authors have shown a systolic overestimation of 8 mm Hg and a diastolic underestimation of 9 mm Hg. In 1965, forearm and standard upper-arm auscultatory blood pressure measurements were performed in 76 middle-aged men without description of subject's BMI . Forearm approach underestimated systolic, whereas diastolic blood pressure was overestimated compared to the intra-arterial method. Another study, performed with a mercury manometer in 98 nonobese women reported that upper-arm systolic values were higher and diastolic values were lower compared to the forearm . Between 1999 and 2010, seven studies were conducted using the oscillometric method with cuff/bladder positioned on the upper arm to validate forearm blood pressure measurement [15-21]. These studies showed a general trend toward the forearm measurements to overestimate both systolic and diastolic values compared to upper-arm with a statistically significant correlation between methods. Four studies included obese participants (BMI > 30 kg/m2) [15, 17, 18, 21], whereas the three others did not report BMI [16, 19, 20]. None reported severely obese patients separately. Findings revealed discrepancies between forearm vs. upper-arm blood pressure methods without reporting on conformity with the Association for the Advancement of Medical Instrumentation (AAMI) regulations .
The aim of this study is to assess the degree of agreement between the intra-arterial method taken as the gold standard vs. the noninvasive method of forearm blood pressure measurement and the noninvasive usual upper-arm method.
Fifty-one severely obese adults undergoing a biliopancreatic diversion with duodenal switch surgery were enrolled in the study. Men and women were ≥18 years old with a BMI ≥40 or ≥35 kg/m2 with comorbidities. Volunteers were excluded if they had significant peripheral arterial disease, resulting in at least a 20 mm Hg blood pressure difference between both arms. The severely obese patients were separated in two groups. The first group (G1, n = 26) had their blood pressure measured with the intra-arterial line and the noninvasive forearm method in the operating and the recovery rooms. The upper-arm blood pressure was not measured in G1 because it was not technically possible to do so in the operating room. The second group (G2, n = 25) had their blood pressure measured in the recovery room only using noninvasive usual upper-arm and forearm methods on the same arm in the addition to the intra-arterial blood pressure taken at the contra-lateral arm.
The protocol was approved by the ethic committee of the Institut Universitaire de Cardiologie et de Pneumologie de Québec (IUCPQ). Each participant provided written informed consent before inclusion in the study.
Patients were admitted to IUCPQ one day before or on the same day of their bariatric surgery. Baseline blood samples were drawn after a 12-hours overnight fast to determine plasma glucose and lipid profile. Body weight (kg) and height (cm) were measured on calibrated scales: a Detectomedical scale (Brooklyn, USA) for weight <140 kg or a Toledo-scale (Model 4181-8140, Canada) for weight ≥140 kg. BMI was calculated as the weight divided by the height in square meter (kg/m2). Arm circumference was measured at mid-point between the shoulder and the olecranon in centimetres. Forearm circumference was measured at mid-point between the olecranon and the styloid apophysis in centimetres. Baseline blood pressure measurements were performed according to the American Heart Association (AHA) hypertension guidelines  with a validated oscillometric device recommended on dablEducational website (Welch Allyn, Life sign monitor 5200 Series NY)  in a sitting position after at least a 5-min resting period. Different cuff sizes were used according to the participant's arm circumference and defined as follow: standard cuff (23-33, 25-35, or 21-31 cm), large cuff (31-40 cm), and very large cuff (33-47 cm).
Blood pressure measurements
Intra-arterial blood pressure
The intra-arterial blood pressure measurement was considered the reference method. The intra-arterial line was inserted in the radial artery before surgery on a routine basis by an anesthesiologist.
Upper-arm blood pressure
Upper-arm blood pressure was measured at the contra-lateral arm (Welch Allyn, Life sign monitor 5200 Series NY) of the intra-arterial line with the appropriate cuff size . The intra-arterial measures were taken simultaneously with the upper-arm method with at least 5-min interval between each measurement in the supine position in alternating sequence, i.e., upper-arm/intra-arterial, forearm/intra-arterial, etc. Upper-arm and forearm blood pressures were not taken simultaneously because cuffs were placed on the same side.
Forearm blood pressure
We arbitrarily determined forearm cuff position at 3 cm from the styloid process of the ulna, palpable at the wrist. Forearm blood pressure was measured at the contra-lateral forearm (Welch Allyn, Life sign monitor 5200 Series NY) of the intra-arterial line. Intra-arterial and forearm blood pressure assessments were performed simultaneously with at least 5-min interval between measurements in the supine position.
Data are reported as mean ± standard deviation (SD) unless stated otherwise. An unpaired t-test for independent samples was used to assess differences in baseline variables between groups. A chi-square analysis was performed to determine differences between groups in categorical variables. Pearson coefficient correlation was calculated to assess the strength of the relationship between systolic and diastolic blood pressures for methods compared two-by-two: intra-arterial vs. forearm, intra-arterial vs. upper-arm, and upper-arm vs. forearm blood pressure and expressed with a squared correlation (r2). As a result of the uneven number of blood pressure measurements per patient, the first three blood pressure measurements were averaged to get a comparable number of values in each group (G1, 26 data in the operating room; G2, 25 data in the recovery room). Differences between the mean of the first three readings were calculated for systolic/diastolic blood pressure and compared with a paired t-test. A Bland-Altman graphic was drawn to assess the relationship between methods. Straight dark line represents the mean difference between methods and the dotted line represents the limits of agreement (mean ± 2 SD) [24, 25]. A repeated measures ANOVA model was used to estimate the between and the within-participants variance proportion on the difference of systolic and diastolic blood pressure measurement differences for the 1285 sets of intra-arterial and forearm blood pressure measurements. This model was also used to explain the potential bias between methods. The proportion of within-participants variance, explained by identifiable factors, was defined with the squared correlation (r2) between the measured and the predicted values of the model. A P ≤ 0.05 (two-sided) was considered significant. Analyses were conducted using SPSS (version 19.0) and SAS (version 9.2) softwares.
Characteristics of the severely obese patients are described in Table 1. An average of 23 sets (585 data) of blood pressure measurements was taken with the intra-arterial and forearm methods for the G1 in the operating room only and an average of 14 sets (700 data) was taken for each individual in G1 and G2 in the recovery room. Regarding the upper-arm method, an average of 14 sets (352 data) of blood pressure measurements was taken with the intra-arterial and the forearm method in the recovery room.
Differences between intra-arterial vs. forearm blood pressure
The 2570 intra-arterial and forearm data (1285 systolic and 1285 diastolic blood pressure measurements) were closely correlated (r2 = 0.90, P < 0.001) (Figure 1A). Intra-arterial and forearm systolic values ranged from 70 to 208 mm Hg and from 78 to 227 mm Hg, respectively; whereas diastolic values extended from 31 to 119 mm Hg and from 23 to 111 mm Hg, respectively. When we considered blood pressure measured in the recovery room only, the 704 intra-arterial and forearm data (352 systolic, 352 diastolic) were slightly more closely correlated (r2 = 0.92, P < 0.001). The first three measurements were averaged for patients in G1 and G2. Compared to the intra-arterial readings, the forearm method overestimated systolic blood pressure by 6 ± 16 mm Hg (P < 0.001) and underestimated diastolic blood pressure by 2 ± 11 mm Hg (P = 0.03) (Table 2). Better agreement between methods using Bland-Altman representation for diastolic than for systolic was observed in the operating room (Figure 2A and B) and in the recovery room (Figure 2C and D). Overall, the agreement between intra-arterial and forearm methods was better in the recovery room. Data dispersion remained within the 95% confidence interval for systolic and diastolic blood pressure in all analyses.
Differences between upper-arm vs. forearm blood pressures
The subset of 704 upper-arm and forearm data were closely correlated (r = 0.89, P < 0.001) (Figure 1B). Upper-arm systolic and diastolic values ranged from 102 to 217 mm Hg and from 57 to 109 mm Hg, respectively. Forearm systolic and diastolic values paired with upper-arm method extended from 102 to 227 mm Hg and from 63 to 117 mm Hg, respectively. The first three measurements were averaged for patients in the recovery room. Compared to the upper-am reference, the forearm method overestimated systolic blood pressure by 13 ± 14 mm Hg (P < 0.001) and underestimated diastolic blood pressure by 4 ± 7 mm Hg (P < 0.001) (Table 2). The limits of agreement between methods are represented with a Bland-Altman graphic for systolic and diastolic (Figure 3A and B). Data dispersion remained within a 95% confidence interval for systolic and diastolic in all analyses.
Differences between upper-arm vs. intra-arterial blood pressures
The subset of 704 intra-arterial and upper-arm data were closely correlated (r = 0.89, P < 0.001) (Figure 1C). Intra-arterial systolic and diastolic values paired with upper-arm readings ranged from 102 to 217 mm Hg and from 63 to 117 mm Hg, respectively. Compared to the intra-arterial reference, the upper-arm method underestimated systolic blood pressure by 8 ± 16 mm Hg (P < 0.001) and overestimated diastolic blood pressure by 9 ± 7 mm Hg (P < 0.001) (Table 2). Again, data dispersion remained within a 95% confidence interval for systolic and diastolic in all analyses.
Potential bias for differences between sets of intra-arterial and forearm blood pressure measurements
The repeated measures ANOVA model used to assess the intra-arterial and forearm blood pressure measurement differences between methods showed that potential bias for systolic and diastolic blood pressure measures were mainly related to within-participant factors with 63% and 60% compared to between-participant factors with 39% and 40% of total variance for systolic and diastolic, respectively. Three within-participant factors were identified as potential bias influencing the progression of differences between intra-arterial and forearm methods: 1) the number of repeated blood pressure measurements over time, 2) the influence of the location where blood pressure was performed (operating or recovery room) and, 3) the impact of blood pressure level variability.
Number of blood pressure
The r2 for the number of repeated blood pressure measurements taken over time was 0.04 and 0.25 for systolic and diastolic, respectively (all P < 0.001), regardless of the location.
The progression of differences between methods occurring with the number of repeated blood pressure measurements taken over time shown a statistically significant progression for systolic and diastolic blood pressures (P < 0.0001) in the operating room (Figure 4A and B, dotted line) in contrast to the recovery room where it was unchanged for systolic (P = 0.813) and diastolic (P = 0.157) blood pressure measurements (Figure 4A and B, solid line).
Blood pressure level variability
The r2 for the systolic and diastolic blood pressure levels was 2% and 1%, respectively, regardless of the location (all P < 0.05). Taking into consideration the stability of systolic and diastolic differences between intra-arterial and forearm methods over blood pressure measurements throughout time, systolic and diastolic blood pressure measurements were divided in quartiles based on the intra-arterial values in the recovery room only, including 700 blood pressure measures (data not shown). The agreement for the first (≤114 mm Hg) and the fourth quartile (≥157 mm Hg) of systolic blood pressure showed the lowest mean difference between methods with intra-arterial blood pressure overestimation of 6 ± 4 and 3 ± 2 mm Hg, with the forearm methods, respectively. For diastolic blood pressure, the lowest difference was showed for the lower quartile only (≤61 mm Hg), with an intra-arterial blood pressure overestimation of 5 ± 3 mm Hg.
This study is, to our knowledge, the first to validate forearm blood pressure measurement taken simultaneously using an oscillometric device vs. an intra-arterial method in severely obese patients in the operating and the recovery room. Our main findings were related to the degree of agreement between the forearm vs. the intra-arterial methods, which showed that both methods are significantly related (r2 = 0.90, P < 0.001). The overall trend in all patients was that the forearm method overestimates systolic and underestimates diastolic blood pressure compared to intra-arterial. When the forearm method was compared to upper-arm blood pressure measurements, the same trend was observed for systolic and diastolic blood pressure albeit greater than the differences between the forearm and the intra-arterial methods in both rooms. Differences between all methods taken two-by-two all exceeded the AAMI criteria  excepted for diastolic blood pressure measured in the recovery room with intra-arterial and forearm methods with differences ≤ 5 ± 8 mm Hg (intra-arterial overestimation of 5 ± 7; P < 0.01).
Better agreement is observed regarding blood pressure measured in the recovery room rather than in the operating room, possibly because of greater blood pressure fluctuation in the operating room. Within-participant factors explained 63% and 60% of the total variance for systolic and diastolic blood pressure differences between intra-arterial and forearm methods. The within-participants variables were accounted for most of the factors explaining differences between the two blood pressure measurement techniques. Other within-participants variables related to intra-arterial (blood clots dampening the arterial catheter, patient positioning changes from fowler to Trendelenburg position), to the oscillometric technical detection (cuff sliding toward the wrist), to the intravenous medication given to patients during and following the surgery or factors related to the awakening process, which are independent of each other and change over time, may explain our findings.
This validation study of a forearm blood pressure measurement method is clinically relevant taking into consideration the limitations related to blood pressure assessment in severely obese patient, especially when the appropriate upper-arm cuff size may not be readily available. Study performed with a similar population and design has shown that wrist and intra-arterial blood pressure method could not be interchanged . Also, important local upper-arm adiposity or excess adipose tissue in the axilla can cause artefacts during readings. Using a smaller cuff than needed will provide falsely high blood pressure readings that may result in an unnecessary administration of antihypertensive medications . Some investigators have tried to examine this issue 60 years ago by investigating the intra-arterial vs. the auscultatory methods and have strongly recommended the use of the intra-arterial method in obese individual, while also reporting imprecise and useless indirect blood pressure measurements [28, 29]. Intra-arterial, auscultatory and oscillometric methods can provide different readings in the same individual if methods are used simultaneously. The source of variation can be associated with mechanisms inherent to each method and/or involves observer, patient, or equipment features . As observed in our study, intra-arterial blood pressure measurement records precisely blood waves whereas the cuff used with the noninvasive auscultatory method compresses the artery wall and usually underestimated the true systolic value . It was shown in critically ill obese patient (BMI ∼ 34 kg/m2) that both upper-arm oscillometric and auscultatory methods also underestimate the intra-arterial systolic value .
Validation organizations suggest instructions using the auscultatory with mercury and cuff/stethoscope method as the gold standard and have discouraged the use of intra-arterial monitoring as a reference [22, 33, 34]. Consequently, most oscillometric devices are now validated with the auscultatory method using mercury manometer and it is, therefore, difficult to know how different would have been our results if the intra-arterial method would have been used to assess external validity. In our study, we used a validated oscillometric method, which meets the criteria of the AAMI protocol related to instrument validation with a mercury manometer only, but not with an intra-arterial approach . Recently, Smulyan and Safar  have criticized the approach of validation of oscillometric devices with a mercury manometer. They emphasized that the AAMI protocol was originally developed based on the results of five studies that validated the cuff/stethoscope vs. the intra-arterial method with an average difference ranging from 0.9 to 12.3 mm Hg for systolic and from 1.3 to 13 mm Hg for diastolic blood pressures . Nevertheless, despite slight disagreements between the sets of intra-arterial and forearm data, we consider that the mean difference for systolic and diastolic blood pressure measurements are clinically acceptable taken as a whole.
At the present time, there are no guidelines regarding where exactly to position the cuff on the forearm for blood pressure measurement in the literature. We standardized forearm cuff positioning at 3 cm of the styloid process as no distinction is often made between wrist and forearm location in studies [36, 37]. Our study was performed in a clinical setting different from the usual blood pressure assessment performed on the majority of individuals as the patients in our study had clinically severe obesity and were studied during their bariatric surgery. They were in the supine position receiving intravenous medications responsible for blood pressure fluctuations. For a better comparison of the noninvasive usual upper-arm and forearm methods, blood pressure measurements should be performed simultaneously rather with an elapsed time between methods. This short delay could not explain all differences between upper-arm and forearm readings, but is a factor to consider. There is no way to measure upper-arm and forearm blood pressures adequately simultaneously on the same arm. We have chosen a 5-min delay between blood pressure measurements (upper-arm vs. forearm) to eliminate a potential carry over effect. Indeed, even if the upper-arm and forearm blood pressure methods would have been used simultaneously, a major problem remains: cuff fitting. For the majority of severely obese patients, the upper-arm cuff was adjusted with difficulty during testing relatively to the extreme upper-arm circumference (∼40 cm) as well as the conical upper-arm morphology. This situation reflected the actual and problematic usual practice that we tried to address through our study design.
We show that the mean difference between intra-arterial and forearm blood pressure measurement methods slightly exceeds limits set by guideline recommendations regarding validation of a new method. Intrinsic pitfalls related to the intra-arterial and oscillometric methods may be responsible for the slight difference encountered between methods. Looking at the upper-arm vs. forearm blood pressure measurement methods, no guidelines related to the instrument validation provide information regarding severely obese upper-arm as a reference, because none has included extreme arm circumference or severely obese individual in their methodology [22, 33, 34]. On the other hand, we found that the magnitude of differences between intra-arterial and the forearm method was less than differences between intra-arterial and the usual upper-arm methods. We also found that the differences between intra-arterial and the forearm methods were stable over time when blood pressure measurements were performed in the supine position in the recovery room. Our results suggest that the noninvasive forearm blood pressure method may be used as a more accurate alternative especially when usual upper-arm blood pressure measurement is technically challenging.
The authors would like to thank all patients participating in the study, Mr. Serge Simard for his help with the statistical analysis and Ms. Jocelyne Bellemare in the preparation of this manuscript.
- 6Recommendations for blood pressure measurement in humans and experimental animals: part 1: blood pressure measurement in humans: a statement for professionals from the subcommittee of Professional and Public Education of the American Heart Association Council on high blood pressure research. Circulation 2005;111:697-716., , , et al.
- 7On behalf of the American Heart Association Obesity Committee of the Council on Nutrition, Physical Activity and Metabolism. bariatric surgery and cardiovascular risk factors: a scientific statement from The American Heart Association. Circulation 2011;123:1683-1701., , , et al.
- 12Measurement of blood pressure in obese persons. JAMA 1956;162:970-971., , .
- 20A comparison of forearm and upper arm blood pressure measurements in a sample of healthy young adults. J Undergrad Nurs Scholarsh 2009;11:1-9., , , et al.
- 22American National Standard Non-Invasive Sphygmomanometers—Part 2: Clinical validation of automated measurement type. ANSI/AAMI/ISO 81060-2: 2009;1-38.
- 30Calibration of automated non invasive blood pressure measurement devices. In: Mukhopadhyay SC, Lay-Ekuakille A, editors. Advances in Biomedical Sensing, Measurements, Instrumentation and Systems. Vol 55: Berlin Heidelberg: Springer; 2010:281-304.., .
- 33The British Hypertension Society protocol for the evaluation of blood pressure measuring devices. J Hypertens 1993;11(Suppl 2):S43-S62., , , et al.