Monitoring systemic arterial blood pressure is considered the standard of care for assessing anesthetized patients.1 The criterion-referenced standard for measuring blood pressure involves placement of an intra-arterial catheter that is connected to a transducer to measure arterial pulsations.2,3 This method causes pain to patients and is technically challenging and impractical in many clinical situations.4 For those reasons, noninvasive equipment such as oscillometric units or Doppler ultrasonography detectors are commonly used. In clinical settings, these indirect blood pressure measurements are often used when deciding whether therapeutic intervention for hypertension or hypotension is necessary. For that reason, it is imperative that a noninvasive device provides accurate estimates of arterial blood pressure.
Numerous studies5–12 have been conducted during the past 11 years in an attempt to clarify the relationship between results for oscillometric devices and direct blood pressure measurement. Some of these studies yielded conflicting results. For example, investigators of 1 study5 found that noninvasive blood pressure measurements obtained from a forelimb or hind limb of normotensive patients had poor agreement and underestimated direct blood pressure measured at a cranial tibial artery. Investigators of another study6 found that SAP and MAP noninvasively measured at the forelimb of normotensive patients more closely agreed with blood pressures measured invasively at the lingual and dorsal pedal arteries than with DAP. Although methods differed among the previous studies, a common theme is the use of a traditional cylindrical blood pressure cuff.
In veterinary medicine, most blood pressure cuffs used for canine and feline patients are modeled on blood pressure cuffs intended for humans. These blood pressure cuffs were designed to fit the relatively cylindrical structure of the brachium of humans and, hence, their rectangular shape. Evaluation of the canine antebrachium reveals that the anatomic structures do not resemble a cylinder. In many dogs, the proximal portion of the antebrachium is of greater diameter than is the distal portion of the antebrachium, which results in a conical, tapering antebrachium.13
In human medicine, a conical cuff has been developed for patients with a tapering bicipital area, which is typically found in bariatric patients. In a study14 of human subjects with an arm circumference of 37.5 to 42.5 cm, 15% of individuals initially identified as hypertensive were normotensive when assessed with a conical cuff. That study14 helped stimulate the development of a conical blood pressure cuff for use in bariatric patients. A conical cuff offers certain theoretical benefits, such as better fit on a tapering antebrachium, less slippage during inflation, and more uniform inflation. These benefits should lead to better arterial compression, compared with that obtained by use of a cylindrical cuff. For a blood pressure cuff to provide accurate measurements, it must adhere to the antebrachial area and apply uniform force on the forelimb.15 However, when a cylindrical cuff is placed on a tube-shaped structure, there is a gradual decrease in the pressure on the tissue when moving from the center of the cuff to the edge of the cuff.15 In canine and feline subjects with conical antebrachial areas, the gap created at the distal end of a cylindrical blood pressure cuff could unpredictably exacerbate this pressure change, which could lead to errors in blood pressure measurement.14
The purpose of the study reported here was to compare the ability of an anatomically modified conical blood pressure cuff and a cylindrical oscillometric blood pressure cuff to estimate blood pressure measured invasively at the median sacral artery. The median sacral artery was chosen on the basis of results of a recent study16 that suggest the median sacral and superficial palmar arch arteries have the best agreement for validating indirect blood pressure systems. We believed that the failure of some investigators to validate indirect blood pressure devices may have been attributable to the use of a cylindrical cuff on a conical area. We hypothesized that an oscillometric device with a conical cuff would provide better agreement, compared with that obtained by use of a cylindrical cuff, with blood pressure measured invasively at the median sacral artery. Validation of the noninvasive blood pressure devices was defined in accordance with the ACVIM consensus guidelines on blood pressure measurement.17 Those guidelines indicate that for a noninvasive blood pressure device to be validated, the difference of paired measurements must be ≤ 10 mm Hg, SD must be ≤ 15 mm Hg, there should be a correlation of ≥ 0.9 between paired pressures across the range of measured values, 50% of all measurements must be within 10 mm Hg of the values for the reference method, and 80% of all measurements must be within 20 mm Hg of the values for the reference method. In addition, an agreement analysis was used to compare the performance of the 2 cuffs.
Materials and Methods
Animals
Hound-type dogs (n = 17) from the Louisiana State University Division of Laboratory Animal Medicine Research Colony were enrolled in a prospective study. The Louisiana State University Institutional Animal Care and Use Committee approved the study protocol.
Procedures
Use of an anatomically modified conical blood pressure cuffa was evaluated (Figure 1). Dimensions of the conical cuff were determined prior to commencement of the study. Measurements of the circumference at the proximal and distal aspects of the middle third of the antebrachium at the borders of the cylindrical cuff (proximal and distal circumference, respectively) were obtained on a subpopulation of 14 of the dogs. Comparing the measurements for each dog revealed that the distal circumference (mean ± SD, 12.7 ± 1.2 cm) was approximately 80% (mean, 81.2%) of the proximal circumference (mean, 15.6 ± 1.3 cm). This resulted in a mean slant angle of 86.2%, which would have caused a substantial cuff gap if a cylindrical cuff were used. The conical cuff was designed so that the distal border of the cuff had a circumference 80% of the circumference of the proximal border of the cuff to create a cone shape and eliminate the cuff gap. This specific conical cuff had a width of 6.25 cm, which was 30% to 40% of the mean proximal circumference.
Illustration of a cylindrical blood pressure cuff placed on the tapering antebrachium of a dog. Notice the larger circumference of the proximal aspect of the forelimb (proximal circumference), compared with the circumference of the distal aspect of the forelimb (distal circumference) and the gap between the inflated cuff and antebrachial surface.
Citation: American Journal of Veterinary Research 77, 1; 10.2460/ajvr.77.1.59
Dogs were premedicated with a combination of dexmedetomidineb (0.002 mg/kg) and hydromorphonec (0.2 mg/kg) administered IM. Once dogs were adequately sedated, anesthesia was induced with propofold (3 to 5 mg/kg, IV). Dogs were then intubated with an endotracheal tube of an appropriate size, placed in left lateral recumbency, and connected to a circular breathing system. Anesthesia was maintained with isoflurane in oxygen. Heart rate and rhythm, blood pressure, esophageal temperature, expired concentration of carbon dioxide, and pulse oximetry were continuously monitored. Each dog then was fitted at the middle third of the antebrachium with a cylindrical blood pressure cuff of an appropriate size (30% to 40% of the limb's circumference).
The ventral aspect of the tail was prepared by use of dilute chlorhexidine and alcohol. Using sterile technique, a 20-gauge cathetere was inserted into the median sacral artery. Noncompliant tubingf filled with saline (0.9% NaCl) solution was used to connect the arterial catheter to a transducer.g The transducer and blood pressure cuff were connected to a multifunction monitor with an integrated indirect blood pressure measurement system.h A new sterile tubing system was used for each dog, but the same data collection system was used for all dogs.
The transducer was positioned at the level of the base of the heart and calibrated to zero on the basis of atmospheric pressure for each new subject. The system was connected to a pressurized bag of saline solution, and calibration was verified against a mercury manometer by use of a 3-point calibration technique (0, 50, and 150 mm Hg). Arterial catheters were flushed periodically to prevent clots and remove extraneous air bubbles, which could have changed the damping coefficient of the system.
Before data collection commenced, all values were noted to be stable with consistent waveforms. Four paired indirect and direct measurements of SAP, DAP, and MAP were obtained. There was a 2-minute interval between subsequent sets of measurements. Data were collected for the cylindrical cuff, which was then removed.
The conical cuff was then placed in the same location by use of an area marked on the antebrachium. The conical cuff was connected to the same data acquisition system. Direct blood pressure waves were monitored, and stable consistent values were obtained before data recording commenced. Four paired indirect and direct measurements of SAP, DAP, and MAP were obtained. There was a 2-minute interval between subsequent sets of measurements.
Statistical analysis
Data distribution for body weight and age was examined with the Shapiro-Wilk test of normality. Correlation (Pearson correlation coefficient) between paired pressures across the range of measured values was calculated for SAP, DAP, and MAP for each of the cuffs. Correlations were considered significant at P < 0.05. Agreement between direct and indirect blood pressure measurements for the standard (cylindrical) and anatomically modified (conical) cuff was examined by use of Bland-Altman analysis.18 Because the sampling strategy involved a repeated-measures approach, the mean for each of the pressure measurements was calculated and used for comparison purposes. Bias was defined as the mean difference between the 2 methods; 95% LOA were calculated as bias ± (1.96 × SD). Because of the potential for underestimating the SD of the differences when a repeated-measures approach is used, calculated SDs were corrected as described elsewhere.19 All statistical analyses were performed with commercially available software.i
Results
The 17 dogs of the study comprised 10 females and 7 males. Body weight and age were normally distributed. Mean ± SD body weight was 24.24 ± 1.63 kg, and mean age was 5.8 ± 2.5 years.
Four sets of direct and indirect measurements were collected for each dog with each blood pressure cuff. For the standard (cylindrical) cuff, agreement analysis revealed bias for SAP of 3.96 mm Hg (LOA, −8.40 to 16.33 mm Hg), DAP of 2.65 mm Hg (LOA, −6.26 to 11.55 mm Hg), and MAP of 2.12 mm Hg (LOA, −7.95 to 12.19 mm Hg; Figure 2). Adjusted SD for SAP, DAP, and MAP was 6.31, 4.54, and 5.14 mm Hg, respectively. Correlation between paired measurements across the range of measured values for SAP, DAP, and MAP was 0.95, 0.97, and 0.98, respectively; all correlations were significant.
Bland-Altman plots of direct and indirect SAP (A), DAP (B), and MAP (C) for 17 dogs obtained at the median sacral artery (direct blood pressure) and at the antebrachium with an oscillometric device and cylindrical blood pressure cuff (indirect blood pressure). The dotted horizontal line represents the bias, and the solid horizontal lines represent the 95% LOA.
Citation: American Journal of Veterinary Research 77, 1; 10.2460/ajvr.77.1.59
For the conical cuff, agreement analysis revealed bias for SAP of 2.16 mm Hg (LOA, −12.19 to 16.51 mm Hg), DAP of 0.09 mm Hg (LOA, −9.98 to 10.16 mm Hg), and MAP of 0.07 mm Hg (LOA, −9.06 to 9.21 mm Hg; Figure 3). Adjusted SD for SAP, DAP, and MAP was 7.32, 5.14, and 4.66 mm Hg, respectively. Correlation between the paired measurements across the range of measured values for SAP, DAP, and MAP was 0.63, 0.78, and 0.81, respectively; all correlations were significant.
Bland-Altman plots of direct and indirect SAP (A), DAP (B), and MAP (C) for 17 dogs obtained at the median sacral artery (direct blood pressure) and at the antebrachium with an oscillometric device and conical blood pressure cuff. The dotted horizontal line represents the bias, and the solid horizontal lines represent the 95% LOA.
Citation: American Journal of Veterinary Research 77, 1; 10.2460/ajvr.77.1.59
Discussion
In the study reported here, directly measured SAP, DAP, and MAP at the median sacral artery were compared with pressures measured by use of an oscillometric device with a cylindrical cuff and a conical cuff. Dimensions of the conical cuff were determined by use of antebrachial measurements obtained from a subset of 14 dogs in the study. The ACVIM consensus guidelines on blood pressure measurement17 state that for an NIPB device to be validated against a reference method, the difference of paired measurements must be ≤ 10 mm Hg, the SD must be ≤ 15 mm Hg, there should be a correlation of ≥ 0.9 between paired pressures across the range of measured values, 50% of all measurements must be within 10 mm Hg of the values for the reference method, and 80% of all measurements must be within 20 mm Hg of the values for the reference method. The cylindrical cuff met all the ACVIM consensus guidelines for SAP, MAP, and DAP. The conical cuff met the ACVIM consensus guidelines for difference of paired measurements, SD, and percentages of measurements within 10 and 20 mm Hg of the value for the reference method; however, it failed the correlation analysis. In addition, although bias for the conical cuff was less than that of the cylindrical cuff for SAP, MAP, and DAP measurements, LOA for the conical cuff was wider than the LOA for the cylindrical cuff for SAP and MAP measurements.
Questions have been raised as to the appropriateness of applying correlation analysis to results of studies that compare 2 techniques for evaluating the same physiologic variable, as is required by the ACVIM guidelines.18,20 Limitations include the fact that correlation coefficients only measure the strength of the relationship between 2 variables, so a coefficient can be close to 1 even when there is considerable bias between the methods.18 In addition, the range of the values being evaluated can dramatically affect the correlation. Because of these limitations, most investigators seeking to validate a new technique for measuring a biological variable will use agreement analysis.19 Therefore, it is not clear whether the correlation coefficient requirement of the ACVIM guidelines is appropriate. Nevertheless, even if we consider the conical cuff validated because it met the other ACVIM requirements, it failed to have superior agreement with invasively measured SAP and MAP, compared with results for the cylindrical cuff.
Oscillometric devices can provide automated indirect measurement of SAP, DAP, and MAP. These devices work by inflating a cuff, which has been placed around a limb. This creates external pressure that impedes arterial flow. The cuff is then slowly deflated, and the device records the strength of each arterial pulsation as arterial flow returns. These data are then used to determine peak amplitude, which correlates to MAP. Proprietary algorithms then are used to calculate SAP and DAP.21–23 Fitting a subject with a cuff of an appropriate size that is suitably placed on a limb is imperative for accurate measurements24; therefore, we hypothesized that a conical cuff that better conformed to the shape of the antebrachium would yield more accurate measurements.
A possible limitation of the present study was that blood pressure was directly and indirectly measured at 2 anatomic locations. Investigators of a recent study16 found that blood pressure is not constant throughout the body and is dependent on both the site of measurement and body position. Nevertheless, bias in that study16 was relatively low when directly measured blood pressure at the superficial palmar arch and median sacral artery was compared. For that reason, the antebrachium and median sacral artery were selected for comparison of indirect and direct blood pressure measurement in the present study. In addition, the cylindrical cuff used in the study reported here met all of the ACVIM guidelines for blood pressure measurement.
Another limitation of the present study was that all dogs had blood pressure values within normotensive ranges, and no attempt was made to pharmacologically manipulate pressures above or below the normotensive range. It is possible that when values are outside the normotensive range, the relationship between noninvasively and invasively measured blood pressures could change. Adult laboratory dogs with a narrow range of body weight and body condition were used in the study. A greater variety in the body weight and breed of dogs could have affected bias, LOA, and SD.
Another possible shortcoming of the present study was the degree of curvature of the conical cuff. This cuff was specifically designed for use in this study on the basis of antebrachial measurements obtained from a subset of 14 dogs used in the study. Evaluating a larger population of subjects could yield another set of antebrachial measurements and therefore a different curvature of the conical cuff. In the present study, we used only 1 type of conical cuff. A similar study14 on humans involved the use of 4 conical cuffs. Therefore, further studies could be performed with canine subjects to evaluate whether differences in the curvature of a cuff would provide better agreement for blood pressure measurements.
It can be concluded from the study reported here that a conical blood pressure cuff with an 80% ratio should not be used to replace a cylindrical blood pressure cuff because the cylindrical cuff met all the ACVIM consensus guidelines and the conical cuff failed to meet the correlation guideline as defined by the ACVIM guidelines.17 In addition, the conical cuff failed to provide better agreement between noninvasively and invasively measured blood pressure than was obtained with the cylindrical cuff.
Acknowledgments
Support for Ms. Domingues was provided by a Merck Merial summer scholar stipend.
ABBREVIATIONS
| ACVIM | American College of Veterinary Internal Medicine |
| DAP | Diastolic arterial blood pressure |
| LOA | Limits of agreement |
| MAP | Mean arterial blood pressure |
| SAP | Systolic arterial blood pressure |
Footnotes
Provided by Ramsey Medical Inc, Tampa, Fla.
Dexdormitor, Zoetis, Florham Park, NJ.
Hydromorphone HCL injection, USP, Hospira Inc, Lake Forrest, Ill.
PropoFlo, Abbott Laboratories, North Chicago, Ill.
BD Insyte, Becton Dickinson Infusion Therapy System Inc, Sandy, Utah.
Microbore extension set, Hospira Inc, Lake Forest, Ill.
BD DTX Plus, Becton Dickinson, Sandy, Utah.
Advantage OEM BP module-veterinary module, SunTech Medical Inc, Morrisville, NC.
GraphPad Prism, version 6.0, GraphPad Inc, San Diego, Calif.
References
1. Haskins SC. Monitoring anesthetized patients. In: Tranquilli WJ, Thurmon JC, Kurt AG, eds. Lumb & Jones' veterinary anesthesia and analgesia. 4th ed. Ames, Iowa: Blackwell Publishing, 2007;533–558.
2. Brown SA, Henik RA. Diagnosis and treatment of systemic hypertension. Vet Clin North Am Small Anim Pract 1998; 28: 1481–1494.
3. Binns SH, Sisson DD, Buoscio DA, et al. Doppler ultrasonographic, oscillometric sphygmomanometric, and photoplethysmographic techniques for noninvasive blood pressure measurement in anesthetized cats. J Vet Intern Med 1995; 9: 405–414.
4. Acierno MJ, Labato MA. Hypertension in renal disease: diagnosis and treatment. Clin Tech Small Anim Pract 2005; 20: 23–30.
5. Sawyer DC, Guikema AH, Siegel EM. Evaluation of a new oscillometric blood pressure monitor in isoflurane-anesthetized dogs. Vet Anaesth Analg 2004; 31: 27–39.
6. McMurphy RM, Stoll MR, McCubrey R. Accuracy of an oscillometric blood pressure monitor during phenylephrine-induced hypertension in dogs. Am J Vet Res 2006; 67: 1541–1545.
7. Wernick M, Doherr M, Howard J, et al. Evaluation of high-definition and conventional oscillometric blood pressure measurement in anaesthetised dogs using ACVIM guidelines. J Small Anim Pract 2010; 51: 318–324.
8. Shih A, Robertson S, Vigani A, et al. Evaluation of an indirect oscillometric blood pressure monitor in normotensive and hypotensive anesthetized dogs. J Vet Emerg Crit Care (San Antonio) 2010; 20: 313–318.
9. Acierno MJ, Fauth E, Mitchell MA, et al. Measuring the level of agreement between directly measured blood pressure and pressure readings obtained with a veterinary-specific oscillometric unit in anesthetized dogs. J Vet Emerg Crit Care (San Antonio) 2013; 23: 37–40.
10. Drynan EA, Raisis AL. Comparison of invasive versus noninvasive blood pressure measurements before and after hemorrhage in anesthetized Greyhounds using the Surgivet V9203. J Vet Emerg Crit Care (San Antonio) 2013; 23: 523–531.
11. MacFarlane PD, Grint N, Dugdale A. Comparison of invasive and non-invasive blood pressure monitoring during clinical anaesthesia in dogs. Vet Res Commun 2010; 34: 217–227.
12. Garofalo NA, Teixeira Neto FJ, Alvaides RK, et al. Agreement between direct, oscillometric and Doppler ultrasound blood pressures using three different cuff positions in anesthetized dogs. Vet Anaesth Analg 2012; 39: 324–334.
13. Dyce KM, Wensing CJG. The forelimb of the dog and cat. In: Textbook of veterinary anatomy. 4th ed. St Louis: WB Saunders Co, 2010; 476–489.
14. Palatini P, Benetti E, Fania C, et al. Rectangular cuffs may overestimate blood pressure in individuals with large conical arms. J Hypertens 2012; 30: 530–536.
15. Lan H, Al-Jumaily AM, Lowe A, et al. Effect of tissue mechanical properties on cuff-based blood pressure measurements. Med Eng Phys 2011; 33: 1287–1292.
16. Acierno MJ, Domingues ME, Ramos SJ, et al. Comparison of directly measured arterial blood pressure at various anatomic locations in anesthetized dogs. Am J Vet Res 2015; 76: 266–271.
17. Brown S, Atkins C, Bagley R, et al. Guidelines for the identification, evaluation, and management of systemic hypertension in dogs and cats. J Vet Intern Med 2007; 21: 542–558.
18. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–310.
19. Bland JM, Altman DG. Applying the right statistics: analyses of measurement studies. Ultrasound Obstet Gynecol 2003; 22: 85–93.
20. Patton N, Aslam T, Murray G. Statistical strategies to assess reliability in ophthalmology. Eye (Lond) 2006; 20: 749–754.
21. Mauck GW, Smith CR, Geddes LA, et al. The meaning of the point of maximum oscillations in cuff pressure in the indirect measurement of blood pressure—part ii. J Biomech Eng 1980; 102: 28–33.
22. Ramsey M III. Noninvasive automatic determination of mean arterial pressure. Med Biol Eng Comput 1979; 17: 11–18.
23. Lee TK, Westenkow DR. Comparison of blood pressure measured by oscillometry from the supraorbital artery and invasively from the radial artery. J Clin Monit Comput 1998; 14: 113–117.
24. Moens Y, Coppens P. Patient monitoring and monitoring equipment. In: Seymour C, Duke-Novakovski T, eds. BSAVA manual of canine and feline anaesthesia and analgesia. 2nd ed. Quedgeley, Gloucestershire, England: British Small Animal Veterinary Association, 2007;62–79.
