Precision of repeated, Doppler-derived indirect blood pressure measurements in conscious psittacine birds.
Key words: indirect blood pressure, precision, Doppler, avian, psittacine birds
Systemic arterial blood pressure is the product of cardiac output (defined as the product of heart rate and stroke volume) and total peripheral resistance. (1) Direct measurement of arterial pressure is achieved by placing a catheter directly into an artery and connecting the catheter to an electronic pressure transducer. (2) Although direct pressure measurement is considered the "gold standard" method of evaluating blood pressure, disadvantages associated with its use include the need for specific technical skill, the invasive nature of the procedure, the cost of equipment, the need for maintenance of the catheter with frequent heparinization, and the risk of thrombosis and infection. (2) Several methods of indirect blood pressure measurement are available; all of which are based on detecting blood flow beneath an inflated cuff (2) and provide a noninvasive, less technically demanding option for evaluating blood pressure in a clinical setting.
Although the use of indirect methods for measuring blood pressure has become commonplace in dogs and cats, (3) it is uncertain whether these methods can be extended to avian species with any proven accuracy or precision, despite advocacy to that end. (4,5) A recent report (6) demonstrated substantial disagreement between indirect blood pressure measurements obtained with a Doppler ultrasonic flow detector and those directly measuring systolic arterial blood pressure in anesthetized Hispaniolan Amazon parrots (Amazona ventralis). Results of another recent report (7) comparing Doppler-derived blood pressure measurements to direct arterial blood pressure measurements in awake and anesthetized red-tailed hawks (Buteo jamaicensis) suggest that Doppler-derived blood pressure measurements more closely approximate the mean arterial pressure than the systolic arterial pressure.
A pilot study we conducted on 11 apparently healthy psittacine birds suggested that Doppler blood pressure measurements were highly variable and may differ by site of measurement. Sites explored in that pilot study were the proximal tibiotarsus, distal femur, proximal humerus, and distal humerus, with multiple methods of selecting cuff size evaluated at each site. The sites and cuff selection methods with the least variation in Doppler blood pressure measurements (distal femur and distal humerus, with cuff size approximately equal to 40% of the circumference of the limb) were identified for use in the current study. These pilot data were supported by the results from the study mentioned above on red-tailed hawks, (7) which found that the most reliable cuff size was approximately 40%-50% of the circumference of the limb.
The primary objective of the current study was to determine whether Doppler-derived blood pressure measurements at previously identified sites in individual psittacine birds were repeatable and thus precise; a secondary objective was to identify factors associated with significant variation in Doppler blood pressure measurements in psittacine birds. Our hypotheses were that indirect blood pressure measurements obtained from a site on the wing and the leg would not be significantly different and that repeated indirect blood pressure measurements obtained from the same site would result in precise measurements in individual birds.
Materials and Methods
This study was approved by the Animal Care and Use Committee at Colorado State University (Fort Collins, CO, USA). Twenty-five psittacine birds, weighing 230-1263 g and representing 17 commonly kept species, were examined (Table 1). All parrots were from a flock of approximately 400 birds that were individually housed or housed in pairs and were allowed access to outdoor group flight cages when weather permitted. Each bird was fed a freshly prepared diet consisting of a bean/grain mash, sprouted seeds, nuts, and a variety of fresh produce. All birds were healthy based on results of physical assessments and a minimum 3-month history of customary appetite and droppings, as well as visual daily assessment by the animal caretakers of the aviary. However, health was not a criterion for inclusion in this study because the objective of the study was to evaluate the precision of indirect blood pressure measurements on individual birds and not to attempt to establish reference ranges for the species.
Birds were manually restrained in a towel, and indirect blood pressure measurements were obtained with a Doppler ultrasonic flow detector (model 811-B, Parks Medical Electronics Inc, Aloha, OR, USA). Indirect blood pressure, presumed to correspond more closely to mean arterial pressure, was measured by placing a cuff around the limb proximal to a Doppler crystal held over either the basilic or cranial tibial artery. The cuff was deflated and the Doppler crystal was adjusted until a steady auditory pulse was detected. The cuff was then inflated with a manual sphygmomanometer to a pressure that caused a disruption of the auditory pulse signal from the Doppler detector. The pressure in the cuff was gradually released until the pulse signal could again be heard. This pressure was marked as the indirect blood pressure. A single investigator obtained all measurements.
[FIGURE 1 OMITTED]
Indirect blood pressure measurements from the leg were collected with the cuff placed around the distal femur and the Doppler ultrasonic flow detector placed over the cranial tibial artery. The width of the cuff selected was estimated to be 40% of the circumference of the leg at the level of the distal femur (Fig 1). Blood pressure measurements from the wing were obtained with the cuff around the distal humerus and the flow detector over the basilic artery. The width of the cuff selected was estimated to be 40% of the circumference of the distal humerus with the patagium compressed (Fig 1). Feathers overlying the area of cuff placement were manipulated to allow smooth and even contact between the cuff and limb. The free edge of the cuff was secured with a small piece of white tape before inflation. Feathers overlying the area of ultrasonographic probe placement were moistened with ethyl alcohol and similarly manipulated to allow the probe direct contact with the skin before applying a coupling gel. Three measurements, with complete cuff deflation between each measurement, were taken. Then, the cuff was completely removed and replaced at the site, and another set of 3 readings were taken. This procedure was repeated until 3 separate indirect blood pressure measurement sets were obtained from first the wing and then the leg site on each individual bird. All birds were carefully monitored throughout the procedure and recovered uneventfully from restraint. Birds were restrained and measurements were obtained within a range of 15-20 minutes.
A linear regression mixed effects analysis was performed to determine whether mean blood pressure measurements obtained from 2 limbs (leg versus wing) and from 3 different cuff placements at each site differed significantly. A multilevel modeling approach was undertaken to account for the effect of clustering (nesting) of one variable within another in a hierarchical manner, (8) with "measurement" nested within "cuff placement," which was nested within "site," which was nested within "bird identification (ID)." Hence, a 4-level, random effects modeling was performed with MLwiN software (9) (version 2.0, Centre for Multilevel Modeling, London, England) to account for those nesting effects. The variance explained by each random effect variable was evaluated by z tests to determine whether any random effect had a nonzero variance for blood pressure measurements in psittacine birds; that is, in a 1-sample location test comparing the mean variance of each individual component to 0, the z tests assessed whether the variances due to bird ID or limb or cuff placement were statistically significant. "Site" and "cuff size" were evaluated as fixed effects in the model. Site was also included in the hierarchical structure of the nesting as a random effect. Models with univariable fixed effects were evaluated initially, and variables of P < .25 were eligible for the multivariable model. In the multivariable model, P values [less than or equal to] .05 were considered significant.
In addition, to help with the clinical assessment of the data, a mean value was calculated for each cuff placement on each bird, and the difference between the highest and lowest mean values for each bird ([DELTA]mean) was determined.
Data concerning all bird samples are presented in Table 2. The difference in highest and lowest mean pressure measurements varied from 3 to 113 mm Hg, with an average difference of 32 mm Hg. On univariate analysis, neither cuff site (P = .70) nor limb (P = .23) significantly affected arterial pressure after accounting for the nested nature of the data. The variance contributed by each random effect variable within the hierarchical nesting structure indicated that most of the variance in arterial pressure was attributed to the individual bird (bird ID), followed by cuff placement (with each placement representing a measurement set corresponding to the cuff being removed and replaced at the same site in an individual bird), with the remainder attributed to limb (wing versus leg) and residual variation (Table 3). The variation associated with both bird ID and cuff placement was significant (P < .05), indicating blood pressure varies significantly among individual birds and between cuff placements on the same limb in the same bird. Variation attributable to limb (leg versus wing in an individual bird) was not significant (P = .08).
We attempted to address 2 questions in this study: is significant variation present in repeated Doppler-derived blood pressure measurements in individual psittacine birds, and what factors account for the variability, if present. Our results show that sets of readings obtained at each instance of cuff placement in an individual bird differ significantly. This statistical variability has the potential to be clinically significant, although that is difficult to evaluate because of the lack of reference intervals for indirect blood pressure measurements in avian species. The difference in highest and lowest mean values for each bird, ([DELTA]mean, calculated as highest mean in cuff placement group - lowest mean in cuff placement group) ranged from 3 to 113 mm Hg, with a mean of 32 mm Hg. Although differences in the low mean value are unlikely to be clinically relevant, a mean difference of 113 mm Hg between cuff placements in an individual bird could lead to incorrect clinical assumptions. A reference range would need to be established to see whether the average [DELTA]mean of 32 mm Hg would negatively affect clinical assumptions. In short, sets of measurements taken with the same cuff placed at the same site within a very brief time frame (time for the cuff to be removed and subsequently replaced) were highly variable; therefore, the precision of these indirect blood pressure measurements was poor.
Every attempt to keep the methodology of cuff placement consistent was made, including having a single, experienced individual perform each placement and subsequent measurements to minimize individual variability in technique. The statistical analyses performed took into account the hierarchical structure of the data, the results of which were robust. (9-11) However, a larger sample size, particularly one that represented more diverse psittacine species and weights, might increase the power of our findings.
In a clinical setting, the results of this study suggest that blood pressure measurements in several species of psittacine birds taken during separate cuff placement episodes (eg, monitoring indirect blood pressure in a patient serially over weeks to months) may not provide precise, easily interpretable results. Because most variance was due to individual bird and cuff placement, monitoring indirect blood pressure via the Doppler flow method during a single cuff placement episode (for example, as a method of monitoring an anesthetized patient) could possibly allow for recognizable trends within a single patient. Further evaluation of the precision of indirect blood pressure techniques in anesthetized birds may prove valuable for that purpose. Additionally, investigation into the precision of a larger number of measurements in conscious birds taken within the same cuff placement episode may be of clinical interest in relation to the treatment of hypovolemic shock and in further understanding the cardiovascular effects of handling and restraint on indirect blood pressure measurements. However, when evaluating the individual cuff placement in the individual psittacine bird and multiple cuff placements under the same restraint (Table 2), neither prolonged restraint nor learned anticipation of the inflating cuff appeared to have a consistent effect on the results of Doppler-derived blood pressure.
Interestingly, the site of cuff placement was not statistically significant for variance in a data set, accounting for only 9.25% of the variance. However, the P value of .08, suggests that, with a larger sample size and hence more statistical power, the site of cuff placement might become statistically significant. Compared with individual bird and cuff placement, the chosen site of cuff placement appears to be the least important in the resulting lack of precision.
Our results, coupled with those of the previously mentioned study in Hispaniolan Amazon parrots, (6) suggest that the meaning and clinical value of Doppler-derived indirect blood pressure measurements obtained in psittacine birds remain in question, warranting further research.
Acknowledgments: We thank the Gabriel Foundation (Elizabeth, CO, USA) for allowing access to parrots from their aviary for this study.
(1.) Stepien RL. Blood pressure assessment. In: Ettinger SJ, Feldman EC, eds. Textbook of Veterinary Internal Medicine. 6th ed. St Louis, MO: Saunders/ Elsevier; 2005:470-472.
(2.) de Laforcade AM, Rozanski EA. Central venous pressure and arterial blood pressure measurements. Vet Clin North Am Small Anim Pract. 2001;31(6): 1163-1174.
(3.) Stepien RL. Diagnostic blood pressure measurement. In: Ettinger SJ, Feldman EC, eds. Textbook of Veterinary Internal Medicine. 6th ed. St Louis, MO: Saunders/Elsevier; 2005:282-285.
(4.) Lichtenberger M. Determination of indirect blood pressure in the companion bird. Semin Avian Exotic Pet Med. 2005;14(2):149-152.
(5.) Lichtenberger M. Principles of shock and fluid therapy in special species. Semin Avian Exotic Pet Med. 2004;13(3):142-153.
(6.) Acierno MJ, da Cunha A, Smith J, et al. Agreement between direct and indirect blood pressure measurements obtained from anesthetized Hispaniolan Amazon parrots. J Am Vet Med Assoc. 2008;233(10):1587-1590.
(7.) Zehnder AM, Hawkins MG, Pascoe PJ, et al. Evaluation of indirect blood pressure monitoring in awake and anesthetized red-tailed hawks (Buteo jamaicensis): effects of cuff size, cuff placement, and monitoring equipment. Vet Anaesth Analg. 2009;36(5):464-479.
(8.) Leyland AH, Goldstein H, eds. Wiley Series in Probability and Statistics: Multilevel Modeling of Health Statistics. Chichester, England: J Wiley & Sons; 2001.
(9.) Rasbah J, Steele F, Browne W, Goldstein H. A User's Guide to MLwiN 2.0. London, England: Centre for Multilevel Modeling, Institute of Education, University of London; 2003.
(10.) Dohoo I, Martin W, Stryhn H. A structural approach to data analysis. In: Veterinary Epidemiologic Research. Charlottetown, PE, Canada: University of Prince Edward Island; 2007:581-589.
(11.) Dohoo I, Martin W, Stryhn H. Mixed models for continuous data. In: Veterinary Epidemiologic Research. Charlottetown, PE, Canada: University of Prince Edward Island; 2007:473-498.
Matthew S. Johnston, VMD, Dipl ABVP (Avian), Leslie A. Davidowski, DVM, Sangeeta Rao, MVSc, PhD, and Ashley E. Hill, DVM, MPVM, PhD
From the Department of Clinical Sciences (Johnston) and the Animal Population Health Institute (Rao, Hill), College of Veterinary Medicine and Biomedical Sciences, Colorado State University, 300 W Drake Rd, Fort Collins, CO 80523, USA; and the Southside Hospital for Animals, 10515 Abercorn St, Savannah, GA 31419, USA (Davidowski).
Table 1. Common and scientific names of 25 psittacine birds used for indirect blood pressure measurement. Common name Scientific name No. African grey parrot Psittacus erithacus erithacus 2 Blue and gold macaw Ara ararauna 1 Blue-headed pionus Pionus menstruus 1 Chestnut-fronted macaw Ara severus 1 Double yellow-headed Amazon Amazona oratrix 2 Eclectus parrot Eclectus roratus Goffin's cockatoo Cacatua goffiniana 1 Green-winged macaw Ara chloroptera 1 Lesser sulfur-crested cockatoo Cacatua sulphurea 1 Maximilian's pionus Pionus maximiliani 1 Mealy Amazon Amazona farinosa 1 Medium sulfur-crested cockatoo Cacatua galerita eleonora 1 Orange-winged Amazon Amazona amazonica 2 Red-lored Amazon Amazona auttunnalis 2 Scarlet macaw Ara macao l Umbrella cockatoo Cacatua alba 3 Yellow-naped Amazon Avnazona auropalliata 1 Table 2. Individual indirect blood pressure measurement data sets collected from 25 psittacine birds. Cuff Mean, Species Limb placement mm Hg [DELTA]Mean (a) African grey parrot Leg 1 138 2 155 46 3 184 Wing 1 144 2 151 7 3 149 African grey parrot Leg 1 147 2 151 4 3 150 Wing 1 110 2 156 46 3 138 Blue and gold macaw Leg 1 219 2 211 59 3 160 Wing 1 198 2 203 6 3 197 Blue-headed pionus Leg 1 131 2 157 26 3 157 Wing 1 151 2 146 20 3 131 Chestnut-fronted macaw Leg 1 140 2 131 11 3 129 Wing 1 134 2 125 15 3 140 Double yellow-headed Leg 1 245 Amazon 2 243 48 3 197 Wing 1 183 2 189 27 3 162 Double yellow-headed Leg 1 144 Amazon 2 159 30 3 174 Wing 1 152 2 148 4 3 152 Eclectus parrot Leg 1 146 2 163 35 3 181 Wing 1 160 2 212 52 3 182 Eclectus parrot Leg 1 157 2 215 58 3 212 Wing 1 249 2 234 65 3 184 Eclectus parrot Leg 1 201 2 219 18 3 207 Wing 1 215 2 195 20 3 205 Goffin's cockatoo Leg 1 262 2 247 29 3 233 Wing 1 283 2 301 22 3 279 Green-winged macaw Leg 1 206 2 172 57 3 149 Wing 1 177 2 165 12 3 167 Lesser sulfur-crested Leg 1 201 cockatoo 2 202 3 3 199 Wing 1 203 2 215 36 3 179 Maximilian's pionus Leg 1 195 2 225 30 3 223 Wing 1 220 2 182 50 3 170 Mealy Amazon Leg 1 265 2 260 28 3 237 Wing 1 221 2 254 39 3 215 Medium sulfur-crested Leg 1 229 cockatoo 2 191 38 3 197 Wing 1 217 2 226 10 3 227 Orange-winged Amazon Leg 1 121 2 234 113 3 217 Wing 1 217 2 175 42 3 (b) Orange-winged Amazon Leg 1 193 2 193 6 3 187 Wing 1 189 2 149 40 3 161 Red-lored Amazon Leg 1 211 2 191 31 3 180 Wing 1 165 2 161 6 3 159 Red-lored Amazon Leg 1 233 2 218 28 3 205 Wing 1 191 2 185 40 3 151 Scarlet macaw Leg 1 181 2 165 16 3 171 Wing 1 167 2 225 58 3 222 Umbrella cockatoo Leg 1 186 2 156 30 3 173 Wing 1 169 2 155 32 3 137 Umbrella cockatoo Leg 1 223 2 171 54 3 169 Wing 1 156 2 149 7 3 151 Umbrella cockatoo Leg 1 296 2 263 63 3 233 Wing 1 225 2 250 40 3 265 Yellow-naped Amazon Leg 1 163 2 163 34 3 197 Wing 1 207 2 203 11 3 214 (a) [DELTA]mean = (highest mean in cuff placement group)--(lowest mean in cuff placement group); average Amean for all 50 cuff placementgroups was 32 mm Hg. (b) Data collection was stopped because of the bird's perceived stress. Table 3. Variance explained by each random effect variable within the hierarchical nesting structure. The percentage provided represents how much of the total variance was attributable to a specific effect. Variance, % Variable (total = 1562.33) P value Bird ID 59.78 [less than or equal to] .05 Cuff placement 24.59 [less than or equal to] .05 Limb 9.25 0.08 Residual variance 6.38 [less than or equal to] .05 Abbreviation: ID indicates identification number.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Original Studies|
|Author:||Johnston, Matthew S.; Davidowski, Leslie A.; Rao, Sangeeta; Hill, Ashley E.|
|Publication:||Journal of Avian Medicine and Surgery|
|Date:||Jun 1, 2011|
|Previous Article:||Coming meetings.|
|Next Article:||Plasma osmolality reference values in African grey parrots (psittacus erithacus erithacus), Hispaniolan Amazon parrots (Amazona ventralis), and...|