During general anesthesia, monitoring blood pressure is often performed as a marker of cardiac performance and tissue perfusion. Recording invasive arterial blood pressure (IBP) is the most accurate method of monitoring blood pressure.1 However, arterial catheter placement is not always clinically practical, as it can be technically challenging and requires specialized equipment and skilled personnel. Non-invasive blood pressure (NIBP) monitoring using oscillometric monitors, Doppler photoplethysmographic, or photoacoustic probes with a sphygmomanometer may be preferred in clinical practice due to the simplicity of the equipment and technique as well as its widespread availability. Several studies in dogs, cats, and horses have attempted validation of these indirect techniques but were unsuccessful in satisfying American College of Veterinary Internal Medicine (ACVIM) guidelines.2–7 To the author’s knowledge, limited data exists describing blood pressure monitoring in bats. Cannulation of the radial artery of several microchiropteran species (Myotis lucifugus and sodalis) was first described in the mid-1970s.8 Since then, a wide range of systolic and mean arterial pressures have been reported in bats weighing less than 500 g.8–11 Only recently has arterial catheterization and IBP monitoring been described in megachiropterans.12 However, despite the potential advantages of utilizing NIBP monitoring, the use of this technique has not been described in bats.
The accuracy of blood pressure measurements in animals can be influenced by the technique (Doppler vs oscillometric), the specific monitor used, cuff size and placement site, the blood pressure state (ie, normotension vs hypotension vs hypertension), and concurrent administration of drugs that alter systemic vascular resistance.13 Appropriate cuff size and placement are critical for accurately measuring NIBP. The current proposed range for cuff width: limb circumference for many domestic mammalian species is 30% to 40%.14–16 The effect of the cuff placement site is dependent on the species of interest. Although one study showed minimal difference in oscillometric NIBP between the pelvic and pectoral limbs in dogs,17 the site of measurement has been shown to affect the accuracy of NIBP measurements in cats and rabbits.18,19 Understanding how the accuracy of NIBP monitoring techniques is affected by extremes in blood pressure is imperative for interpreting abnormalities and minimizing the chance of treatment errors.
Given the unique limb structure found in bats that facilitates both prolonged hanging and flight, it is currently unknown whether NIBP can be successfully measured in chiropterans and whether the pectoral or pelvic limb cuff placement affects the accuracy of NIBP monitoring. The primary objectives of this study were to compare IBP measurement to both non-invasive oscillometric and Doppler techniques and to evaluate the effect of different cuff sizes, cuff placement sites, and blood pressure states (hypotension, hypertension) on the accuracy of NIBP measurement. We hypothesized that results for the oscillometric and Doppler devices would have good agreement with IBP results and would fulfill the ACVIM recommendation that: (1) the mean difference between paired IBP and NIBP measurements must be ≤ 10 mmHg with a SD ≤ 15 mmHg; and (2) 50% of all measurements must lie within 10 mmHg of the reference method, whereas 80% of all measurements must lie within 20 mmHg of the reference method.20,21
Materials and Methods
Study animals
Eight (3 intact males and 5 intact females) large flying foxes (Pteropus vampyrus) with ages ranging from 3 to 22 years old (mean 12.5 years) and weighing 1.1 ± 0.2 kg were used for the study. Bats were permanent residents of a private conservatory and breeding center and were transported to the University of Florida on the day of the experiment. They were deemed healthy based on physical examination, CBC, biochemistry panel, and fecal evaluation results from 2 weeks before the study. All bats were fasted at least 6 hours before the study. The study was approved by the Institutional Animal Care and Use Committee at the University of Florida and Lubee Bat Conservancy (UF Protocol #201609608).
Anesthetic procedure and instrumentation
Bats were manually restrained in a head-down position and anesthesia was induced with 5% isoflurane (Fluriso; VetOne) in oxygen at a 2 L/min flow rate administered via a face mask attached to a semi-open non-rebreathing Mapleson F circuit (Jorgensen Laboratories). Once muscular relaxation (assessed as a loss of muscle tone in the pelvic and pectoral limbs) and loss of palpebral reflex were apparent, the bats were weighed and then placed in dorsal recumbency with the head slightly flexed ventrally for orotracheal intubation. Intubation was achieved with a laryngoscope with a Magill pediatric blade and an appropriately sized uncuffed endotracheal tube (2–3 mm ID). Using the same semi-open non-rebreathing system, anesthesia was maintained with isoflurane in 100% oxygen and the vaporizer setting was adjusted to maintain a plane of anesthesia based on clinical signs including spontaneous ventilation maintaining end-tidal carbon dioxide concentration (Petco2) between 35 and 45 mmHg, generalized muscle relaxation, and a weak palpebral reflex. A mainstream infrared capnometer (EMMA; Masimo) was attached between the machine end of the endotracheal tube and the breathing circuit for continuous monitoring of Petco2 and respiratory rate. A lead II ECG, peripheral oxygen saturation, and end-tidal concentration of isoflurane (ETISO) were continuously measured by a multiparameter monitor (S/5; GE-Datex Ohmeda). External heat sources including a circulating water blanket (Jorgensen Laboratories) and a forced air warming device (SurgiVet Equator Convective Warming System; Smiths Medical) were used to maintain body temperature near 36–37 °C as measured by a digital rectal thermometer. Significant caution was taken to avoid contact with heat support over 37 °C, as a thermal injury with extensive patagial necrosis and sloughing are possible with both forced air heating and thermal blankets. Bats remained in dorsal recumbency for the duration of anesthesia. Physiologic data were collected and recorded every 5 minutes throughout the study.
The right cephalic vein was aseptically catheterized using a 22-gauge catheter (SurFlash; Terumo Medical). Immediately after venous catheterization, arterial access was achieved using a 24-gauge catheter (SurFlash; Terumo Medical) aseptically introduced in a right posterior tibial artery as previously described.12 The arterial catheter was connected to a short 15.24 cm non-compliant tubing and a disposable pressure transducer system (Deltran II, Utah Medical Products Inc.) that were flushed with heparinized saline (5 IU/mL) and connected to a power lab (ADInstruments PowerLab® Systems) for continuous measurements of invasive, systolic arterial blood pressure (SAPinvasive), mean arterial blood pressure (MAPinvasive), and diastolic arterial blood pressure (DAPinvasive) at a sampling rate of 1 measurement per second. Once connected to the bat, the transducer was zeroed to atmospheric pressure at the level of the manubrium before initiating IBP measurements.
Before experimental procedures were initiated for the day, the transducer and multiparametric monitoring system were calibrated using a digital pressure manometer at 3 different pressures (0 mmHg, 100 mmHg, and 200 mHg). Calibration was deemed acceptable when the difference between the pressure exerted by the manometer and the displayed pressure on the multiparametric monitor was < 2 mmHg. At experiments concluded on each individual bat, drift was calculated by comparing a 100 mmHg pressure exerted by the manometer to the value displayed on the multiparametric monitor. The pressure transducer and the multiparametric monitor were re-calibrated once the drift exceeded 5 mmHg. The pressure transducer and multiparametric monitor were also re-calibrated each morning of the study before data collection. Before recording IBP, damping of the arterial pulse pressure waveform was assessed using a fast-flush test as previously described.22
Blood pressure cuffs (Welch Allyn Neonate Cuffs sizes 1, 2, and 3) were placed midway between the elbow and carpal joints in the pectoral limb and midway between the stifle and tarsal joints in the pelvic limb. All NIBP measurements were performed in the left limbs. Cuff sizes were chosen based on approximate cuff width-to-limb circumferences of 20% to 30% (for cuff 1), 30% to 40% (for cuff 2), and 40% to 50% (for cuff 3). Before data collection, pectoral and pelvic limb circumferences were measured at the site of cuff placement for calculation of the cuff width-to-limb circumference ratios. The 8.2 MHz piezoelectric crystal probe of a Doppler unit (ultrasonic Doppler flow detector Model 811-BL; Parks Medical Electronics) was gently held in place over either the median artery of the pectoral limb or the posterior tibial artery of the pelvic limb using ultrasound contact gel. A sphygmomanometer (Welch Allyn Hand Aneroid Sphygmomanometer) was attached to the cuff and used to inflate the cuff until there was a loss of audible pulsatile flow. The Doppler blood pressure (NIBP-D) was measured when the audible pulsatile flow returned upon deflation of the cuff. Another multiparametric monitor (IntelliVue MP50, Philips Healthcare) connected to the cuff was used to record non-invasive, systolic (SAPnoninvasive), mean (MAPnoninvasive), and diastolic arterial blood pressure (DAPnoninvasive) using an oscillometric (NIBP-O) technique.
Experimental protocol
After instrumentation, blood pressure cuffs were placed on the bats in random order by an investigator blinded to IBP readings. This same individual performed all NIBP-O and NIBP-D measurements to minimize inter-individual variability. Three IBP measurements (SAPinvasive, MAPinvasive, and DAPinvasive) were taken 30 seconds apart immediately after arterial catheterization. The average of these measurements was defined as baseline normotension for the individual bat. Three measurements were taken for NIBP-O and NIBP-D at each location (pectoral and pelvic limb) for each cuff size (1, 2, and 3) with a delay of approximately 10 seconds between readings. All bats had NIBP-D and NIBP-O recorded for both locations using each cuff size while in a normotensive state. Normotension was defined as a MAPinvasive > 60 mmHg and < 110 mmHg23 and a value no more than 10 mmHg greater or less than baseline readings. After this data was collected, bats were randomly assigned into either a hypertensive or a hypotensive group (4 bats in each group). Hypotension was defined as MAPinvasive < 60 mmHg23 and was induced by increasing the isoflurane concentration to 5% on the vaporizer dial. If bats became apneic during the hypotensive phase, manual ventilation was instituted to a maximum peak inspiratory pressure of 15 cmH2O to maintain Petco2 between 35–45 mmHg. Hypertension was defined as MAPinvasive > 110 mmHg23 and was initiated with a constant rate infusion of norepinephrine (Levophed; CIBA Pharmaceuticals Company) dosed at 1–5 µg/kg/minute as well as by reducing the isoflurane concentration. The dose of norepinephrine was based on doses used to create hypertension in dogs and birds.24,25 Definitions for normotension, hypotension, and hypertension were based on the literature found in anesthetized small animals and horses,23 considering these definitions are not specifically reported in bats. All blood pressure measurements were collected in a similar randomized fashion as previously described for the normotensive state after an acclimation period of 5 minutes after initiating either hypertension or hypotension.
Once all measurements were obtained in the abnormal blood pressure state, isoflurane concentrations were adjusted to maintain the patient in a light plane of anesthesia based on clinical signs as previously described, and norepinephrine was immediately discontinued for hypertensive bats. Catheters were removed, isoflurane was discontinued, and the bats were extubated after observation of swallowing or coughing. Once recovery was deemed complete, bats were observed until transportation back to the conservatory.
Statistical analysis
Data were analyzed for normality using the Shapiro-Wilk test. Parametric data were expressed as mean ± SD and non-parametric data were reported as median (interquartile range [IQR]). Arterial SAPinvasive, MAPinvasive, and DAPinvasive during each blood pressure state were averaged for all bats and reported as mean ± SD. The time bats spent in the hypertensive state compared with the hypotensive state and the difference in circumference between pectoral and pelvic limbs were analyzed using a 2-tailed Student t-test. A P < .05 was considered statistically significant for all tests performed.
The percentage of machine failure for the NIBP-O readings at each limb location, cuff size, and pressure state was calculated. An error was defined as when the NIBP-O monitor failed to record a reading with a corresponding error message. The total percentage of machine failure for all attempted NIBP-O readings considering SAPnoninvasive, MAPnoninvasive, and DAPnoninvasive separately was calculated. If the percentage of error for an individual limb location, cuff size, or pressure state was greater than the total machine error for all attempted NIBP-O readings, it was excluded from further statistical analysis, as any statistical analysis would be significantly affected by the small number of paired non-invasive and invasive measurements for comparison. For the NIBP-O measurements, mean bias ± SD, limits of agreement (LOA), and 95% CI were reported for all data that was not excluded from the analysis based on the percentage of error calculations. A 2-way analysis of variance with a post hoc Tukey test was used to assess the effects of blood pressure state, location, and cuff size on blood pressure measurements.
Any measurement obtained by NIBP-D that was greater than 300 mmHg, which is the upper limit of the sphygmomanometer, was reported as 300 mmHg for further statistical analysis. The mean ± SD of the NIBP-D measurements at each cuff location, cuff size, and pressure state were calculated. Bias for the population SAPinvasive and MAPinvasive at each cuff location, cuff size, and pressure state were calculated by subtracting the NIBP-D measurements from the corresponding IBP measurement. The DAPinvasive was excluded from the analysis as it was not expected to relate to the NIBP-D.25 Also, LOA and 95% CI were calculated based on the mean bias using standard error for the mean of the population at each cuff location, cuff size, and pressure state. Mean bias calculated for the SAPinvasive and MAPinvasive was assessed for the effects of cuff position, cuff size, and the pressure state using a 3-way ANOVA with a post hoc Tukey test. Agreement between IBP and each NIBP measurement previously analyzed for bias was calculated using the Bland-Altman method26 for each cuff location, cuff size, and blood pressure state. The mean bias ± 1.96 SD was included for all plots. All statistical analyses were performed using commercially licensed statistical software (Sigma Plot version 15.0; Systat Software).
Results
All bats completed the study and recovered from general anesthesia without complications. Arterial catheterization was successfully achieved in all bats. The mean ± SD anesthesia time from induction to extubation was 103.1 ± 24.6 min. There was no significant difference between the total time bats spent in the hypertensive state (38.3 ± 6.2 min) compared with the hypotensive state (35.3 ± 10.2 min). The mean ± SD ETISO that maintained a light plane of anesthesia in un-premedicated bats was 2.0 ± 0.6%. To achieve a target arterial MAPinvasive during the hypotensive state, the mean ± SD ETISO was 3.8 ± 1.2%. During the period of hypertension, mean ± SD ETISO 1.4 ± 0.5% combined with norepinephrine infusion 3 ± 0.5 µg/kg/min was required to achieve the target blood pressure goals. The combined median (IQR) heart rate of all bats during normotension was 224 beats/min [187, 275 beats/min]. Heart rate was significantly increased (P < .001) during the hypertensive phase (257 beats/min [215, 285 beats/min]) compared with the hypotensive phase (181 beats/min [162, 204 beats/min]). No arrhythmias or other complications occurred during norepinephrine infusions in these bats.
The mean ± SD IBP values during various blood pressure states are reported (Table 1). The mean ± SD time to the first IBP measurement after induction was 12.8 ± 6.9 min. There was a significant difference (P < .001) between the circumference of the pectoral and pelvic limbs (8.2 ± 0.6 cm and 6.6 ± 0.7 cm, respectively). In the pectoral limb, the relationship between cuff width and limb circumference was 26.6 ± 0.02% for cuff 1, 33.5 ± 0.03% for cuff 2, and 48.1 ± 0.06% for cuff 3. In the pelvic limb, the relationship between cuff width and limb circumference was 38.5 ± 0.09% for cuff 1, 49.3 ± 0.1% for cuff 2, and 68.4 ± 0.2% for cuff 3. Even though the present study goals did not include validation of specific NIBP devices in bats, ACVIM Consensus guidelines are compared with the summary statistics for NIBP-D and NIBP-O methods tested on pectoral and pelvic limbs, in relation to IBP values in our study bats (Table 2). One of the ACVIM criteria20,21 is that the correlation between the paired measurements between NIBP and IBP (reference) methods must be < 0.90. Correlation between NIBP and IBP was not tested in this study, however, it was observed that the other ACVIM criteria were also not consistently met in the present study.
Mean ± SD for invasive arterial blood pressure values measured during normo-, hypo-, and hypertensive states in anesthetized, adult healthy bats (n = 8) with total of 288 measurements for normotensive phase and 144 measurements each for hyper- and hypotensive phases.
| Blood pressure | Normotension | Hypotension | Hypertension |
|---|---|---|---|
| SAP (mmHg) | 100 ± 13 | 69 ± 10 | 140 ± 18 |
| MAP (mmHg) | 83 ± 12 | 47 ± 6 | 119 ± 5 |
| DAP (mmHg) | 68 ± 14 | 36 ± 7 | 96 ± 12 |
The American College of Veterinary Internal Medicine (ACVIM) Consensus guidelines compared with the summary statistics for Doppler and oscillometric devices that were tested on pectoral and pelvic limbs, in relation with invasive arterial systolic (SAP), diastolic (DAP), and mean (MAP) blood pressure (BP) in 8 anesthetized, adult healthy bats.
| Oscillometric technique on pectoral limb | |||||
|---|---|---|---|---|---|
| BP state, cuff size | Variable | Bias (mmHg) | SD (mmHg) | ≤ ± 10 mmHg (%) | ≤ ± 20 mmHg (%) |
| Normotension, cuff 1 | SAP | −5.8* | 20.8 | 33.3 | 33.3 |
| MAP | 21.6 | 27.4 | 50.0* | 16.6 | |
| DAP | 16.2 | 34.6 | 33.3 | 33.3 | |
| Normotension, cuff 2 | SAP | 11.0 | 30.5 | 15.4 | 46.1 |
| MAP | 23.5 | 25.9 | 23.1 | 30.7 | |
| DAP | 13.4 | 28.9 | 15.4 | 30.7 | |
| Normotension, cuff 3 | SAP | 17.3 | 21.3 | 20.0 | 40.0 |
| MAP | 29.4 | 19.9 | 10.0 | 5.0 | |
| DAP | 18.7 | 20.9 | 10.0 | 20.0 | |
| Hypotension, cuff 1 | SAP | −9.2* | 20.0 | 40.0 | 20.0 |
| MAP | −15.2 | 13.9* | 40.0 | 0.0 | |
| DAP | −24.0 | 17.5 | 20.0 | 20.0 | |
| Hypotension, cuff 2 | SAP | 2.7* | 25.0 | 30.0 | 50.0 |
| MAP | 1.1* | 18.3 | 40.0 | 30.0 | |
| DAP | −0.6* | 18.6 | 40.0 | 30.0 | |
| Hypotension, cuff 3 | SAP | 0.6* | 14.2* | 58.3* | 33.3 |
| MAP | 0.9* | 13.9* | 25.0 | 66.6 | |
| DAP | −3.7* | 12.3* | 41.6 | 58.3 | |
| Hypertension, cuff 1 | SAP | −6.2* | 16.5 | 20.0 | 60.0 |
| MAP | 32.8 | 23.2 | 20.0 | 20.0 | |
| DAP | 36.2 | 29.8 | 20.0 | 20.0 | |
| Hypertension, cuff 2 | SAP | 3.0* | 10.1* | 60.0* | 25.0 |
| MAP | 31.2 | 9.0* | 0.0 | 25.0 | |
| DAP | 29.2 | 17.1 | 25.0 | 0.0 | |
| Hypertension, cuff 3 | SAP | 22.5 | 22.3 | 42.8 | 14.2 |
| MAP | 30.7 | 23.9 | 28.5 | 14.2 | |
| DAP | 27.1 | 32.5 | 28.5 | 0.0 | |
| Oscillometric technique on pelvic limb | |||||
| BP state, cuff size | Variable | Bias (mmHg) | SD (mmHg) | ≤ ± 10 mmHg (%) | ≤ ± 20 mmHg (%) |
| Normotension, cuff 1 | SAP | 4.7* | 22.1 | 56.5* | 34.7 |
| MAP | 14.6 | 20.8 | 34.7 | 43.4 | |
| DAP | 4.3* | 25.8 | 26.0 | 30.4 | |
| Normotension, cuff 2 | SAP | −0.09* | 13.6* | 47.8 | 43.4 |
| MAP | 15.3 | 16.8 | 47.8 | 21.7 | |
| DAP | 12.3 | 22.0 | 26.0 | 30.4 | |
| Normotension, cuff 3 | SAP | 6.9* | 10.2* | 43.4 | 56.5 |
| MAP | 21.3 | 15.4 | 17.3 | 30.4 | |
| DAP | 16.9 | 18.2 | 39.1 | 21.7 | |
| Hypotension, cuff 1 | SAP | −4.7* | 14.7* | 41.6 | 50.0 |
| MAP | −8.0* | 21.3 | 33.3 | 41.6 | |
| DAP | −6.8* | 17.7 | 16.6 | 58.3 | |
| Hypotension, cuff 2 | SAP | −11.2 | 18.8 | 41.6 | 16.6 |
| MAP | −2.3* | 15.4 | 50.0* | 41.6 | |
| DAP | −5.9* | 17.6 | 25.0 | 58.3 | |
| Hypotension, cuff 3 | SAP | −2.2* | 15.0 | 41.6 | 41.6 |
| MAP | 2.2* | 12.8* | 66.6* | 25.0 | |
| DAP | −0.4* | 14.5* | 41.6 | 41.6 | |
| Hypertension, cuff 1 | SAP | −16.0 | — | 0.0 | 0.0 |
| MAP | 3.0* | — | 0.0 | 0.0 | |
| DAP | −24.0 | — | 0.0 | 0.0 | |
| Hypertension, cuff 2 | SAP | −6.6* | 7.2* | 50.0* | 50.0 |
| MAP | 17.8 | 8.9* | 16.6 | 33.3 | |
| DAP | 10.5 | 11.2* | 33.3 | 66.6 | |
| Hypertension, cuff 3 | SAP | 25.8 | 36.2 | 66.6* | 0.0 |
| MAP | 47.4 | 33.4 | 11.1 | 11.1 | |
| DAP | 35.4 | 28.7 | 22.2 | 11.1 | |
| Doppler technique on pectoral limb | |||||
| BP state, cuff size | Variable | Bias (mmHg) | SD (mmHg) | ≤ ± 10 mmHg (%) | ≤ ± 20 mmHg (%) |
| Normotension, cuff 1 | SAP | 4.7* | 22.1 | 43.4 | 34.7 |
| MAP | −11.5 | 21.6 | 26.0 | 43.4 | |
| Normotension, cuff 2 | SAP | −17.4 | 41.7 | 20.8 | 16.6 |
| MAP | −48.9 | 50.8 | 12.5 | 4.1 | |
| Normotension, cuff 3 | SAP | 6.1* | 18.9 | 41.6 | 25.0 |
| MAP | −22.3 | 21.5 | 25.0 | 20.8 | |
| Hypotension, cuff 1 | SAP | −13.5 | 11.5* | 41.6 | 33.3 |
| MAP | −44.1 | 8.5* | 0.0 | 0.0 | |
| Hypotension, cuff 2 | SAP | 6.17* | 8.93* | 75.0* | 25.0 |
| MAP | −26.6 | 17.84 | 8.3 | 58.3 | |
| Hypotension, cuff 3 | SAP | 9.0* | 15.7 | 66.6* | 8.3 |
| MAP | −25.8 | 22.8 | 16.6 | 16.6 | |
| Hypertension, cuff 1 | SAP | −23.5 | 35.3 | 0.0 | 0.0 |
| MAP | −60.6 | 36.5 | 25.0 | 0.0 | |
| Hypertension, cuff 2 | SAP | −33.4 | 64.8 | 25.0 | 25.0 |
| MAP | −79.5 | 63.4 | 0.0 | 0.0 | |
| Hypertension, cuff 3 | SAP | 24.4 | 64.7 | 16.6 | 8.3 |
| MAP | −26.2 | 73.11 | 8.3 | 8.3 | |
| Doppler technique on pelvic limb | |||||
| BP state, cuff size | Variable | Bias (mmHg) | SD (mmHg) | ≤ ± 10 mmHg (%) | ≤ ± 20 mmHg (%) |
| Normotension, cuff 1 | SAP | 14.9 | 15.0* | 37.5 | 41.6 |
| MAP | −20.0 | 19.0 | 50.0* | 20.8 | |
| Normotension, cuff 2 | SAP | 20.5 | 16.4 | 29.1 | 45.8 |
| MAP | −12.0 | 12.4* | 51.4* | 41.6 | |
| Normotension, cuff 3 | SAP | 29.4 | 15.9 | 12.5 | 16.6 |
| MAP | 0.33 | 17.4 | 54.1* | 33.3 | |
| Hypotension, cuff 1 | SAP | 9.1* | 6.2* | 50.0* | 50.0 |
| MAP | −21.2 | 11.0* | 16.6 | 41.6 | |
| Hypotension, cuff 2 | SAP | 18.9 | 15.6 | 33.3 | 33.3 |
| MAP | −9.8* | 14.2* | 66.6* | 0.0 | |
| Hypotension, cuff 3 | SAP | 33.9 | 8.8* | 0.0 | 8.3 |
| MAP | 1.4* | 7.6* | 75.0* | 25.0 | |
| Hypertension, cuff 1 | SAP | −5.0* | 14.6* | 66.6* | 16.6 |
| MAP | −41.0 | 21.5 | 0.0 | 25.0 | |
| Hypertension, cuff 2 | SAP | −5.1* | 13.0* | 58.3* | 33.3 |
| MAP | −40.0 | 14.6* | 0.0 | 16.6 | |
| Hypertension, cuff 3 | SAP | 23.0 | 23.9 | 8.3 | 33.3 |
| MAP | −10.9 | 26.8 | 41.6 | 25.0 | |
Bias = mean of all differences (invasive BP – non-invasive BP). ACVIM criteria include: (a) the mean difference between paired invasive and non-invasive BP measurements must be ≤ 10 mmHg with a SD ≤ 15 mmHg; (b) 50% of all measurements must lie within 10 mmHg and 80% of all measurements must lie within 20 mmHg of the invasive BP readings (reference method).
Readings complying with the ACVIM recommendations.
From a total of 864 measurements taken from the NIBP-O monitor (considering SAPnoninvasive, MAPnoninvasive, and DAPnoninvasive separately), 252 (29.1%) measurements failed to obtain a reading. A higher incidence of measurement failure occurred with the NIBP-O monitor when blood pressure was measured with a size 1 cuff regardless of location (132/288, 45.8%) and for readings taken from the pectoral limb (183/432, 42.4%). The prevalence of IBP-O measurement failure was highest for bats experiencing hypertension (120/216, 55.6%) and lowest during hypotension (27/216, 12.5%). After analyzing these findings, it was noted that there was a higher incidence of measurement failure reported, which caused a lack of a significant number of paired NIBP-O readings with corresponding IBP readings from the pectoral limb and size 1 cuff, and during hypertension. Because this percentage of error for an individual limb location (ie, pectoral), cuff size (ie, cuff 1), and pressure state (ie, hypertension) was greater than the total machine error for all attempted NIBP-O readings, this data was excluded from further statistical analysis.
Mean ± SD values for SAPinvasive, MAPinvasive, and NIBP-D during various blood pressure states are presented (Table 3). As mentioned earlier, DAPinvasive was excluded from analysis as it was not expected to relate to the NIBP-D.25 In the pectoral limb, NIBP-D measurements using a size 1 cuff significantly over-estimated (P < .001) both SAP invasive and MAP invasive regardless of the blood pressure state (Figures 1 and 2). These findings were similar in the pectoral limb using a size 2 cuff with the exception of the hypotensive state, in which NIBP-D underestimated SAPinvasive (mean bias = 11 mmHg, 95% CI = 0 to 21 mmHg, LOA = −21 to 43 mmHg). Combining all blood pressure states, the NIBP-D demonstrated the best agreement with SAPinvasive in the pectoral limb when using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%) with a mean bias of 2 mmHg (95% CI = −5 to 9 mmHg, LOA = −42 to 46 mmHg). This is compared with a mean bias of −14 mmHg (95% CI = −28 to −1 mmHg, LOA = −105 to 77 mmHg) with size 2 cuff and a mean bias of −18 mmHg (95% CI = −32 to −5 mmHg, LOA = −111 to 74 mmHg) with a size 1 cuff. However, utilizing a size 3 cuff in the pectoral limb caused the NIBP-D to underestimate the SAP invasive during states of hypertension (mean bias = 24 mmHg, 95% CI = −17 to 66 mmHg, LOA = −103 to 151 mmHg) and hypotension (mean bias = 11 mmHg, 95% CI = 0 to 21 mmHg, LOA = −21 to 43 mmHg). During periods of hypertension and hypotension, NIBP-D measurements utilizing a size 3 cuff in the pectoral limb displayed closer agreement to the MAP invasive, with a mean bias of 1 mmHg (95% CI = −45 to 47 mmHg, LOA = −138 to 140 mmHg) during the hypertensive state and a mean bias of −9 mmHg in the hypotensive state (95% CI = −24 to 6 mmHg, LOA = −54 to 36 mmHg). However, results from a 3-way ANOVA indicated there was no significant effect of blood pressure state on NIBP-D measurements.
Invasive arterial systolic (SAP) and mean (MAP) blood pressure measurements and non-invasive Doppler (NIBP-D) blood pressure measurements are presented as mean ± SD (mmHg) with mean bias, SD of the bias, the total number of readings per pressure state (N), limits of agreement (LOA), and 95% CI of the mean bias for the normo-, hypo-, and hypertensive states using cuff sizes 1 through 3 in the pectoral and pelvic limb of 8 anesthetized, adult, healthy bats.
| Cuff 1 | Cuff 2 | Cuff 3 | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Normo- | Hypo- | Hyper- | Normo- | Hypo- | Hyper- | Normo- | Hypo- | Hyper- | |
| Pectoral | |||||||||
| SAP ± SD (mmHg) | 104 ± 12 | 62 ± 20 | 155 ± 25 | 105 ± 15 | 74 ± 22 | 143 ± 17 | 96 ± 13 | 74 ± 15 | 152 ± 27 |
| NIBP-D ± SD (mmHg) | 128 ± 59 | 76 ± 12 | 165 ± 32 | 122 ± 51 | 63 ± 16 | 177 ± 63 | 86 ± 28 | 63 ± 19 | 127 ± 86 |
| N | 24 | 12 | 12 | 24 | 12 | 12 | 24 | 12 | 12 |
| Mean bias | −24 | −14 | −11 | −17 | 11 | −33 | 10 | 11 | 24 |
| ± SD | 62 | 12 | 34 | 42 | 17 | 65 | 21 | 16 | 65 |
| LOA | −147, 98 | −38, 10 | −77, 55 | −99, 64 | −21, 44 | −160, 94 | −30, 51 | −21, 43 | −103, 151 |
| 95% CI | [−51 to 2] | [−22 to −6] | [−32 to 11] | [−35 to 0] | [0 to 22] | [−75 to 8] | [2 to 19] | [0 to 21] | [−17 to 66] |
| Pelvic | |||||||||
| SAP ± SD (mmHg) | 101 ± 13 | 68 ± 10 | 127 ± 11 | 100 ± 15 | 70 ± 10 | 125 ± 14 | 98 ± 12 | 72 ± 10 | 123 ± 15 |
| NIBP-D ± SD (mmHg) | 87 ± 12 | 59 ± 11 | 132 ± 19 | 79 ± 18 | 51 ± 19 | 130 ± 19 | 68 ± 16 | 39 ± 10 | 98 ± 32 |
| N | 24 | 12 | 12 | 24 | 12 | 12 | 24 | 12 | 12 |
| Mean bias | 14 | 9 | −5 | 20 | 19 | −5 | 29 | 34 | 26 |
| ± SD | 15 | 6 | 15 | 16 | 16 | 13 | 16 | 9 | 24 |
| LOA | −15, 44 | −3, 21 | −34, 24 | −12, 53 | −12, 50 | −31, 20 | −2, 61 | 16, 51 | −22, 74 |
| 95% CI | [8 to 21] | [5 to 13] | [−15 to 4] | [14 to 28] | [9 to 29] | [−14 to 3] | [23 to 36] | [28 to 39] | [10 to 41] |
| Pectoral | |||||||||
| MAP ± SD (mmHg) | 85 ± 9 | 48 ± 18 | 132 ± 15 | 90 ± 9 | 58 ± 22 | 120 ± 13 | 81 ± 12 | 54 ± 16 | 128 ± 19 |
| NIBP-D ± SD (mmHg) | 128 ± 59 | 76 ± 12 | 165 ± 32 | 122 ± 51 | 63 ± 16 | 177 ± 63 | 86 ± 28 | 63 ± 19 | 127 ± 86 |
| N | 24 | 12 | 12 | 24 | 12 | 12 | 24 | 12 | 12 |
| Mean bias | −39 | −28 | −34 | −33 | −5 | −56 | −5 | −9 | 1 |
| ± SD | 57 | 10 | 33 | 48 | 25 | 64 | 23 | 23 | 71 |
| LOA | −150, 72 | −48, −9 | −98, 30 | −62, 128 | −55, 44 | −182, 69 | −50, 41 | −54, 36 | −138, 140 |
| 95% CI | [−61 to −17] | [−35 to −22] | [−55 to −13] | [−53 to −12] | [−22 to 11] | [−97 to −15] | [−14 to 5] | [−24 to 6] | [−45 to 47] |
| Pelvic | |||||||||
| MAP ± SD (mmHg) | 82 ± 10 | 54 ± 11 | 108 ± 5 | 82 ± 14 | 55 ± 12 | 108 ± 9 | 81 ± 13 | 57 ± 11 | 108 ± 14 |
| NIBP-D ± SD (mmHg) | 87 ± 12 | 59 ± 11 | 132 ± 19 | 79 ± 18 | 51 ± 19 | 130 ± 19 | 68 ± 16 | 39 ± 10 | 98 ± 32 |
| N | 24 | 12 | 12 | 24 | 12 | 12 | 24 | 12 | 12 |
| Mean bias | −5 | −5 | −24 | 2 | 4 | −22 | 13 | 18 | 9 |
| ± SD | 15 | 11 | 19 | 11 | 10 | 13 | 17 | 4 | 29 |
| LOA | −35, 24 | −26, 17 | −61, 14 | −20, 25 | −15, 23 | −47, 66 | −21, 46 | 10, 26 | −47, 66 |
| 95% CI | [−12 to 1] | [−12 to 2] | [−36 to −11] | [−2 to 7] | [−2 to 10] | [−30 to −13] | [6 to 20] | [16 to 21] | [−9 to 28] |
In the pectoral limb, the relationship between cuff width and limb circumference was 26.6 ± 0.02% for cuff 1, 33.5 ± 0.03% for cuff 2, and 48.1 ± 0.06% for cuff 3. In the pelvic limb, the relationship between cuff width and limb circumference was 38.5 ± 0.09% for cuff 1, 49.3 ± 0.1% for cuff 2, and 68.4 ± 0.2% for cuff 3.
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive systolic arterial blood pressure (SAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the SAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurements and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
Bland-Altman plots of the difference between invasive mean arterial blood pressure (MAPinvasive) and non-invasive blood pressure measured with a Doppler technique (NIBPD) plotted against the average of the MAPinvasive and NIBPD measurements in anesthetized, adult healthy bats (n = 8). The mean bias (dark continuous line) and limits of agreement (mean bias ± 1.96 SD, continuous dotted lines) are shown. Each plot represents the limb and cuff size (cuff width: limb circumference ratio) used for NIBPD measurement and includes: (A) the pectoral limb using a size 1 cuff (cuff width: size circumference ratio of 26.6 ± 0.02%), (B) the pectoral limb using a size 2 cuff (cuff width: size circumference ratio of 33.5 ± 0.03%), (C) the pectoral limb using a size 3 cuff (cuff width: size circumference ratio of 48.1 ± 0.06 %), (D) the pelvic limb using a size 1 cuff (cuff width: size circumference ratio of 38.5 ± 0.09%), (E) the pelvic limb using a size 2 cuff (cuff width: size circumference ratio of 49.3 ± 0.1%), and (F) the pelvic limb using a size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%)
Citation: American Journal of Veterinary Research 84, 5; 10.2460/ajvr.22.12.0208
In the pelvic limb, NIBP-D measurements utilizing a size 3 cuff significantly underestimated (P < .001) the SAPinvasive and MAPinvasive regardless of the blood pressure state (Table 3; Figures 1 and 2). The NIBP-D measurements also consistently overestimated the MAPinvasive utilizing a size 1 cuff for all blood pressure states. NIBP-D measurements obtained with a size 1 cuff measurements tended to over-estimate SAPinvasive during hypertension (mean bias = −5 mmHg, 95% CI = −15 to 4 mmHg, LOA = −34 to 24 mmHg) and underestimate during hypotension (mean bias = 9, 95% CI = 5 to 13 mmHg, LOA = −3 to 21 mmHg). Combining all blood pressure states, the NIBP-D demonstrated the best agreement with MAPinvasive in the pelvic limb when using a size 2 cuff with a mean bias of −3 mmHg (95% CI = −8 to 1 mmHg, LOR = −34 to 27 mmHg). However, NIBP-D measurements using this size cuff tended to overestimate the MAPinvasive during periods of hypertension with a mean bias of −22 (95% CI = −30 to −13 mmHg, LOA = −47 to 4 mmHg). On the other hand, using a size 2 cuff on the pelvic limb significantly underestimated SAPinvasive during normotension and hypotension, causing poor agreement between NIBP-D and SAPinvasive.
The mean bias (CI, LOA) of NIBP-O measurements compared with SAPinvasive, MAPinvasive, and DAPinvasive measurements utilizing a size 2 cuff in the pelvic limb were −2 mmHg (−7 to 3 mmHg, −27 to 12 mmHg) for SAPnoninvasive, 13 mmHg (8 to 20 mmHg, −14 to 42 mmHg) for MAPnoninvasive and 9 mmHg (1 to 17 mmHg, −28 to 46 mmHg) for DAPnoninvasive during periods of normotension. During normotension, NIBP-O measurements significantly underestimated MAPinvasive and DAPinvasive using size 2 cuff, while during hypotension, there was good agreement between NIBP-O and IBP (Table 4). The NIBP-O measurements taken with a size 3 cuff in the pelvic limb tended to consistently underestimate IBP, with a mean bias (CI, LOA) of 7 mmHg (2 to 11 mmHg, −13 to 27 mmHg) for SAPnoninvasive, 22 mmHg (16 to 29 mmHg, −9 to 53 mmHg) for MAPnoninvasive and 18 mmHg (9 to 26 mmHg, −21 to 57 mmHg) for DAPnoninvasive.
Invasive arterial systolic (SAP), mean (MAP), and diastolic (DAP) blood pressure measurements and non-invasive oscillometric systolic (SAPnoninvasive), mean (MAPnoninvasive), and diastolic (DAPnoninvasive) blood pressure measurements are presented as mean ± SD (mmHg) with mean bias, SD of the bias, the total number of readings per pressure state (N), limits of agreement (LOA), and 95% CI of the mean bias for the normo- and hypotensive states using a size 2 cuff (cuff width: circumference ratio 49.3 ± 0.1%) and size 3 cuff (cuff width: size circumference ratio of 68.4 ± 0.2%) in the pelvic limb of 8 anesthetized, adult, healthy bats.
| Cuff 2 | Cuff 3 | |||
|---|---|---|---|---|
| Normo- | Hypo- | Normo- | Hypo- | |
| Systolic blood pressure | ||||
| SAP ± SD (mmHg) | 100 ± 13 | 65 ± 17 | 95 ± 13 | 76 ± 19 |
| SAPnoninvasive ± SD (mmHg) | 102 ± 15 | 71 ± 15 | 89 ± 12 | 76 ± 14 |
| N | 23 | 12 | 23 | 12 |
| Mean bias | −2 | −6 | 7 | 0 |
| ± SD | 13 | 17 | 10 | 16 |
| LOA | −27, 12 | −39, 26 | −13, 27 | −32, 33 |
| 95% CI | [−7 to 3] | [−17 to 4] | [2 to 11] | [−10 to 11] |
| Mean blood pressure | ||||
| MAP ± SD (mmHg) | 84 ± 12 | 51 ± 16 | 80 ± 13 | 58 ± 21 |
| MAPnoninvasive ± SD (mmHg) | 71 ± 16 | 48 ± 17 | 58 ± 13 | 47 ± 13 |
| N | 23 | 12 | 23 | 12 |
| Mean bias | 13 | 3 | 22 | 12 |
| ± SD | 14 | 16 | 16 | 15 |
| LOA | −14, 42 | −29, 25 | −9, 53 | −18, 41 |
| 95% CI | [8 to 20] | [−7 to 13] | [16 to 29] | [2 to 21] |
| Diastolic blood pressure | ||||
| DAP ± SD (mmHg) | 69 ± 15 | 40 ± 17 | 67 ± 16 | 46 ± 23 |
| DAPnoninvasive ± SD (mmHg) | 60 ± 15 | 41 ± 17 | 50 ± 15 | 38 ± 13 |
| N | 23 | 12 | 23 | 12 |
| Mean bias | 9 | −1 | 18 | 8 |
| ± SD | 19 | 20 | 20 | 18 |
| LOA | −28, 46 | −39, 37 | −21, 57 | −28, 43 |
| 95% CI | [1 to 17] | [−14 to 12] | [9 to 26] | [−4 to 20] |
Discussion
According to the ACVIM Consensus guidelines,20,21 criteria to be met for validation include: (1) the mean difference between paired IBP and NIBP measurements must be ≤ 10 mmHg and the SD must be ≤ 15 mmHg, (2) the correlation between paired measurements when treated as a separate entity (eg, SAP or DAP) should be ≥ 0.9 across the range of measured blood pressure values and (3) 50% of all measurements must be within 10 mmHg of the reference IBP readings, and 80% must lie within 20 mmHg of the reference IBP recordings. The present study showed that during all blood pressure states, MAPinvasive estimation by Doppler technology demonstrated the best agreement when using a cuff with a width that is about 50% of the limb circumference (size 2 cuff) and when the cuff is placed mid-way between the stifle and the tarsus. Moreover, the NIBP-D was in good agreement with SAPinvasive in the pectoral limb when using a cuff with a width that is about 70% of the limb circumference (size 3 cuff). Agreement between NIBP-D and MAPinvasive occurred over a range of clinical blood pressure values (MAPinvasive = 55–82 mmHg) in anesthetized bats during the present study. Hence, during general anesthesia where hypotension is a common occurrence, Doppler ultrasound may prove clinically useful in estimating this range of blood pressures in this species of bat when a proper cuff size and measurement location is selected. Even though a reasonable agreement between NIBP-D and IBP was established in our study, which was similar to what has been previously reported in red-tailed hawks,25 this non-invasive device did not fulfill all the ACVIM criteria for validation in our study bats. The bias between the test and the reference methods can be influenced by the range of the measured variables.26 We observed that the LOA widened significantly, despite relatively low bias values when comparing NIBP-D with MAPinvasive. Overall, this lowered the accuracy of using the Doppler technique to measure NIBP in this species of bats over a wider range of blood pressure values. Interestingly, a study in Hispaniolan parrots failed to show appropriate agreement between NIBP-D and IBP.27
The NIBP measured in the pectoral limbs using the Doppler ultrasound technique significantly overestimated the SAP invasive and MAPinvasive for size 1 and 2 cuffs. Also, the NIBP-D obtained in the pectoral limb resulted in greater over-estimation of the MAPinvasive compared with the pelvic limb for each cuff size, while the SAPinvasive was significantly underestimated with all cuff sizes in the pelvic limb compared with NIBP-D values obtained in the pectoral limb. These observations are similar to other studies in birds, where blood pressure obtained by Doppler ultrasound tended to underestimate the SAPinvasive.25,27 There are several conflicting studies in veterinary patients demonstrating the effect of cuff location on the accuracy of non-invasive devices to estimate IBP.17,28–30 Cuff location can influence the agreement between NIBP and IBP when comparing the same cuff size due to alterations in cuff width-to-limb circumference ratios between different locations and variations in anatomy, precluding uniform occlusion of underlying arterial structures during cuff inflation.4,30 The bony and soft tissue structures of the pectoral limb are embedded in the patagium, making the pectoral limb less cylindrical than the pelvic limb.11,31 In this species of bat, the pectoral limb circumference was significantly greater due to the presence of the patagium, resulting in a lower cuff width-to-limb circumference ratio for each cuff size in the pectoral limb compared with the pelvic limb. It is generally accepted that a smaller cuff width in relation to limb circumference results in over-estimation of IBP.32 The results of this study are in line with this observation, as there was a trend for greater overestimation of arterial pressures measured in the pectoral limb compared with the pelvic limb at each cuff size.
The NIBP-O monitor used in this study displayed a high proportion of error readings during all pressure states (29.1%) with the highest error readings with cuff 1 (45.8%) more often on the pectoral limb (42.4%). This could be due to the inability of the small bladder of the size 1 cuff to appropriately inflate the entire circumference of either limb. The patagium, in particular, may have interfered with appropriate circumferential occlusion of underlying arteries in the pectoral limb, leading to error messages. The applied cuff pressure may have been impacted by the patagium, that in turn affected the amplitude of oscillations occurring with pulsatile blood flow within an artery located underneath the cuff, preventing the estimation of MAPnoninvasive. This observation agrees with other studies utilizing oscillometric blood pressure devices in birds25,27,33 and green iguanas.34 Similar to results found in red-tailed hawks,25 the error percentage in the present study was highest during hypertension (55.6%) and lowest during hypotension (12.5%). Increased error readings were likely observed during hypertension, as these values were beyond the upper limit of measurement for the NIBP-O monitor used in our study (150 mmHg) resulting in error messages. Similar observations have been reported in anesthetized and conscious dogs, where elevations in arterial blood pressure resulted in greater measurement error using oscillometric monitors, as the SAPinvasive is underestimated to a greater extent than the MAPinvasive or DAPinvasive during periods of hypertension.17,28,29,35 Chiropterans on average have higher heart rates31 that may have also contributed to the overall increased error readings with the NIBP-O method that relies on the detection of regular pulsatile flow during cuff deflation for accurate pressure readings. Likewise, it is possible that periods of hypotension were associated with lower heart rates closer to domesticated small animals, resulting in lower rates of measurement error. Based upon the findings of this study, the NIBP-O monitor is unsuitable for non-invasive measurement of blood pressure in anesthetized bats; however, if a situation mandates its use, then a size 2 cuff (cuff width:limb circumference ratio 0.5) can be chosen. This should be done with the understanding that oscillometric measurements will significantly underestimate both the MAPinvasive and DAPinvasive during normotension, and will be inaccurate entirely during periods of hypertension. However, during hypotension, the agreement between NIBP-O and IBP could be better with this cuff size.
In domesticated animals, a cuff width-to-limb circumference ratio of 0.4 to 0.6 is recommended to optimize the accuracy of indirect blood pressure measurements.14,29,36–38 Extrapolating data from small mammalian species and birds, three cuff sizes measuring approximately 20–30% (cuff 1), 30–40% (cuff 2), and 40–50% (cuff 3) of limb circumference, were selected for the purpose of this study and the actual cuff width-to-limb circumference ratio was calculated before data collection. Despite the relatively larger cuff widths obtained at the pelvic limb, the most acceptable agreement between NIBP-D and MAPinvasive occurred with a size 2 cuff correlating with a cuff width-to-limb circumference ratio of approximately 0.5. This is within the recommended range of 0.4 to 0.6, but slightly higher than reported in other exotic and domesticated mammalian species to maximize agreement between invasive pressures and pressures obtained non-invasively.25,30,39
Several steps were taken to ensure that reference values remained accurate throughout the data collection process, specifically to avoid underestimation of SAPinvasive that can occur from over-damping. Previous studies in microchiropterans reported over-damping of arterial waveforms, although details regarding the IBP measurement system including the size of the catheter were not described.8 In this study, over-damping was avoided by minimizing elasticity in the system through the use of short non-compliant pressure tubing and catheters while also ensuring that no air bubbles were present within the collection system40 to offset any damping caused by the narrow catheter (24 gauge) required for arterial cannulation in this species. To assess for inappropriate damping, a manual fast flush test was performed and 2 oscillations were required to be observed before return to baseline and a distinct dicrotic notch was noted before data collection was initiated. One significant limitation of this study is the natural resonance frequency was not calculated from dynamic response testing. This is especially important in animals, such as chiropterans, with a heart rate above 180 beats/minute that would require a system resonance frequency of > 24 Hz to prevent underestimation of the SAPinvasive, although the MAPinvasive and DAPinvasive would likely be unaffected.30,41 Therefore, it is possible that the SAPinvasive was underestimated in this study.
Arterial pressures can vary depending on the site of measurement42 and the SAPinvasive measured from peripheral vessels is often higher than in central vessels due to retrograde reflection of resonant waves that travel down the arterial tree that rapidly reflect from areas of arterial branching, leading to amplification of the SAPinvasive in peripheral arteries.43 The magnitude of variation in the SAPinvasive measured at peripheral vessels compared with more centrally located vessels is enhanced during periods of hypertension and tachycardia.44 Therefore, a possible limitation of this study is that all non-invasive measurements in both the pectoral and pelvic limbs were compared with invasive pressure measurements from only the posterior tibial artery. Due to potential variation in arterial pressure at different anatomic locations that can be amplified during alterations in hemodynamic status, it would have been beneficial to compare NIBP measured in the pectoral limb to IBP collected at the median artery. However, it is the authors’ opinion that arterial catheterization of vessels within the wing is more technically challenging than in the pelvic limb, which is why the right posterior tibial artery was chosen for IBP measurement. Despite choosing this location, it is important to consider that only cuff size 2 yielded relatively good agreement with IBP measurements for NIBP-O and NIBP-D.
As the second largest mammalian order, bats comprise over 1,400 species45 representing over 20% of extant mammals.46 It should be noted that P. vampyrus is one of the largest bat species,47 and this should be taken into consideration when applying our data to other bat species. Our clinical experience is that 24-gauge catheters are often too large for the posterior tibial artery of smaller Pteropus species. We caution that smaller bat species would most likely have increased complications in non-invasive blood pressure measurements, exacerbating the discrepancies between NIBP and IBP measurements.
Given the high degree of error, the NIBP-O was not a reliable method of estimating IBP in this species of bats anesthetized with isoflurane. However, across all blood pressure states, the NIBP-D showed acceptable agreement with the MAPinvasive, but not with the SAPinvasive, when a cuff size with a width-to-limb circumference ratio of 0.5 (size 2 cuff) was placed mid-way between the stifle and tarsus. However, with the same size cuff, an agreement between NIBP-D and MAPinvasive decreased with hypertension. On the pectoral limb, NIBP-D values with a width-to-limb circumference ratio of 0.5 (size 3 cuff) exhibited good agreement with SAPinvasive for all blood pressure states. None of the non-invasive methods were able to fulfill all the criteria proposed by ACVIM Consensus. As this study was performed in a healthy population of large flying foxes, extrapolation of results from this study to other species of bats or unhealthy animals should be done with caution.
Acknowledgments
No third-party funding or support was received in connection with this study or the writing or publication of the manuscript.
The authors declare that there were no conflicts of interest.
We would like to thank Lubee Bat Conservancy (Gainesville, FL) for providing the subject animals (study #189) and Sarah Tillman Crevasse for her expertise in restraining and handling bats during the research study.
References
- 1.↑
Wagner AE, Brodbelt DC. Arterial blood pressure monitoring in anesthetized animals. J Am Vet Med Assoc.1997;210(9):1279–1285.
- 2.↑
Bodey AR, Young LE, Bartram DH, et al. A comparison of direct and indirect (oscillometric) measurements of arterial blood pressure in anesthetized dogs using tail and limb cuffs. Res Vet Sci. 1994;57(3):265–269. doi:10.1016/0034-5288(94)90116-3
- 3.
Pederson KM, Butler MA, Ersboll AK, et al. Evaluation of an oscillometric blood pressure monitor for use in anesthetized cats. J Am Vet Med Assoc. 2004;221(5):646–650. doi:10.2460/javma.2002.221.646
- 4.↑
Da Cunha AF, Ramos SJ, Domingues M, et al. Agreement between two oscillometric blood pressure technologies and invasively measured arterial pressure in the dog. Vet Anaesth Analg. 2016;43(2):199–203. doi:10.1111/vaa.12312
- 5.
Cremer J, da Cunha AF, Paul LJ, Liu CC, Acierno MJ. Assessment of a commercially available veterinary blood pressure device used on awake and anesthetized dogs. Am J Vet Res. 2019;80(12):1067–1073. doi:10.2460/ajvr.80.12.1067
- 6.
Cremer J, da Cunha A, Aulakh K, Liu CC, Acierno MJ. Validation of the oscillometric blood pressure monitor Vet20 SunTech in anesthetized healthy cats. Vet Anaesth Analg. 2020;47(3):309–314. doi:10.1016/j.vaa.2019.12.007
- 7.↑
Twele L, Neudeck S, Delarocque J, et al. Agreement of high-definition oscillometry (HDO) and invasive blood pressure measurements at a metatarsal artery in isoflurane-anaesthetised horses. Animals (Basel). 2022;12(3):363. doi:10.3390/ani12030363
- 8.↑
Longnecker DE, Miller FN, Harris PD. Small artery and vein response to ketamine HCl in the bat wing. Anesth Analg. 1974;53(1):64–68.
- 9.
Wiederhielm CA, Weston BV. Microvascular, lymphatic, and tissue pressures in the unanesthetized mammal. Am J Physiol. 1973;225(4):992–996. doi:10.1152/ajplegacy.1973.225.4.992
- 10.
Miller FN, Harris PD. Sensitivity of subcutaneous small arteries and veins to norepinephrine, and epinephrine, and isoproterenol in the unanesthetized bat. Microvasc Res. 1975;10(3):340–351. doi:10.1016/0026-2862(75)90037-0
- 11.↑
Kallen FC. The cardiovascular systems of bats: structure and function. In: Wimsatt WA, ed. Biology of Bats. 1st ed. Academic Press Inc.; 1977;289–483.
- 12.↑
Gatson BJ, Paranjape V, Wellehan JFX, et al. A description of arterial blood pressure measurement in two species of flying foxes (Pteropus vampyrus and Pteropus hypomelanus). J Zoo Wildl Med. 2019;50(3):665–671. doi:10.1638/2018-0218
- 13.↑
Skelding A, Valverde A. Non-invasive blood pressure measurement in animals: part 1 - Techniques for measurement and validation of non-invasive devices. Can Vet J. 2020;61(4):368–374.
- 14.↑
Valtonen MH, Eriksson LM. The effect of cuff width on accuracy of indirect measurement of blood pressure in dogs. Res Vet Sci. 1970;11(4):358–362.
- 15.
Geddes LA, Combs W, Denton W, et al. Indirect mean arterial pressure in the anesthetized dog. Am J Physiol. 1980;238(5):H664–H666. doi:10.1152/ajpheart.1980.238.5.H664
- 16.↑
Sparkes AH, Caney SM, King MC, Gruffydd-Jones TJ. Inter- and intraindividual variation in Doppler ultrasonic indirect blood pressure measurements in healthy cats. J Vet Intern Med. 1999;13(4):314–318. doi:10.1892/0891-6640(1999)013<0314
- 17.↑
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(9):1541–1545. doi:10.2460/ajvr.67.9.1541
- 18.↑
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(6):405–414. doi:10.1111/j.1939-1676.1995.tb03301.x
- 19.↑
Ypsilantis P, Didilis VN, Politou M, et al. A comparative study of invasive and oscillometric methods of arterial blood pressure measurement in the anesthetized rabbit. Res Vet Sci. 2005;78(3):269–275. doi:10.1016/j.rvsc.2004.08.003
- 20.↑
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(3):542–558. doi:10.1892/0891-6640(2007)21[542:gftiea]2.0.co;2
- 21.↑
Acierno MJ, Brown S, Coleman AE, et al. ACVIM consensus statement: guidelines for the identification, evaluation, and management of systemic hypertension in dogs and cats. J Vet Intern Med. 2018;32(6):1803–1822. doi:10.1111/jvim.15331
- 22.↑
Kleinman B, Powell S, Kumar P, et al. The fast-flush test measures the dynamic response of the entire blood pressure monitoring system. Anesthesiology. 1992;77(6):1215–1220. doi:10.1097/00000542-199212000-00024
- 23.↑
Skelding A, Valverde A. Review of non-invasive blood pressure measurement in animals: part 2 - evaluation of the performance of non-invasive devices. Can Vet J. 2020;61(5):481–498.
- 24.↑
Karzai W, Reilly JM, Hoffman WD, et al. Hemodynamic effects of dopamine, norepinephrine, and fluids in a dog model of sepsis. Am J Physiol. 1995;268(2):H692–H702. doi:10.1152/ajpheart.1995.268.2.H692
- 25.↑
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. doi:10.1111/j.1467-2995.2009.00485.x
- 26.↑
Bland JM, Altman DG. Comparing methods of measurement: why plotting difference against standard method is misleading. The Lncet. 1995;346(8982):1085–1087. doi:10.1016/s0140-6736(95)91748-9
- 27.↑
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. doi:10.2460/javma.233.10.1587
- 28.↑
Sawyer DC, Guikema AH, Siegel EM. Evaluation of a new oscillometric blood pressure monitor in isoflurane-anesthetized dogs. Vet Anaesth Analg. 2004;31(1):27–39. doi:10.1111/j.1467-2995.2004.00141.x
- 29.↑
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(4):324–334. doi:10.1111/j.1467-2995.2012.00711.x
- 30.↑
Cerejo SA, Teixeira-Neto FJ, Garofalo NA, et al. Effects of cuff size and position on the agreement between arterial blood pressure measured by Doppler ultrasound and through a dorsal pedal artery catheter in anesthetized cats. Vet Anaesth Analg. 2020;47(2):191–199. doi:10.1016/j.vaa.2019.11.001
- 31.↑
Heard D. Chiropterans (Bats). In: West G, Heard D, Caulkett N (eds.). Zoo Animal and Wildlife Immobilization and Anesthesia. 2nd ed. John Wiley & Sons Inc.; 2014:543–550.
- 32.↑
Grandy JL, Dunlop CI, Hodgson DS, et al. Evaluation of the Doppler ultrasonic method of measuring systolic arterial blood pressure in cats. Am J Vet Res. 1992;53(7):1166–1169.
- 33.↑
Lichtenberger M. Determination of indirect blood pressure in the companion bird. Semin Avian Exotic Pet Med. 2005;14:149–152.
- 34.↑
Chinnadurai SK, DeVoe R, Koenig A, et al. Comparison of an implantable telemetry device and an oscillometric monitor for measurement of blood pressure in anaesthetized and unrestrained green iguanas (Iguana iguana). Vet Anaesth Analg. 2010;37(5):434–439. doi:10.1111/j.1467-2995.2010.00557.x
- 35.↑
Haberman CE, Kang CW, Morgan JD, Brown SA. Evaluation of oscillometric and Doppler ultrasonic methods of indirect blood pressure estimation in conscious dogs. Can J Vet Res. 2006;70(3):211–217.
- 36.↑
Coulter DB, Keith JC Jr. Blood pressures obtained by indirect measurement in conscious dogs. J Am Vet Med Assoc. 1984;184(11):1375–1378.
- 37.
Haberman CE, Morgan JD, Kang CW et al. Evaluation of doppler ultrasonic and oscillometric methods of indirect blood pressure measurement in cats. Intern J Appl Res Vet Med. 2004; 2(4): 279–289.
- 38.↑
Vachon C, Belanger MC, Burns PM. Evaluation of oscillometric and Doppler ultrasonic devices for blood pressure measurements in anesthetized and conscious dogs. Res Vet Sci. 2014;97(1):111–117. doi:10.1016/j.rvsc.2014.05.003
- 39.↑
Hapfelmeier A, Cecconi M, Saugel B. Cardiac output method comparison studies: the relation of the precision of agreement and the precision of method. J Clin Monit Comput. 2016;30(2):149–155. doi:10.1007/s10877-015-9711-x
- 40.↑
Gardner RM. Direct blood pressure measurement–dynamic response requirements. Anesthesiology. 1981;54(3):227–236. doi:10.1097/00000542-198103000-00010
- 41.↑
Hipkins SF, Rutten AJ, Runciman WB. Experimental analysis of catheter-manometer systems in vitro and in vivo. Anesthesiology. 1989;71(6):893–906. doi:10.1097/00000542-198912000-00013
- 42.↑
Acierno MJ, Domingues ME, Ramos SJ, Shelby AM, da Cunha AF. Comparison of directly measured arterial blood pressure at various anatomic locations in anesthetized dogs. Am J Vet Res. 2015;76(3):266–271. doi:10.2460/ajvr.76.3.266
- 43.↑
McGhee BH, Bridges EJ. Monitoring arterial blood pressure: what you may not know. Crit Care Nurse. 2002;22(2):60–64, 66–70, 73 passim.
- 44.↑
Monteiro ER, Campagnol D, Bajotto GC, Simões CR, Rassele AC. Effects of 8 hemodynamic conditions on direct blood pressure values obtained simultaneously from the carotid, femoral and dorsal pedal arteries in dogs. J Vet Cardiol. 2013;15(4):263–270. doi:10.1016/j.jvc.2013.07.002
- 45.↑
Wilson DE, Mittermeier RA (chief eds.). Handbook of the Mammals of the World. Vol. 9 Bats. Lynx Edicions Barcelona; 2019:1008.
- 46.↑
Teeling EC, Springer MS, Madsen O, Bates P, O’Brien SJ, Murphy WJ. A molecular phylogeny for bats illuminates biogeography and the fossil record. Science. 2005;307(5709):580–584. doi:10.1126/science.1105113