Discussion
The primary goal of the present study was to determine whether MAP
measured via NIBP was equivalent to invasive MAP during normovolemic
normotension and hypovolemic hypotension in dorsally-recumbent
anaesthetised horses. The NIBP was measured at the base of the unshaved
tail using a goal cuff width-to-tail circumference ratio of 0.25 because
previous studies have shown that this was the only anatomic site and
cuff width to be equivalent to invasive MAP.12,13,20Invasive MAP was measured at three peripheral arteries (FA, TFA, MT)
based on their clinical relevance and because MAP might vary
substantially among different peripheral arteries in different
animals.8,20 By demonstrating equivalence between MAP
measured via NIBP and invasively in dorsally-recumbent isoflurane
anaesthetized horses over a wide range of systemic arterial pressures
encompassing normovolemic normotension and hypovolemic hypotension, the
current results corroborated and extended previous findings in
normotensive, laterally recumbent, isoflurane-anaesthetized horses and
ponies.11,12 However, equivalence between NIBP and
invasive MAP was obtained for the FA but it was not equivalent for the
TFA in the dorsally-recumbent horses of the present study. This result
differs from the findings using the TFA in laterally-recumbent horses of
a previous study,11 suggesting that body positioning
during anaesthesia might affect equivalence between NIBP and invasive
MAP at least in the TFA. A goal cuff width-to-tail circumference ratio
of approximately 0.25 is critical for accurate
measurements,11,20 with larger ratios of 0.4-0.6 being
consistently found to yield non-equivalent results in anaesthetised and
awake standing horses under normotensive, hypotensive and hypertensive
conditions.6,7,11,20 Taken together, the current
results and the available literature supports the conclusion that NIBP
can be used to accurately monitor arterial blood pressure in
anaesthetised horses during normovolemic normotension and hypovolemic
hypotension provided that it is measured at the tail base with an
appropriate goal cuff width-to-tail circumference ratio of 0.25.
The sizable vertical distance between the middle coccygeal artery and
the heart base requires measurement adjustment to account for
differences in gravity for horses in dorsal recumbency. If this is not
done, readings from the NIBP on the tail will be falsely elevated by
0.735 mmHg for each centimetre of vertical distance. The distribution of
vertical distances in the horses studied was fairly homogeneous around a
median of 27 mmHg, despite the range of breeds and sizes. Equivalence
between NIBP and FA was still demonstrated when the NIBP data was
corrected using a standardised 27 mmHg for all horses. Using this
approach, equivalence was also demonstrated between NIBP and MT. While
ideally the correction should be based on each horse’s individual
vertical distance, these results demonstrate that a standard correction
of 27 mmHg might be acceptable if the vertical distance is not known or
cannot be accurately determined as may be the case in a clinical
setting.
Calculation of sensitivity and specificity based on MAP from the FA that
would prompt treatment for hypotension (MAP < 70 mmHg)
demonstrated that the NIBP had a false negative rate of 12%
(sensitivity of 88%) and a false positive rate of 27% (specificity
73%) when corrected to heart level using each horse’s vertical
distance. When NIBP was corrected to heart level using the standardised
27 mmHg, the false negative rate increased to 20%, and the false
positive rate decreased to 22%. Thus, although equivalence was
demonstrated during a wide range of arterial pressures, NIBP
measurements would cause true hypotension to go untreated in 12-20% of
instances depending on the heart level correction method used. Since in
the 12% false negative instances the NIBP overestimated invasive MAP by
an average of 11 mmHg, hypotension was redefined as NIBP MAP <
80 mmHg and < 90 mmHg while using a standardised heart level
correction of 27 mmHg and sensitivity, specificity, false negative and
false positive data were recalculated. Using this approach, a definition
of hypotension of NIBP MAP < 80 mmHg (corrected to heart level
using 27 mmHg for all horses) or < 107 mmHg (uncorrected to
heart level) resulted in the best balance between false negatives (5%
of hypotensive horses not treated) and false positives (35% of
normotensive horses being treated). Successive measurements at
individual time points could offer additional protection for cases where
values are borderline. Collectively, these results suggest that, when
using NIBP at the tail base with an appropriately sized cuff (goal: 0.25
ratio), hypotension should be defined as NIBP < 80 mmHg
(corrected to heart level) or < 107 mmHg (uncorrected to heart
level) in dorsally-recumbent, anaesthetized horses. However, the
limitations to this approach must also be recognized such that invasive
arterial pressure monitoring should remain the standard in equine
anaesthesia whenever possible.
The variation in blood pressure measurements between different
peripheral arteries found in the current study align with results from
previous studies.20-22 A second goal of the present
investigation was to describe the variability of the differences between
MAP invasive measured in FA, TFA and MT in relation to that measured
centrally in the thoracic aorta, seeking to gain insight into whether or
not there is an ideal or preferred peripheral site for invasive
measurement of arterial blood pressure in anaesthetised horses. The FA
was found to have the least variability, followed by the MT and the TFA,
suggesting that the FA might be the preferred peripheral artery site for
invasive MAP measurement. It is unclear why the variability was greatest
in the TFA. The finding of a mean positive MAP gradient between Ao and
TFA of 2 mmHg (95% CI 0, 4), which is not physiologically possible,
suggests that there were external factors, such as neck/head orientation
during dorsal recumbency, that influenced the accuracy of invasive MAP
determination at this site. The above results suggest that the FA and
TFA are, respectively, the most and least desirable sites for invasive
MAP determination in dorsally recumbent, isoflurane-anaesthetised
horses. These observations along with the finding of equivalence between
NIBP and FA, provide further assurance regarding the usefulness of
measuring MAP via NIBP at the tail base with a goal cuff width-to-tail
circumference ratio of 0.25 in dorsally recumbent,
isoflurane-anaesthetised horses.
Several limitations in this study should be considered. Invasive blood
pressure measurements from the Ao were only available for 4 horses in
the study, which may underpower those results. Invasive blood pressures
in the middle coccygeal artery were not concurrently measured during the
experiment. Our results, similar to other
studies,20,22,23 indicate that MAP varies among
peripheral arterial sites, and therefore NIBP may or may not be an
accurate reflection of the true pressure of the artery it is
measuring.8 Another limitation of this study is that
the experiment was performed on adult horses and may not be translatable
to draft breeds, foals, or ponies.
In conclusion, NIBP measured at the tail (goal cuff
width-to-tail-circumference ratio of 0.25) is equivalent to invasive MAP
from the FA in anaesthetised horses in dorsal recumbency. The results
suggest that hypotension with NIBP should be considered as MAP less than
80 mmHg (when 27 mmHg is subtracted as a heart level correction) or less
than 107 mmHg when no heart level correction is made. Based on
comparisons between invasive MAP in peripheral locations and MAP from
the Ao, the invasive MAP from the FA had the lowest coefficient of
variation, suggesting it is the peripheral artery that is most
reflective of aortic pressures. This study evaluated hypotensive and
normotensive blood pressure readings in adult horses and further
research is needed to guide use in draft breeds, juveniles, or
hypertensive states.
Table 1: Population data, measurements for 0.25 cuff
width-to-tail circumference ratio, cuff selection, and heart level
adjustment based on the vertical distance between the tail and the
scapulohumeral joint (approximation of the right atrium; 1 cm = 0.735
mmHg) for invasive and non-invasive oscillometric blood pressure (NIBP)
comparisons in horses anaesthetised in dorsal recumbency during
normotension and in a hypovolemic haemorrhage model of hypotension
(n = 9).