Ultrasonographic caudal vena cava to aorta ratio changes during blood donation and crystalloid fluid resuscitation in cats

Abstract

Objectives

The aim of the present study was to evaluate caudal vena cava:aorta (CVC:Ao) ratio changes at the hepatorenal location as an ultrasonographic indicator for detecting acute blood loss and monitoring fluid therapy response in cats undergoing blood donation and subsequent crystalloid administration.

Methods

A total of 10 healthy adult cats scheduled for routine blood donation were sedated with ketamine and midazolam. Ultrasonographic measurement of the caudal vena cava (CVC) and aorta (Ao) at the hepatorenal location was performed at baseline, after 10 ml/kg blood donation, and at 5 and 15 mins after intravenous administration of an equal volume of Hartmann’s solution. Triplicate measurements of the CVC:Ao ratio were averaged at each time point and compared using linear mixed models.

Results

The mean (±SD) CVC:Ao ratio at baseline was 1.01 ± 0.07 and decreased to 0.92 ± 0.09 after blood donation, although this change was not statistically significant (P = 0.06). After fluid administration, mean ratios increased to 1.02 ± 0.12 and 1.00 ± 0.08 at 5 and 15 mins, respectively. Only the difference between post-donation and 5 mins post-fluid resuscitation values reached statistical significance (mean difference –0.1, 95% confidence interval [CI] –0.2 to –0.1; P = 0.03). Intra-observer reliability was modest (intraclass correlation coefficient = 0.48, 95% CI 0.29–0.66), although the within-observer SD was 0.09.

Conclusions and clinical relevance

The CVC:Ao ratio at the hepatorenal location demonstrated limited sensitivity for detecting hypovolemia induced by 10 ml/kg blood donation in cats. However, the significant increase observed after fluid resuscitation suggests this measurement may be useful for monitoring response to fluid therapy in cats with previously documented hypovolaemia. The modest intra-observer reliability despite low SD highlights challenges in consistently measuring small vascular structures in cats.

Introduction

Hypovolaemia ranks among the leading causes of shock in feline patients presenting to emergency veterinary services. ¹ The rapid identification and treatment of hypovolaemia in these patients is critical to avoid progressive organ dysfunction and death. Hypovolaemia resulting in the clinical manifestations of shock in feline patients is difficult to distinguish from other causes of shock as feline patients often present with similar shock manifestations regardless of the cause, with common signs including hypotension, hypothermia and bradycardia. ² The development of non-invasive approaches to assess intravascular volume and response to treatment is therefore of utmost importance for this species to avoid fluid administration that may be unnecessary and even dangerous.

Point-of-care ultrasound (POCUS) techniques offer non-invasive methods to assess emergent and critical patients in both human and veterinary medicine.^3,4 There have been limited descriptions of these techniques in feline patients, with most of the veterinary literature thus far focusing on use in canine patients. The caudal vena cava (CVC) diameter is a promising POCUS parameter as its highly compliant nature leads to size alterations with changes in blood volume, and its degree of distension has a potential relationship with cardiac preload. ⁵ The caudal vena cava to aorta (CVC:Ao) ratio has gained traction in human paediatric medicine as it accounts for differences in CVC diameter in paediatric patients of different sizes.^{6 –9} The CVC:Ao ratio has been successfully measured in healthy conscious cats, with reference intervals (RIs) generated at four different image acquisition sites. ¹⁰ In that study, the hepatorenal and iliac bifurcation locations appeared most promising because of ease of acquisition. Multiple canine haemorrhage-resuscitation models have demonstrated that the CVC:Ao ratio correlates with hypovolaemia and demonstrates utility as a static measure of fluid responsiveness.^11,12 The CVC:Ao ratio has not been examined in cats as a method to identify hypovolaemia or assess response to treatment. This non-invasive method has potential as a valuable assessment measure alongside more traditional assessment methods, such as physical examination, laboratory parameters and more invasive hemodynamic monitoring.

The primary aim of this study was to assess the CVC:Ao ratio in cats with iatrogenic hypovolaemia after fixed volume haemorrhage. The secondary aim was to investigate the CVC:Ao ratio after subsequent rapid intravenous crystalloid fluid resuscitation. It was hypothesised that the CVC:Ao ratio would decrease after intravascular volume depletion and return to baseline (pre-haemorrhage) values immediately after fluid resuscitation.

Materials and methods

The experimental protocol incorporated ultrasound assessment of cats that were already enrolled as regular blood donors in an established feline blood donor programme. University teaching and research colony cats scheduled for routine blood donation procedures between January 2023 and February 2025 were enrolled. An a priori sample size calculation to detect an expected effect of 20% with a desired power of 0.80 and an alpha level of 0.05 was conducted using gigacalculator.com (Georgiev GZ) and generated a suggested sample size of seven cats. Ten cats were enrolled to allow for up to 30% dropout. Cats were included in the study if they met the following inclusion criteria: aged older than 1 year; body weight exceeding 3.5 kg; and all parameters within normal limits on physical examination, complete blood count and biochemistry. Cats also underwent a focused ultrasonographic cardiac assessment and were included if they had a normal left atrial:aortic root ratio (<1.6) and left ventricular posterior wall thickness at end diastole (<6 mm). ¹³ All blood donation procedures were performed according to the institution’s standard operating protocols for feline blood donors and in line with established feline blood donation guidelines. These procedures were scheduled independently of the research study and were conducted by trained veterinary personnel to ensure donor welfare and collection quality. This study was conducted in accordance with the University of Queensland Animal Ethics Committee guidelines and received approval under protocol number 2022/AE000408.

To facilitate blood collection, each cat received ketamine (5 mg/kg IV) and midazolam (0.25 mg/kg IV) via a 22 G cephalic intravenous catheter. Cats then underwent blood collection of 10 ml/kg over approximately 5 mins via jugular venepuncture, as previously described. ¹⁴ Blood collection occurred once appropriate sedation was achieved (approximately 2 mins after administration of sedation) and after baseline CVC:Ao ratio measurements were taken. Immediately after blood collection, cats were administered lactated Ringer’s solution (10 ml/kg IV, Baxter Compound Sodium Lactate Hartmann’s solution; Baxter Healthcare) over an additional 5 mins via the cephalic intravenous catheter.

Ultrasound of the CVC:Ao ratio was performed at the hepatorenal site with the cats positioned in left lateral recumbency as previously described. ¹⁰ Ultrasound examinations were performed at four time points: immediately after sedation but before blood donation (time 0); immediately after blood collection (time 1); and at 5 (time 2) and 15 mins (time 3) after the completion of intravenous fluid resuscitation. The 15-min post-resuscitation time point (time 3) was specifically included to assess CVC:Ao ratio changes during the early phase of crystalloid redistribution, as previous feline studies have demonstrated that significant vascular to interstitial fluid shifts occur within this timeframe. ¹⁵ Ultrasound examinations were performed with a microconvex 4.5–8 MHz probe (Mindray Z5; Shenzhen Mindray Bio-Medical Electronics).

Statistical analysis

Statistical analyses were conducted using SAS software version 9.4 (SAS Institute), with the exception of intraclass correlation coefficients (ICCs), which were calculated using the irr package (version 0.84.1, 2012) in R statistical software (R Foundation for Statistical Computing). A significance level of alpha = 0.05 was used for all statistical tests.

For each subject and at each time point, three measurements were obtained by the same observer and performed by lifting and then replacing the probe for each measure. The triplicate CVC:Ao ratios for each cat were averaged before analysis. Normality of the distribution of mean ratios was assessed through visual inspection of quantile-quantile plots, probability-probability plots, histograms and calculation of skewness values. Results are presented as mean ± SD. To evaluate the effect of time on CVC:Ao ratios, linear mixed models were constructed with time as a fixed factor and animal as a random intercept to account for repeated measurements on the same subjects. Model assumptions of normality and homoscedasticity of residuals were verified through inspection of conditional quantile-quantile plots, histograms and residual scatter plots. Multiple pairwise comparisons between time points were performed using Tukey’s adjustment to control for type I errors. The Satterthwaite approximation for degrees of freedom and restricted maximum likelihood estimation methods was employed. Intra-observer reliability was assessed using ICC estimates and their 95% confidence intervals (CIs), calculated based on a single-rating, absolute-agreement, two-way mixed-effects model. ICC values were interpreted according to Koo and Li’s ¹⁶ guidelines: above 0.90 indicated excellent reliability; 0.75–0.90 good reliability; 0.50–0.75 moderate reliability; and below 0.50 poor reliability. The within-observer measurement variability was quantified by calculating the square root of the residual variance estimate from the two-way mixed-effects model.

Results

A total of 10 cats (six males, four females, all neutered) with a mean age of 3.4 ± 1.0 years and mean weight of 4.3 ± 0.65 were enrolled and included in the final analysis. Ultrasound measures of the CVC:Ao ratio were successfully obtained for all cats at all time points.

The mean (± SD) CVC:Ao ratios and their comparisons to baseline (time 0) are reported in Table 1. At baseline (time 0), the mean CVC:Ao ratio was 1.01 ± 0.07. After blood donation (time 1), the ratio decreased to 0.92 ± 0.09, although this change did not reach statistical significance (mean difference 0.09, 95% confidence interval [CI] –0.002 to –0.19; P = 0.06). After fluid administration, the ratio increased to 1.02 ± 0.12 at 5 mins (time 2) and remained increased at 1.00 ± 0.08 at 15 mins (time 3), neither of which differed significantly from baseline (P = 1.00 for both comparisons).

Table 1

Caudal vena cava:aorta ratios for 10 healthy cats undergoing induced hypovolaemia through routine blood donation

Time	Mean ± SD	Mean difference (95% CI) compared with time 0	P value
0	1.01 ± 0.07	–	–
1	0.92 ± 0.09	0.09 (–0.002 to 0.19)	0.06
2	1.02 ± 0.12	–0.01 (–0.10 to 0.09)	1.00
3	1.00 ± 0.08	0.01 (–0.09 to 0.10)	1.00

Measurements were performed in triplicate for each cat at each time point and averaged for each subject before inclusion in the cohort analysis. Time 0 = baseline (after sedation); time 1 = after blood donation; time 2 = 5 mins after intravenous fluid administration; time 3 = 15 mins after intravenous fluid administration

CI = confidence interval

Pairwise comparisons between sequential time points revealed that only the increase in CVC:Ao ratios from after donation (time 1) to 5 mins after fluid administration (time 2) was statistically significant (mean difference –0.10, 95% CI –0.20 to –0.01; P = 0.03) (Table 2). This finding indicates a significant increase in the CVC:Ao ratio after crystalloid administration in blood donation-induced hypovolemia. Individual cat measurements across all time points are displayed in Figure 1.

Table 2

Expanded linear mixed model assessing the change in caudal vena cava:aorta ratios measured at the hepatorenal site in 10 cats undergoing blood donation of 10 ml/kg, followed by intravenous fluid resuscitation

Time	Vs time	Mean difference (95% CI)	P value
0	1	0.09 (–0.002 to 0.19)	0.06
0	2	–0.01 (–0.10 to 0.09)	1.00
0	3	0.01 (–0.09 to 0.10)	1.00
1	2	–0.10 (–0.20 to –0.01)	0.03
1	3	–0.09 (–0.18 to 0.01)	0.09
2	3	0.02 (–0.08 to 0.11)	0.97

Time 0 = baseline (after sedation); time 1 = after blood donation; time 2 = 5 mins after intravenous fluid administration; time 3 = 15 mins after intravenous fluid administration

CI = confidence interval

Figure 1

Mean caudal vena cava:aorta (CVC:Ao) ratio measurements for 10 healthy, sedated cats at baseline (time 0), after induced hypovolaemia during routine blood donation (time 1), and 5 and 15 mins after fluid resuscitation (times 2 and 3, respectively)

Assessment of measurement reliability revealed modest intra-observer agreement for the CVC:Ao ratio, with an ICC of 0.48 (95% CI 0.29–0.66), indicating poor to moderate reliability according to established criteria. Despite this, the within-observer SD was only 0.09, suggesting that absolute measurement variability was relatively small in magnitude.

Discussion

This study is the first to investigate CVC:Ao ratio dynamics during experimentally induced hypovolaemia in cats. Our findings demonstrate that the CVC:Ao ratio measured at the hepatorenal location did not significantly decrease during a standardised 10 ml/kg blood donation, and nor did values fall below previously published RIs for unsedated cats. ¹⁰ However, the CVC:Ao ratio did exhibit a significant increase after crystalloid resuscitation, thus suggesting a potential use for monitoring fluid therapy response rather than using it as a primary diagnostic tool for detecting mild to moderate hypovolaemia in cats.

The absence of a significant CVC:Ao ratio decrease after blood donation in our feline cohort aligns with findings from Herreria-Bustillo et al, ¹⁷ who similarly reported no significant CVC:Ao ratio reduction in eight healthy dogs undergoing blood donation when measured at the iliac bifurcation (16.53 ml/kg donated). However, this is in contrast to the findings of Cambournac et al, ¹¹ who observed significant CVC:Ao ratio reduction in 12 dogs undergoing blood donation at the splenorenal measurement site, despite the dogs in that study donating proportionally less blood (9.8 ml/kg). This apparent paradox – where greater blood loss failed to produce the expected response – highlights the complexity of vascular dynamics during acute volume loss and suggests that multiple factors beyond absolute volume removed the influence of CVC:Ao measurements. Several factors may account for these inter-study differences, including measurement site selection, species-specific vascular compliance characteristics, baseline cardiovascular status and compensatory mechanisms, making direct comparison between studies difficult.

Our observation that CVC:Ao ratio values in this study did not fall below previously established RIs for unsedated cats, despite modest induced hypovolaemia, may have several possible explanations. ¹⁰ First, although the standardised 10 ml/kg blood donation volume may be clinically relevant for donor programmes, it may have been insufficient to overcome compensatory mechanisms in healthy cats. In addition, rapid compensatory fluid redistribution likely occurred during the experimental period. The phenomenon of ‘transcapillary refill’ is well documented in human medicine and describes the rapid mobilisation of interstitial fluid into the intravascular space during acute hypovolaemia. ¹⁸ Although this mechanism has not been extensively characterised in cats, our data suggest it may be particularly efficient in this species, potentially limiting the utility of CVC:Ao ratio measurements for detecting mild volume depletion.

The significant CVC:Ao ratio increase observed between post-donation (time 1) and post-fluid resuscitation (time 2) measurements (P = 0.03) represents the most clinically relevant finding of our study. This response likely reflects acute intravascular volume expansion and increased cardiac preload after crystalloid administration. Clinical canine studies have demonstrated a similar pattern when evaluating fluid responsiveness in dogs with hypoperfusion.^4,19 This suggests that although CVC:Ao ratio may demonstrate limited sensitivity for detecting initial hypovolaemia, it appears responsive to acute increases in intravascular volume status, supporting its potential role in monitoring volume expansion in cases of documented hypovolaemia.

The modest intra-observer reliability (ICC = 0.48) observed in our study requires careful interpretation in the context of feline anatomy and measurement constraints. Although this ICC value indicates poor to moderate reliability according to established criteria, ¹⁶ the relatively small within-observer SD (0.09) suggests reasonable measurement consistency. The apparent contradiction between ICC and SD values likely reflects the mathematical influence of limited inter-subject variability on ICC calculations – when the range of true values is narrow, even small measurement errors can disproportionately affect reliability statistics.

The technical challenges of measuring small vascular structures in cats cannot be understated. Feline vascular dimensions are substantially smaller than those in dogs or humans, making precise measurement inherently more difficult and potentially more susceptible to operator-dependent variation. ²⁰ In addition, factors such as motion during ventilation, ultrasound probe pressure and precise anatomical positioning may have magnified this variability in smaller species.

Several limitations should be considered when interpreting our results. First, all cats underwent sedation with midazolam and ketamine, which may have influenced cardiovascular dynamics and CVC:Ao ratio measurements. Although this sedative combination was selected specifically for its cardiovascular-sparing properties, some effect on vascular tone and cardiac preload cannot be completely excluded. ²¹ Future studies in conscious cats with naturally occurring hypovolaemia would help to clarify the clinical relevance of our findings.

Second, our use of controlled blood donation as an experimental model for hypovolaemia, although ethically appropriate and clinically relevant, may not fully replicate the complex pathophysiology of naturally occurring shock states. It is possible that larger donated volumes or an experimental haemorrhage model, including vessel damage, may have produced a significant change in the CVC:Ao ratio and been more representative of the volume status of patient populations presenting to veterinary hospitals with naturally occurring hypoperfusion.

Finally, our sample size of 10 cats, although exceeding our calculated minimum requirement, limits statistical power for detecting smaller effect sizes. The trend towards decreased CVC:Ao ratios after blood donation (P = 0.06) suggests that a larger study population might have detected a statistically significant change. However, the clinical significance of such a small change would require careful evaluation.

Our findings suggest that CVC:Ao ratio measurements at the hepatorenal location may have limited utility as a standalone diagnostic tool for detecting mild to moderate hypovolaemia in cats. However, the significant response to fluid administration indicates potential value in monitoring therapeutic interventions. This application could be particularly relevant in critically ill feline patients where optimising fluid therapy is challenging due to species-specific physiological constraints and susceptibility to both under-resuscitation and volume overload.

Future research should focus on several key areas: evaluation of CVC:Ao ratios in cats with naturally occurring hypovolaemia and shock states to determine clinical sensitivity and specificity; investigation of CVC:Ao ratio as a predictor of fluid responsiveness in cats receiving crystalloid administration, including correlation with haemodynamic improvements and clinical outcomes; assessment of inter-observer reliability in clinical settings; correlation of CVC:Ao changes with other haemodynamic parameters such as cardiac output, blood pressure and lactate levels; and the development of standardised measurement protocols that account for the technical challenges specific to feline patients.

In addition, an investigation into how CVC:Ao ratio measurements integrate with current clinical monitoring methods, such as heart rate, blood pressure, capillary refill time and serial lactate measurements, may provide a more comprehensive approach to volume status assessment. Combining CVC:Ao ratios with established clinical parameters could potentially overcome the limitations of individual measurements and improve overall diagnostic accuracy in feline patients.

Conclusions

This study provides the first systematic evaluation of CVC:Ao ratio dynamics during experimentally induced hypovolaemia in cats. Although the CVC:Ao ratio demonstrated limited sensitivity for detecting mild volume depletion at the hepatorenal measurement site, its significant increase after fluid resuscitation suggests utility for monitoring therapeutic response. The modest intra-observer reliability highlights the challenge of consistently measuring small vascular structures in cats, although the low standard deviation suggests relatively consistent measurements. Future studies in cats with naturally occurring shock states are warranted to fully define the clinical utility of this promising ultrasonographic parameter.