VITALS Deep Dive Article — by Niki DeValk, AAS, LVT, SRS
Interventional & Surgical Specialist | Owner, NiKara Preclinical
Abstract
The electrocardiogram is among the most continuous and information-dense monitoring tools available during preclinical anesthesia and surgery, and in many study designs it is also a primary or secondary endpoint. Accurate interpretation depends on recognizing that normal conduction physiology differs substantially across the species used in preclinical research, that recording conditions such as body position and electrode placement measurably alter the trace, and that common anesthetic agents and electrolyte shifts produce predictable electrocardiographic signatures. This article reviews comparative conduction physiology across swine, ruminants, horses, dogs, rabbits, rodents, guinea pigs, ferrets, and birds; summarizes practical recording methodology; describes the electrocardiographic effects of alpha-2 agonists, opioids, anticholinergics, dissociatives, and inhalant agents; and outlines how the 2022 ICH E14/S7B Questions and Answers have increased the regulatory relevance of high-quality in vivo ECG data. The evidence base is strongest for dogs and laboratory rodents, moderate for swine, and comparatively limited for small ruminants and birds, where much guidance still rests on older reference studies and practitioner experience. Throughout, the emphasis is on interpreting the ECG in context, alongside blood pressure, perfusion, ventilation, and timing, so that expected drug effects are distinguished from developing problems and both animal welfare and data quality are protected.
Introduction
Few monitoring parameters are as familiar, or as easily taken for granted, as the ECG. It is usually the first lead attached after induction and the last removed in recovery. Yet its value depends entirely on the interpreter's frame of reference. A rhythm that signals a problem in one species can be a physiologic norm in another, and a waveform change that looks like pathology may simply reflect the drug given ten minutes earlier.
This matters on two levels in preclinical work. The first is welfare and intraoperative safety: the ECG is often the earliest indicator of autonomic shifts, electrolyte disturbances, hypoxemia, and mechanical irritation of the heart during interventional procedures. The second is data integrity. In cardiovascular device studies, interventional models, and safety pharmacology programs, ECG intervals and rhythm events are frequently study data, and the conditions under which they are recorded become part of the study's scientific validity.
The regulatory landscape has amplified the second point. The ICH E14/S7B Questions and Answers adopted in February 2022 [1, 2] describe how nonclinical data, including in vivo QT studies conducted to best-practice standards, can be integrated with clinical data to reach low-risk QTc conclusions and, in defined scenarios, reduce reliance on a dedicated clinical thorough QT study. As a result, the quality of animal ECG acquisition and interpretation now has a more direct line to drug development decisions than it did a decade ago.
This article is written from the perspective of a practitioner who records and interprets ECGs across many species in both GLP and non-GLP settings. It aims to give surgical, anesthesia, and study teams a comparative framework they can use at the table, while being candid about where the evidence is robust and where it remains thin.
Comparative Conduction Physiology, Recording, and Interpretation
The most useful starting point for comparative electrocardiography is the way the ventricles are activated. In dogs and laboratory rodents, the Purkinje network is largely confined to the subendocardium, so ventricular depolarization spreads in a relatively orderly wave from endocardium to epicardium. Body-surface leads in these species therefore carry meaningful information about the mean electrical axis and chamber size. In ungulates, including swine, cattle, sheep, goats, and horses, Purkinje fibers penetrate deeply into the ventricular myocardium, and large portions of the ventricles depolarize almost simultaneously. Much of that activity cancels out at the body surface, which is why limb-lead axis and amplitude criteria translate poorly to these species and why the base-apex lead is the preferred monitoring configuration for large animals in dorsal or lateral recumbency. Swine sit at an important intersection here: they are a leading translational model for human cardiovascular disease, yet their transmural Purkinje distribution, a sinoatrial node positioned lower on the septal wall, and a more proximal His bundle bifurcation make their ECG distinctly non-human [3, 10]. Their well-recognized susceptibility to ventricular arrhythmias during cardiac handling and catheter manipulation is part of the same physiologic picture [10]. Birds also show deep Purkinje penetration, and in chickens and other galliform species the dominant ventricular deflection in lead II is typically negative, a finding that is entirely normal but can alarm a mammal-trained eye.
Resting autonomic tone adds a second layer of species variation. Horses have high resting vagal tone, and second-degree atrioventricular block and sinus arrhythmia at rest are generally physiologic, resolving as sympathetic tone rises. Dogs commonly show respiratory sinus arrhythmia and a wandering atrial pacemaker, both of which reflect healthy vagal modulation. At the other end of the spectrum, mice and rats operate at heart rates of several hundred beats per minute. Their repolarization depends heavily on fast transient outward potassium currents, so the T wave merges with the end of the QRS and a clearly isoelectric ST segment is often absent. Defining the end of the T wave is correspondingly difficult, and heart-rate correction formulas developed for humans perform poorly in rodents. Rabbits and guinea pigs, whose ventricular repolarization relies more on the delayed rectifier currents that dominate in humans, have long been favored for QT-focused work for exactly this reason.
Recording conditions measurably shape the trace. In anesthetized swine, moving from right lateral to dorsal recumbency produced significant differences in P- and R-wave amplitudes and in several interval durations, and the authors developed separate reference ranges for each position [3]. Similar position effects are well described in dogs. The practical implication is that reference values are only valid for the position, lead configuration, and anesthetic state in which they were generated, and that within-study consistency of setup matters as much as the absolute values. Electrode choice should match patient size: adhesive electrodes on clipped, degreased skin for large animals; padded clips or fine subcutaneous needle electrodes for rabbits, ferrets, guinea pigs, and rodents, whose low-amplitude signals are easily lost to artifact. Sampling rate and filtering also deserve attention, since rodent complexes are short enough that low sampling rates and aggressive filtering can truncate intervals. Keeping lead wires away from electrosurgical cables, warming devices, and infusion pumps remains the simplest way to protect signal quality.
Anesthetic agents leave consistent electrocardiographic fingerprints in research animals. Alpha-2 adrenergic agonists such as dexmedetomidine, medetomidine, and xylazine reliably produce sinus bradycardia and frequently first- or second-degree atrioventricular block. In purpose-bred laboratory beagles, medetomidine produced bradycardia with occasional second-degree atrioventricular block in the first 15 minutes after injection, and atropine or glycopyrrolate abolished the bradycardia only transiently [5]. In rats and rabbits, medetomidine activates the cardiac vagal nerve through modulation of baroreflex control [6], and in isolated rat ventricular myocytes dexmedetomidine reduced calcium currents and contractility in a dose-dependent manner [7]. For interventional and electrophysiology models, an invasive electrophysiology study in pigs concluded that medetomidine and dexmedetomidine can be used as standard premedication without affecting study results [4]. Because adding an anticholinergic to an alpha-2 agonist can raise blood pressure and myocardial workload while vasoconstriction persists, many preclinical teams tolerate the expected conduction pattern when perfusion and pressure are maintained, and set clear thresholds for intervention. Ruminants are notably more sensitive to xylazine than other species. Opioids tend to slow the heart through vagal mechanisms, while anticholinergics increase heart rate but can transiently worsen atrioventricular conduction immediately after administration. Ketamine generally supports heart rate through sympathomimetic effects, and propofol and alfaxalone can produce dose-related bradycardia and hypotension. Inhalant agents influence repolarization: isoflurane prolongs QT and PQ intervals in mice, and baseline intervals under standardized isoflurane anesthesia have been published for C57BL/6 mice [11]. These effects are not complications in themselves; they are expected pharmacology, and recognizing them prevents unnecessary intervention and keeps drug effects from being misread as test-article effects.
Electrolyte, metabolic, and thermal disturbances produce their own recognizable patterns. Hyperkalemia progresses from tall, narrow, peaked T waves to flattened or absent P waves, widening of the QRS complex, and bradycardia. Hypokalemia is associated with ST-segment depression, prominent U waves where they are visible, and increased ectopy. Hypocalcemia prolongs the QT interval and hypercalcemia shortens it. Hypothermia slows the sinus rate and prolongs intervals across the board, which is particularly relevant in small mammals and birds with high surface-area-to-volume ratios. In each case, the ECG is best treated as a prompt for a blood gas, electrolyte panel, or temperature check, so that the cause is corrected rather than only the rhythm.
During interventional and cardiac procedures, the ECG becomes a real-time guide to mechanical and procedural effects. Catheter and wire contact with the atria or ventricles commonly provokes premature beats that resolve when the device is repositioned. Swine warrant particular preparation, and many established porcine cardiac protocols include planned antiarrhythmic strategies, ready defibrillation, and close attention to electrolytes before manipulation begins [10]. In peripheral work, such as renal denervation or embolic procedures, energy delivery and vessel manipulation can produce autonomic responses visible as abrupt rate and rhythm changes, and anticipating these allows analgesic and hemodynamic support to be timed in advance. With advanced transcatheter, transseptal, myocardial injury, and bypass models, rhythm monitoring is often continuous across induction, intervention, and recovery, and it frequently feeds directly into study endpoints.
In safety pharmacology, ECG quality has direct regulatory consequences. The 2022 ICH E14/S7B Q&As set out best-practice considerations for in vivo QT studies, including adequate exposure, appropriate heart-rate correction, sufficient sensitivity, and well-controlled recording conditions, and describe how a "double-negative" integrated nonclinical assessment can support a low-risk QTc conclusion [1]. Conscious telemetry in dogs, swine, and other non-rodent species remains the reference approach, increasingly complemented by jacketed external telemetry and group housing. As recordings grow longer and denser, automated analytics that screen for artifact, abnormal RR intervals, and arrhythmia classes are becoming standard, with the explicit recommendation that automated detection be paired with informed human review [9]. Deep learning methods are also being explored in animal ECGs; a transfer learning model trained on human ECG data, for example, was used to predict cardiac conduction disease from early mouse ECGs [8]. These tools are promising, but they amplify rather than replace the need for trained interpreters who understand species-specific normal patterns.
Discussion and Conclusion
The strength of the evidence behind these principles varies by species and question. Comparative activation patterns, the base-apex lead for ungulates, and the electrocardiographic effects of alpha-2 agonists in dogs rest on decades of consistent findings and can be applied with confidence. Rodent ECG methodology and anesthetic effects are also well characterized, although disagreement persists about the best way to define the end of the T wave and correct QT for heart rate. The porcine literature is solid for anesthetized reference values and position effects but still lacks a standardized 12-lead electrode system [3]. Evidence is thinnest for small ruminants, ferrets, and birds, where much day-to-day practice draws on older reference studies, extrapolation from related species, and experienced practitioner judgment. Much of the foundational alpha-2 electrocardiographic work comes from purpose-bred dogs and rodents, so extrapolation to small ruminants and birds should be made thoughtfully, although invasive electrophysiology data in pigs are reassuring [4].
What emerges consistently across species is that the ECG is most powerful when read in context. A rhythm interpreted alongside blood pressure, perfusion, end-tidal CO2, temperature, drug timing, and procedural events supports better decisions than any single waveform. That contextual reading is what allows a team to tolerate an expected alpha-2 conduction pattern, act early on a peaked T wave, and anticipate catheter-induced ectopy rather than react to it.
For study teams, the practical conclusions are straightforward. Define species-, position-, and protocol-specific normal values before a study begins. Standardize electrode type, placement, and recording conditions, especially when the ECG is an endpoint. Build anesthetic fingerprints and electrolyte patterns into training so that interpreters can tell pharmacology from pathology. And as automated and deep learning tools become part of routine analysis, invest equally in the human expertise that validates their output. Done well, ECG interpretation protects the animal on the table and strengthens every dataset it touches.
Disclaimer
This article is intended for educational purposes for trained research, surgical, and anesthesia professionals. It does not provide drug doses, and nothing in it replaces institutional protocols, IACUC-approved procedures, or the direction of the attending or study veterinarian. Heart rate ranges and electrocardiographic patterns vary with age, breed, strain, body size, and anesthetic protocol.
References
International Council for Harmonisation. E14 and S7B Clinical and Nonclinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential: Questions and Answers. Adopted February 21, 2022. FDA guidance
International Council for Harmonisation. E14/S7B Q&A Training Materials. March 2022. PMDA
Comparative analysis of ECG records depending on body position in domestic swine (Sus scrofa domestica). Porcine Health Management. 2022. doi:10.1186/s40813-022-00282-x. PMC
Cepiel A, Noszczyk-Nowak A, Janiszewski A, Pasławski R, Pasławska U. Effect of xylazine, medetomidine and dexmedetomidine on cardiac conduction in pigs. Medycyna Weterynaryjna. 2018;74(3):187–192. Link
Vainio O, Palmu L. Cardiovascular and respiratory effects of medetomidine in dogs and influence of anticholinergics. Acta Veterinaria Scandinavica. 1989;30:401–408. PDF
Shimizu S, Akiyama T, Kawada T, et al. Medetomidine, an alpha-2 adrenergic agonist, activates cardiac vagal nerve through modulation of baroreflex control. Circulation Journal. 2012;76:152–159.
Effects of dexmedetomidine on cardiac electrical activity and ion currents: an experimental animal study. Turkish Journal of Biochemistry. 2025. doi:10.1515/tjb-2025-0025. Link
Liao Y, Xiang Y, Zheng M, Wang J. DeepMiceTL: a deep transfer learning based prediction of mice cardiac conduction diseases using early electrocardiograms. Briefings in Bioinformatics. 2023;24(3):bbad109. PMC
Kruper S. From data overload to confident interpretation: how advanced analytics improve preclinical telemetry review. Harvard Bioscience (industry publication). May 27, 2026. Link
Swindle MM, Smith AC. Swine in the Laboratory: Surgery, Anesthesia, Imaging, and Experimental Techniques. 3rd ed. CRC Press; 2015.
Shintaku T, Ohba T, Niwa H, Kushikata T, et al. Effects of isoflurane inhalation anesthesia on mouse ECG. Hirosaki Medical Journal. 2015;66(1):1–7. Link
