Hey everyone, welcome back to Wheeler the Healer: Small Animal Emergency and Critical Care—practical ECC education to keep you sharp when it matters most.
My name is Lance Wheeler, and I’m a small animal emergency and critical care specialist.
In this episode, we’re continuing to build on the framework from last time. In Episode 12, we talked about suffering in the ICU and reframed comfort as both an ethical and physiologic priority. We discussed how dyspnea, pain, nausea, anxiety, sleep deprivation, and environmental stress can directly influence patient stability and outcomes. Now we’re shifting into something that sits at the center of ECC decision-making every single day: clinical prediction.
The material for today is inspired by Small Animal Critical Care Medicine, 3rd edition, specifically Chapter 13, “Predictive Scoring Systems in Veterinary Medicine,” by Galina Hayes and Karol Mathews. I’m also going to weave in several newer veterinary studies evaluating systems like the Animal Trauma Triage score, modified Glasgow Coma Scale, APPLE scores, qSOFA-based models, hemorrhage prediction systems like TASH, bleeding assessment tools like DOGiBAT, and disease-specific prediction calculators like HeLP and T-STAT.
Why should an ECC clinician care about predictive scoring systems? Because nearly every important decision we make on ER and ICU shifts involves uncertainty.
- Is this trauma patient actually stable, or are they compensating right before they crash?
- Is this mentation change true traumatic brain injury, or is this patient just severely hypoperfused?
- Is this patient quietly developing occult hemorrhagic shock while still maintaining blood pressure?
- Is this patient sick enough that I need to push for ICU-level monitoring or more aggressive intervention?
- Does this patient needs blood products?
- And how do I communicate all of that risk objectively to the rest of the team and to owners?
And the reality is this—whether we formally calculate a score or not, we’re already making predictive judgments constantly. Every time you say a patient “looks worse,” “looks stable,” or “looks guarded,” you’re already doing predictive medicine. These tools simply try to standardize and quantify that process. But there’s also a trap here. One of the biggest mistakes clinicians make is assuming that because a scoring system is statistically significant, it must therefore be clinically decisive. And those are not the same thing.
A patient with a terrible score can survive. A patient with a reassuring score can still die.
A model can perform beautifully across thousands of patients and still perform imperfectly in the patient standing in front of you at 2 AM while you’re trying to decide whether to transfuse, intubate, operate, or call the owner with a prognosis update. So the goal today is to walk through how these scoring systems actually work, what physiologic processes they’re trying to measure, what outcomes they were designed to predict, where they succeed, where they fail, and how to apply them in real-time ECC medicine without overinterpreting them.
As I call out key numbers or define important terms, take a second to repeat them to yourself. That repetition helps reinforce what you already know and lock in the details that are most testable. At the end, we’ll review them again during the rapid-fire board review so you can check your recall and reinforce them one more time.
Before we discuss specific scoring systems, we first need to understand something fundamentally important about trauma and critical illness physiology. Many patients are not killed primarily by the original injury or disease process. They’re killed by the secondary injury cascade that follows afterward. This is one of the most important conceptual frameworks in trauma and critical care medicine.
The primary injury may involve hemorrhage, fractures, pulmonary contusions, traumatic brain injury (or TBI), gastric dilatation-volvulus, pancreatitis, or severe hypoxemia.
But afterward, the body launches a massive systemic response. You can see systemic inflammatory response syndrome (or SIRS), endothelial injury, glycocalyx shedding, ischemia-reperfusion injury, coagulopathy, mitochondrial dysfunction, microcirculatory failure, oxygen debt, and eventually progression into multiple organ dysfunction syndrome, or MODS.
That should sound familiar from several earlier episodes in this series. We’ve already talked about SIRS, endothelial surface layer injury, oxygen toxicity, shock physiology, and interstitial edema. These models are really trying to quantify the physiologic consequences of those processes.
In other words, many scoring systems are not truly measuring how bad the injury looks.
They’re indirectly measuring how close the patient is to systemic physiologic collapse.
And those are not always the same thing.
One of the recent veterinary trauma papers supporting this idea was published by Neimann and colleagues in 2025. They prospectively evaluated dogs and cats with polytrauma and found that increasing Animal Trauma Triage, or ATT, scores strongly correlated with both mortality and MODS development. More specifically, every one-point increase in ATT score above 8 approximately doubled the odds of death and increased the odds of MODS development nearly sixfold. ATT also demonstrated extremely high discriminatory ability for predicting non-survival, with area under the receiver operating characteristic curve, or AUROC, values approaching 0.98.
That’s a major finding because it reinforces the idea that multisystem dysfunction matters more than isolated injury severity.
The ATT score evaluates six major physiologic systems: perfusion, cardiac, respiratory, eye-muscle-integument, skeletal, and neurologic abnormalities. Each category is scored from 0 to 3, creating a total possible score ranging from 0 to 18, with higher scores indicating more severe overall trauma. Earlier validation studies by Rockar, Ash, and others demonstrated strong associations between increasing ATT scores and mortality risk in traumatized dogs, with ATT scores around 9 out of 18 historically corresponding to roughly 50% survival. Know that number.
And physiologically, this makes sense. Polytrauma patients rarely die from isolated orthopedic injury. They die from combinations of shock, hemorrhage, hypoxemia, coagulopathy, inflammatory organ dysfunction, and secondary organ failure. ATT captures that broader physiologic burden reasonably well.
Now let’s contrast ATT with the modified Glasgow Coma Scale, or mGCS. The mGCS is fundamentally a neurologic severity score. It evaluates motor activity, brainstem reflexes, and level of consciousness. Each category receives a score from 1 to 6, generating a total score ranging from 3 to 18, with lower scores indicating more severe neurologic dysfunction. Historically, studies evaluating TBI suggested that mGCS scores of 8 or lower corresponded to roughly 50% survival.
But one of the most important modern veterinary studies evaluating mGCS was a 2026 VetCOT registry study by Tucker and colleagues. They evaluated over 9,600 traumatized dogs across multiple institutions and found that survival steadily declined as mGCS scores worsened. Dogs with normal scores of 18 had survival rates near 98%, while dogs with abnormal neurologic scores had substantially lower survival overall—around 68%—with mortality and euthanasia rates increasing progressively as neurologic dysfunction worsened.
But the most important finding from the paper was not simply that worsening neurologic scores correlated with worse outcome. It was that low mGCS scores did not reliably correspond with confirmed TBI. In other words, many trauma patients with abnormal mentation were not necessarily suffering from primary intracranial disease.
That’s a huge bedside lesson because altered mentation in a trauma patient is not automatically synonymous with brain injury. Shock can alter mentation. Hypoxemia can alter mentation. Hypercapnia can alter mentation. Severe pain can alter mentation. Sedatives can alter mentation. Intoxications can alter mentation. Post-ictal states can alter mentation. Sepsis can alter mentation. Metabolic encephalopathies can alter mentation.
So clinicians have to avoid anchoring bias. A depressed or abnormal neurologic exam in a trauma patient should trigger a broad physiologic assessment, not an automatic assumption of TBI. And this is one reason ATT often outperforms mGCS for generalized trauma prognosis—because ATT captures the broader systemic physiologic picture rather than focusing primarily on neurologic status.
The Tucker VetCOT paper also identified several interesting epidemiologic findings. Blunt trauma was much more associated with abnormal mGCS scores than penetrating trauma. Younger dogs and smaller dogs were also more likely to have abnormal neurologic scores, and euthanasia frequency rose dramatically as neurologic dysfunction worsened. Mechanism matters clinically. Blunt trauma is more likely to produce both diffuse systemic injury and acceleration-deceleration forces affecting the brain itself. These patients may develop pulmonary contusions, hemorrhagic shock, endothelial injury, reperfusion injury, inflammatory dysfunction, hypoxemia, and secondary organ dysfunction—all of which can alter neurologic status. In contrast, penetrating trauma may create devastating focal injury while preserving other organ systems relatively well.
Now let’s zoom out and discuss one of the most important conceptual frameworks of the episode: different prediction tools are trying to predict completely different outcomes, and clinicians misuse them constantly.
ATT primarily predicts trauma severity and mortality risk. mGCS predicts neurologic injury severity. Acute Patient Physiologic and Laboratory Evaluation, or APPLE, scores estimate global illness severity. Quick Sequential Organ Failure Assessment, or qSOFA, attempts rapid bedside deterioration screening. TASH predicts transfusion requirement. DOGiBAT quantifies bleeding severity. HeLP and T-STAT estimate malignancy probability in dogs with splenic masses.
These systems are not interchangeable. A good hemorrhage prediction score is not automatically a good mortality predictor. A neurologic severity score may perform poorly for MODS prediction. A triage score may perform poorly for long-term outcome prediction. You have to understand what the score was actually designed to predict.
Now let’s move into APPLE scores. APPLE stands for Acute Patient Physiologic and Laboratory Evaluation. APPLEfast incorporates variables like mentation, temperature, blood pressure, lactate, and packed cell volume, while APPLEfull adds additional laboratory variables and body cavity fluid assessment.
The original APPLE studies by Hayes and colleagues demonstrated that increasing APPLE scores correlated with increasing mortality risk in critically ill dogs. But this is another nuance that’s easy to miss: APPLE is probably more valuable as a research stratification tool than as a bedside crystal ball.
One of the major purposes of APPLE scoring is to objectively quantify baseline illness severity in research populations. If one ICU study enrolls dramatically sicker patients than another study, APPLE helps quantify that difference objectively. That’s incredibly important scientifically. But clinicians sometimes overinterpret these scores clinically.
For example, investigators evaluating cats with first-time diabetic ketoacidosis in 2024 found that APPLEfast scores were significantly higher in non-survivors than survivors and that mortality risk increased as APPLEfast scores increased. But there was still substantial overlap between survivors and non-survivors. That overlap matters because these tools shift probabilities—they do not determine destiny.
And this highlights another subtle but extremely important statistical concept. Predictive models are usually much better at predicting what happens across large groups of patients than they are at predicting exactly what will happen to one individual patient.
A scoring system may correctly show that dogs with very high APPLE scores die more often overall than dogs with low APPLE scores. But when you apply that same model to the single dog standing in front of you, uncertainty becomes much larger. One patient may survive despite a terrible score, while another may die despite a reassuring one. That’s what statisticians mean when they talk about wider confidence intervals around individual risk prediction.
In other words, predictive models are often better at describing population trends than predicting individual destiny. That’s one reason prognostic scores should support clinical reasoning—not replace it.
And this has a very important clinical and ethical implication. Severity scores should never be used in isolation to justify withdrawal of care, withholding treatment, or making irreversible decisions for an individual patient. A prognostic score may identify increased statistical risk, but it cannot fully account for factors like response to therapy, reversibility of disease, financial limitations, owner goals, clinician experience, or the unique biology of the individual patient.
In general, these models are far more reliable when applied to research populations than individual patients. Their greatest clinical value is usually helping clinicians communicate risk more objectively, support team discussions, guide realistic owner expectations, and improve informed consent.
Every scoring system makes tradeoffs. And one of the biggest tradeoffs in clinical prediction tools is speed versus precision. More comprehensive systems like APPLE can provide much richer physiologic information, but they require laboratory testing, multiple variables, more time, and more resources to calculate. That limits their usefulness during early triage or rapidly evolving emergencies. Simpler systems like qSOFA sacrifice some detail in exchange for speed and practicality. qSOFA uses just three variables—altered mentation, hypotension, and tachypnea—to quickly identify patients at increased risk of deterioration. And in an emergency setting, that simplicity is often its greatest strength.
A 2025 veterinary study comparing qSOFA, qSIRS, APPLEfast, qSOFA-lactate, and qSOFA-CRP found that modified qSOFA systems outperformed several conventional scoring systems for predicting 28-day mortality in critically ill dogs. The qSOFA-CRP model performed especially well using a C-reactive protein cutoff above 1.55 mg/dL and achieved an AUROC around 0.76. Importantly, the model also demonstrated high sensitivity and high negative predictive value, making it potentially useful as a triage screening tool for identifying patients at lower risk of mortality. Meanwhile, qSOFA-lactate also outperformed conventional qSOFA and APPLEfast models. That reinforces the prognostic importance of tissue hypoperfusion markers.
Now this brings us directly into another major lesson: external validation matters.
One of the biggest mistakes in medicine is assuming that because a model worked well in one study population, it will automatically work equally well everywhere else. It often doesn’t.
One of the best veterinary examples comes from the HeLP and T-STAT splenic mass prediction studies. These were predictive calculators developed to help clinicians estimate the likelihood that a dog with a splenic mass had hemangiosarcoma or another malignant process before surgery.
The HeLP and T-STAT models both combined signalment, laboratory data, physiologic variables, and imaging findings to estimate malignancy risk in dogs with splenic masses. The specific variables differed somewhat between models, but both attempted to predict the likelihood of hemangiosarcoma or other malignant disease before surgery.
Initially, these models appeared promising in their original study populations. But when investigators later tested them in different patient populations through external validation studies, performance deteriorated substantially. The HeLP score still showed acceptable discrimination, with AUROC values around 0.78, while T-STAT performed poorly in that validation population.
And this is one of the most important lessons in predictive modeling: these tools almost always perform better in the populations where they were originally created than they do in external real-world populations.
And this highlights another major principle in predictive modeling: the true test of a scoring system is not how well it performs in the original training population where it was created, but whether it retains predictive accuracy in entirely different patient populations, hospitals, and clinical environments.
And here’s why.
Performance estimates from derivation studies should always be interpreted cautiously because predictive models almost always look best in the populations used to build them. Ideally, a scoring system should maintain reasonable predictive utility when applied to noncontemporary patients at different institutions.
And clinically, that means we should always ask: are the patients used to validate this model actually similar to the patients I’m treating in my own ER or ICU?
Now let’s briefly talk about how predictive models are evaluated statistically, because this is extremely board relevant.
One of the most common measures is the area under the receiver operating characteristic curve, or AUROC. AUROC measures discrimination—basically, how well a score separates patients with an outcome from patients without the outcome.
An AUROC of 0.5 means the model performs no better than random chance. An AUROC between 0.7 and 0.8 is generally considered acceptable. Between 0.8 and 0.9 is considered excellent. And above 0.9 is considered outstanding.
Conceptually, the receiver operating characteristic curve plots sensitivity against specificity across different score cutoffs. An ideal model pushes toward the upper left corner of the graph, where sensitivity and specificity are both high.
Another useful way to think about AUROC is this: if a model has an AUROC of 0.8 for mortality prediction, that means that when comparing a survivor and a non-survivor, the model has roughly an 80% chance of assigning a higher predicted risk to the non-survivor.
But AUROC also has important limitations. When outcomes are rare, AUROC can sometimes look artificially impressive. For example, imagine an ICU where only 5% of patients die and 95% survive. A completely useless model that simply predicts survival for every single patient would still appear correct 95% of the time overall because most patients survive anyway.
But clinically, that model is terrible. It completely fails at the one thing we actually care about: identifying which patients are truly at risk of dying.
And that’s the danger. When nonevents greatly outnumber events, overall statistical performance can look deceptively strong even when the model has very little real-world clinical usefulness.
And that leads into another important concept: calibration. Discrimination asks whether the model can separate high-risk from low-risk patients. Calibration asks whether the predicted probabilities actually match what happens in real patients.
For example, in a perfectly calibrated model, if 10 patients are predicted to have an 80% mortality risk, about 8 of them should actually die. A prediction tool can discriminate reasonably well while still being poorly calibrated. That distinction matters because clinicians often use these scores to estimate real-world probabilities during owner communication and decision-making.
And there are several reasons predictive performance can deteriorate outside the original study population
Why? Because of overfitting, referral bias, population differences, clinician variability, hospital differences, spectrum bias, and changing disease prevalence. This is why external validation matters so much, and it’s why many calculators disappoint clinicians in real practice.
Now let’s talk about hemorrhage prediction. One of the major challenges in trauma medicine is occult hemorrhage recognition because compensatory shock physiology can temporarily preserve blood pressure despite severe blood loss.
The patient may initially look “stable” while accumulating oxygen debt and worsening tissue hypoperfusion.
A 2024 study evaluating the human Trauma-Associated Severe Hemorrhage score, or TASH score, in traumatized dogs found that dogs requiring transfusion had significantly higher TASH scores than dogs that did not require transfusion. The score incorporates hemoglobin concentration, systolic blood pressure, abdominal fluid, heart rate, base excess, fractures, and sex. In this study, a TASH score above roughly 7 to 8 points on the 28-point scale predicted increased likelihood of transfusion requirement.
But the most important lesson wasn’t simply that the score worked. It was which variables emerged as especially predictive: base excess, abdominal fluid, and shock physiology. Those variables are basically physiologic fingerprints of occult hypoperfusion. That’s the real clinical pearl.
Not all shock is obvious.
Compensatory mechanisms may temporarily preserve blood pressure despite severe hemorrhage, allowing occult shock to progress unnoticed unless clinicians integrate perfusion markers like lactate, base excess, abdominal fluid assessment, heart rate trends, and serial reassessment.
Now let’s briefly discuss DOGiBAT. The Daily Canine Bleeding Assessment Tool was developed by Makielski and colleagues in 2018 to standardize assessment of bleeding severity in dogs with immune thrombocytopenia across multiple organ systems. DOGiBAT scores range from 0 to 24, with higher scores indicating more severe bleeding. Unlike some of the other scoring systems we’ve discussed, DOGiBAT was not designed around a strict prognostic cutoff. However, dogs requiring transfusion tended to have substantially higher DOGiBAT scores, with median scores around 5 out of 24 compared with scores around 2 out of 24 in non-transfused dogs.
And this matters because platelet count alone correlates imperfectly with clinically important bleeding risk. And that reinforces another core ECC principle: single numbers rarely tell the whole story. The patient’s clinical phenotype matters. What is the patient actually doing? That question often matters more than any isolated laboratory value.
Another extremely important board-relevant concept is that scoring systems should not automatically trigger interventions simply because a cutoff value is reached. A score can identify increased risk, but whether interventions based on that score actually improve outcome still has to be demonstrated in properly designed clinical trials. That’s a subtle but extremely important distinction between prognostic association and intervention efficacy.
Here’s what I mean by that. Imagine a trauma study finds that dogs with an ATT score above 8 have dramatically increased mortality. That does not automatically mean that every dog with an ATT score above 8 should receive a specific intervention, like automatic transfusion, surgery, vasopressors, or ICU admission. The score tells us those patients deserve closer attention. It does not prove that any particular intervention changes that risk.
Or think about lactate. We know higher lactate concentrations correlate with worse outcome in many diseases. But that doesn’t mean aggressively forcing lactate normalization always improves survival. Sometimes lactate is acting primarily as a marker of disease severity rather than a directly modifiable therapeutic target.
A prognostic marker may help identify sick patients without necessarily telling us how to fix the underlying problem. Correlation with outcome is not the same thing as proof that intervening on the variable improves outcome.
And clinically, this matters because it helps prevent cookbook medicine. The score should inform the bedside assessment—not override it.
And this leads into another extremely important evidence-based medicine concept: confounding. A confounder is an outside factor that distorts the apparent relationship between a treatment and an outcome. In ECC medicine, one of the most common confounders is illness severity itself.
Here’s a classic example. Imagine a retrospective study finds that dogs receiving vasopressors had higher mortality than dogs that did not receive vasopressors. That doesn’t automatically mean vasopressors are harmful. More likely, the sickest dogs were the ones requiring vasopressors in the first place. The treatment becomes associated with mortality because clinicians selectively apply it to patients already at highest risk of death even if the treatment itself is beneficial. That’s one reason retrospective ECC literature can be so difficult to interpret.
This is called confounding by illness severity, and it’s one reason retrospective observational studies can sometimes produce misleading conclusions. It’s also why well-designed prospective studies and randomized controlled trials are so important when evaluating treatment efficacy.
And this is one of the major reasons illness-severity scores became so important in critical care research. Objective illness-severity scoring allows investigators to control for baseline patient severity.
For example, if one treatment group contains dramatically sicker patients than another, comparing outcomes between those groups becomes misleading. Severity scores like APPLE can help researchers account for those differences statistically. Investigators may use illness severity scores as study inclusion criteria, demonstrate similar baseline severity between treatment groups, or incorporate severity scores into regression analyses to help separate the effects of treatment from the effects of underlying disease severity itself.
In other words, severity scores are not just prognostic tools—they’re also tools for improving the quality and interpretability of clinical research.
This also helps explain why severity scoring remains important even in randomized controlled trials. The goal of randomization is to distribute confounding variables—like illness severity, age, comorbidities, or shock severity—roughly equally between treatment groups. That way, differences in outcome are more likely to reflect the treatment itself rather than baseline differences between patients.
But veterinary studies often involve relatively small patient numbers, heterogeneous diseases, and highly variable illness severity. Because of that, randomization may not perfectly balance the groups. One treatment arm may still end up containing much sicker patients purely by chance.
Severity scores help investigators recognize and statistically control for those imbalances. That reduces the risk of mistakenly concluding that a treatment is harmful or ineffective when the treated patients were simply sicker at baseline.
Now let’s zoom out and discuss something broader: triage medicine itself. Triage medicine is fundamentally resource allocation under uncertainty. That’s true in both veterinary and human medicine. Scoring systems help us decide who needs ICU monitoring, who may need transfusion, who requires aggressive reassessment, and who may decompensate.
They also help prioritize limited resources and communicate risk more objectively.
That becomes especially important during mass casualty situations, busy ER shifts, financial limitations, or limited ICU capacity. These tools help standardize decision-making under pressure. But they still have to be interpreted in the context of the patient in front of you.
And this is where clinicians can get into trouble. A model can achieve statistical significance and still be too inaccurate to confidently guide bedside decisions in individual patients. That’s especially true in ECC medicine, where patients evolve rapidly and physiology changes hour to hour.
So here’s my practical framework for using predictive scoring systems: use scores to structure your assessment, use them for communication, use trends more than isolated numbers, understand what outcome the score was designed to predict, recognize the limitations of external validation, and never let the score override what the patient is actually telling you.
Because the patient—not the calculator—is still the gold standard.
And good clinicians know when the score fits the patient—and when it doesn’t.
Rapid-Fire Board Review
- The ATT score evaluates perfusion, cardiac, respiratory, eye-muscle-integument, skeletal, and neurologic systems.
- ATT score ranges from 0 to 18.
- ATT scores around 9 correspond to approximately 50% survival.
- Each one-point increase in ATT score above 8 approximately doubled odds of death in one prospective polytrauma study.
- Increasing ATT scores strongly correlate with mortality risk and MODS development.
- The modified Glasgow Coma Scale evaluates motor activity, brainstem reflexes, and level of consciousness.
- mGCS scores range from 3 to 18.
- Lower mGCS scores indicate more severe neurologic dysfunction.
- Dogs with abnormal mGCS scores had approximately 68% survival versus nearly 98% survival for dogs with normal scores in a large VetCOT study.
- Abnormal mGCS scores do not necessarily indicate confirmed traumatic brain injury.
- APPLEfast incorporates mentation, temperature, blood pressure, lactate, and PCV abnormalities.
- APPLE scores are especially useful for illness severity stratification in research populations.
- qSOFA uses altered mentation, hypotension, and tachypnea for rapid bedside deterioration screening.
- Higher AUROC values indicate better ability of a model to distinguish patients with versus without the outcome of interest.
- AUROC of 0.5 = no better than random chance.
- AUROC of 0.7–0.8 = acceptable discrimination.
- AUROC of 0.8–0.9 = excellent discrimination.
- AUROC greater than 0.9 = outstanding discrimination.
- Calibration refers to how closely predicted risk matches observed real-world outcomes.
- qSOFA-CRP used a CRP cutoff above 1.55 mg/dL and achieved an AUROC around 0.76.
- The Trauma-Associated Severe Hemorrhage, or TASH, score incorporates hemoglobin, systolic blood pressure, abdominal fluid, heart rate, base excess, fractures, and sex to help predict transfusion requirement in traumatized dogs.
- A TASH score above roughly 7–8 out of 28 points was associated with increased likelihood of transfusion requirement in one veterinary trauma study.
- Base excess and abdominal effusion may help identify occult hemorrhagic shock before hypotension becomes obvious.
- Compensatory shock physiology may temporarily preserve blood pressure despite severe blood loss and worsening tissue hypoperfusion.
- Severity scores should never be used in isolation to justify withdrawal of care, withholding treatment, or making irreversible decisions for an individual patient.
- Many predictive models deteriorate during external validation because derivation cohorts often overestimate real-world performance.
- Most veterinary prognostic systems shift probabilities rather than providing certainty.
- Predictive scoring systems support clinical reasoning, but they do not replace bedside assessment.
Thank you for listening—I really appreciate you spending this time with me.
If you found this helpful, make sure to follow or subscribe so you don’t miss future episodes as we work through core ECC topics together.
If you know someone who’s studying for boards or working in emergency and critical care, feel free to share this with them—this is meant to be something we can all learn from together.
And if you have a minute, leaving a rating or review really helps more people find the podcast—and it also helps me understand what’s most useful so I can continue to adapt and focus on the topics that matter most to you.
You can also check out similar educational content, including podcasts and YouTube videos, from my colleague Igor Yankin on VetEmCrit.
And with that, this concludes our initial 13-episode section covering key critical care concepts and foundational ECC physiology. We’ve spent the last several episodes building frameworks around shock, oxygen toxicity, endothelial injury, edema, temperature regulation, suffering in the ICU, and now predictive scoring systems.
I really appreciate all of you who’ve followed along through this first major section of the series.
In the next episode, we’re going to open an entirely new can of worms as we begin the next major block of the podcast: respiratory disorders.
Episode 14 will also mark another first for Wheeler the Healer. Starting with that episode, the podcast will be available in video format on YouTube in addition to the standard podcast format. So, if you prefer to watch rather than just listen, keep an eye out for the YouTube version. And if you’re sticking with the traditional podcast, don’t worry—I’m going to make sure the content works equally well whether you’re watching the video or listening in the car, at the gym, or wherever else you get your podcasts.
Over the next 17-plus episodes, we’ll start diving into oxygenation, ventilation, upper airway disease, lower airway disease, pulmonary vascular disease, pleural space disease, and all the chaos that comes with managing dyspneic patients in the ER and ICU.
Thanks again for being here, and I’ll see you in the next episode.
Long Days and Pleasant Nights.
