Browse the corpus

In people, sepsis is the clinical syndrome that results from the host response to severe or widespread infection [, ] and is the most frequent cause of death in human ICUs []. Sepsis resulted in an estimated 11 million human deaths in 2017, equivalent to ∼20% of global mortality []. Accurate equivalent estimates are not available for dogs and cats, but case fatality rates of up to 68% are reported [, , , , ], with higher fatality rates in animals with more severe disease [] or organ dysfunction [, , ]. Rapid diagnosis and early treatment initiation are crucial because worse outcomes are associated with delayed recognition in people [, , , , ]. Although equivalent data are not available for small animals, it is likely that a similar relationship between delayed intervention and fatality exists, with implications for how sepsis is recognized clinically [, , ].

Disease definitions should be biologically relevant, aid understanding of pathophysiology, and map to readily measurable objective parameters, but codifying sepsis to meet the needs of both clinicians and clinical investigators is challenging. Sensitive criteria that ensure all animals with sepsis are identified might be preferable in the clinic, but the consequent overdiagnosis might result in unnecessary treatment, increased risk of adverse drug effects, diversion of resources away from other patients, or premature euthanasia due to perceived poor prognosis. In contrast, more specific criteria ensure that only animals with sepsis are enrolled in clinical trials, yet this approach could limit generalizability to clinical practice and slow progress because a more stringent definition diminishes the available pool of animals eligible for recruitment.

Since 1991, three sets of sepsis definitions have been proposed for human adults and another three for children. The 1991 and 2001 human adult sepsis definitions established levels of sepsis severity (sepsis, severe sepsis, and septic shock) that remained in use until 2016 []. Central to those definitions was the concept of systemic inflammatory response syndrome (SIRS), in which sepsis was defined as the presence of SIRS due to an infection [, ]. This has also been the basis for veterinary definitions of sepsis but is highly context dependent, and the premise that sepsis is an overexuberant pro‐inflammatory response to infection is of questionable validity. In 2016, sepsis in human adults was redefined to enhance the specificity of the associated clinical criteria [, , ], with a focus on the host response rather than the pathogen trigger. The 2016 definitions recommended that use of the SIRS criteria be discontinued and made the term “severe sepsis” obsolete by making identification of organ dysfunction central to sepsis recognition. The 2016 definitions use the Sequential Organ Failure Assessment (SOFA) score and an abbreviated version termed qSOFA (“quick” SOFA) for identification of organ dysfunction. Septic shock was defined as a subset of sepsis in which underlying circulatory and cellular/metabolic abnormalities are profound enough to cause ∼40% mortality in high‐income countries.

For human pediatrics, the first consensus definitions for sepsis were based on identification of SIRS using variations in physiologic parameters that substantially deviated from those considered normal for the age group []. After the 2016 adult definitions were revised, a Pediatric Sepsis Definition Taskforce was convened and posed two questions: (i) In children with infections, what factors are associated with the development of sepsis? (ii) In children with sepsis, what factors help predict mortality? [, ]. The first question was used to identify clinical sepsis predictors and the second to identify sepsis severity criteria. This approach forms the basis for our efforts to establish the evidence base for sepsis definitions in dogs and cats. After publication of the pediatric sepsis definitions in 2022, a large database‐driven effort redefined sepsis in children in 2024. The resulting Phoenix Sepsis Score was derived from analysis of an international, multicenter database of >3 million care encounters and represents the combination of organ dysfunction criteria most discriminating for sepsis and septic shock in children [, ].

Prior to 2016, veterinary definitions of sepsis were aligned with the 1991 and 2001 human consensus definitions of SIRS plus documented or suspected infection [], albeit with some variation in what constituted infection confirmation [, , , , ]. Multiple sets of SIRS criteria have been reported for both dogs and cats, but without consensus regarding which to use [, , , ]. Data from people indicate that SIRS‐based sepsis definitions are sensitive, but nonspecific [], and SIRS‐based veterinary sepsis definitions are likely similarly affected. Indeed, SIRS‐positive status is common in small animals in emergency and primary care settings, with poor prognostic value []. After publication of Sepsis‐3, veterinary definitions were no longer aligned with those in people, limiting the translational utility of human sepsis literature. The publication of consensus definitions by the Brazilian Veterinary Emergency and Critical Care Society [] also introduced geographic variation in sepsis definitions within small animal medicine.

On this basis, we believed that formally defining sepsis in small animals was warranted to inform clinical care, improve patient outcomes, and facilitate future research []. Here, we report the outcomes of our efforts to establish a consensus definition of sepsis and the results of our systematic review of the available veterinary evidence. We present our recommendations for the use of parameters that can be measured or assessed to aid identification of potential small animal sepsis cases in veterinary hospitals. Limitations of the available data, variation among patient populations, hospitals, and settings, and variation among parameters measured in different laboratories preclude us from making detailed recommendations for cutoff values. Pending further research to evaluate and compare specific cutoffs for objective parameters, clinicians will need to continue to interpret identified abnormalities in the context of individual patient assessments. We have also made suggestions, where appropriate, for the data that we believe should be collected by future veterinary sepsis studies. Although informed by a systematic review, our recommendations represent expert opinion and should be refined as more data become available.

Sepsis is a life‐threatening syndrome associated with a dysregulated host response to infection, resulting in organ dysfunction. There are four key elements in this consensus definition. Sepsis represents a severe condition that is associated with morbidity and increased mortality relative to that expected for an uncomplicated infection, even when that infection is itself serious or difficult to treat. This increased mortality derives from the organ dysfunction that is the result of the host response to the infection. Without infection, there cannot be sepsis, but infection alone is insufficient. The clinical syndrome of sepsis is characterized by organ dysfunction that can be local to the site of infection or distant, and that occurs because of an aberrant, dysregulated, or excessive host response. This updated definition puts organ dysfunction at the center of the pathophysiology of sepsis, making it a requirement for identification of the syndrome and key to successful management of the disorder.

In dogs and cats with infection ( P ), does assessment of SIRS criteria (temperature, heart rate, respiratory rate, leukocyte count) ( I ), compared to not assessing SIRS criteria ( C ), improve prediction of severity (e.g., development of sepsis or new organ dysfunction) or nonsurvival ( O )?

Ten studies (six high quality [, , , , , ], three moderate quality [, , ], one low quality []) provided evidence that supports the use of SIRS criteria. Eighteen studies (eight high quality [, , , , , , , ], seven moderate quality [, , , , , , ], three low quality [, , ]) were not supportive of the use of SIRS criteria. Twenty studies were considered irrelevant to the PICO question [, , , , , , , , , , , , , , , , , , , ].

The evidence supporting the use of SIRS criteria for identifying negative outcomes in dogs and cats with infection is limited and conflicting. Timing of SIRS status adjudication was inconsistent, often unreported, and potentially confounded by administered treatments. Typically, studies of dogs reporting SIRS criteria required satisfaction of at least two of four criteria, whereas studies of cats required that two or three of the four criteria be satisfied. Respiratory rate in panting dogs was inconsistently handled, and there was no uniformity in the cutoffs used for each criterion. Few articles considering SIRS in dogs and cats employed methods or reporting styles that enabled the answering of the PICO question, typically because SIRS‐positive status was an inclusion criterion, infectious and noninfectious causes of SIRS were compared, or a healthy control population was used. The incidence of SIRS varied significantly across the studies considered and between studies evaluating the same disease and reporting similar mortality rates, which could serve as a proxy for patient severity. Conversely, studies with substantial differences in mortality reported similar SIRS rates, which raises questions about the value of SIRS as a robust prognostic marker. Studies supporting the use of SIRS criteria were often matched by similar studies that were not supportive of the use of SIRS criteria. Although a formal meta‐analysis cannot be easily performed, the overall number of patients evaluated in the nonsupportive studies exceeds that of the patients in the supportive studies.

Only three studies considered SIRS in the context of canine sepsis induced by various infection types. In their retrospective cohort study, Hauptmann and colleagues examined the use of SIRS criteria to identify patients with sepsis, defined as an infection plus systemic illness, among a referral surgical population of dogs, most of which did not have underlying infections. In this population, SIRS criteria were sensitive but not specific for identifying sepsis. The criteria for “systemic illness” were ill‐defined, making it difficult to determine if this was an appropriate patient‐centered outcome. A similar retrospective cohort study by Okano and colleagues evaluated SIRS criteria in a general referral canine population and found that SIRS status was associated with survival to discharge, with progressively increasing mortality as more SIRS criteria were fulfilled. However, the underlying diseases were not detailed, precluding interpretation of the results in the context of infection. In a single‐center prospective observational study, Camargo Jr. and colleagues found an association between SIRS status and survival to discharge, but only when SIRS status was evaluated alongside hypoalbuminemia and hypotension; SIRS status was not independently linked to outcome. Similarly, in canine parvovirus infection (CPV), SIRS status is associated with survival to discharge only when considered in conjunction with mucous membrane color and capillary refill time. In dogs with CPV or pyometra, an SIRS‐positive status is associated with worse survival to discharge and longer hospitalization in some, but not all, studies. In dogs with babesiosis, SIRS‐positive status is associated with higher serum concentrations of canine pancreatic‐specific lipase, indicating a potential connection with organ damage, yet SIRS status was not associated with a clinical pancreatitis diagnosis. No association was found between SIRS status and multiple organ dysfunction or mortality in studies of dogs with babesiosis. In dogs with pyometra, SIRS‐positive status was not linked to organ dysfunction incidence, markers of liver injury or kidney dysfunction, or concentrations of lactate or cardiac troponin I (cTnI). In dogs, presence of SIRS status was not associated with outcome in septic peritonitis, pneumonia, or leptospirosis and does not predict the risk of recurrent septic peritonitis in dogs.

Only one study in cats was identified that was partially supportive of the use of SIRS criteria, wherein SIRS‐positive status was associated with higher mortality in cats admitted to an emergency service. Notably, the study population encompassed various etiologies, suggesting that not all cats had infection and reducing the relevance of the results to the PICO question. Other studies were not supportive of the use of SIRS criteria to identify negative outcomes in cats. For instance, in critically ill cats with both infectious and noninfectious diseases, the number of SIRS criteria fulfilled did not correlate with survival to discharge. Similarly, SIRS status was not associated with outcome in critically ill cats with infection, with septic peritonitis, or after gastrointestinal surgery.

In summary, for dogs with infection, evidence supporting the use of SIRS criteria to identify negative outcomes is limited and weak. For cats, there is no evidence to support the use of SIRS criteria for identifying negative outcomes. If sepsis is a syndrome associated with life‐threatening infection, then using SIRS criteria to identify sepsis in clinical settings cannot be recommended. In clinical settings, individual SIRS variables can still be evaluated to identify patients at risk who should be screened for infection and sepsis, provided the potential influence of context, species, and underlying disease process is accounted for. Moreover, individual SIRS variables are likely to be relevant in the context of organ function assessment and might provide helpful information for patient monitoring and for assessing response to treatment when serially evaluated over time. To facilitate future research, we recommend that SIRS variables and SIRS status still be reported in studies evaluating dogs and cats with infection or sepsis to enable comparisons among studies, future meta‐analyses, and scoring system development.

In dogs and cats with infection, use of the SIRS criteria is not required for the diagnosis of sepsis. We suggest measurement of heart rate, respiratory rate, and body temperature at presentation and serially during hospitalization in clinical and research settings. We recommend that SIRS criteria be reported in all future studies of dogs and cats with infection. We recommend that raw data or summary statistics for heart rate, respiratory rate, core temperature, leukocyte, and band neutrophil counts be provided in all future studies of dogs and cats with infection.

In dogs and cats with infection ( P ), does assessment of organ dysfunction or illness severity (e.g., SOFA, APPLE, SPI2) ( I ), compared to not assessing organ dysfunction or illness severity ( C ), improve prediction of severity (e.g., development of sepsis or new organ dysfunction) or nonsurvival ( O )?

Twenty‐four studies (one high quality [], 23 moderate quality [, , , , , , , , , , , , , , , , , , , , , ]) provided evidence that supports the use of organ dysfunction or illness severity assessments in dogs or cats. Ten studies (all moderate quality [, , , , , , , , , ]) were not supportive of the use of organ dysfunction or illness severity assessments in dogs or cats.

Overall, the 24 supportive studies provide a consistent and robust indication of the clinical and prognostic value of assessments of organ dysfunction and illness severity in dogs and cats with infection. Studies using the modified SOFA or validated structured illness severity assessment scores like the Acute Patient Physiologic and Laboratory Evaluation (APPLE) in rapid (APPLE fast ) or complete (APPLE full ) forms also showed consistent associations with mortality across a range of infectious conditions, including septic shock, babesiosis, CPV, leptospirosis, and pyometra. Recent studies demonstrate that both score evolution over time and combinations of scoring with biomarkers can improve prognostication. For instance, presepsin, angiopoietin‐2 (Ang‐2), and procalcitonin (PCT) might offer utility as complementary tools to illness severity scores, even if they cannot currently be recommended as standalone diagnostic criteria. Some markers, like quick SOFA (qSOFA) alone, were less reliable, showing poor sensitivity in multiple studies. In studies that did not apply formal scoring tools, specific markers of organ dysfunction (e.g., creatinine, bilirubin, respiratory failure, thrombocytopenia) were consistently associated with outcome. Data specific to cats are more limited and less consistent, limiting the ability to draw strong conclusions for this species.

Our consensus definition makes identification of organ dysfunction central to establishing a diagnosis of sepsis. To facilitate that, we have constructed a table of potential indicators of organ dysfunction to enable clinicians and researchers to more readily identify patients with sepsis (Table ). The parameters in the table were based on a consensus of expert opinion and will need prospective validation. In addition to requiring identification of organ dysfunction for sepsis diagnosis, we recommend incorporating assessments of organ dysfunction or illness severity, especially APPLE fast , APPLE full , and, where validated, the survival prediction index‐2 (SPI2) or Simplified Acute Physiology Score (SAPS) into clinical decision‐making and research protocols for both dogs and cats. Illness severity scores help stratify risk, identify severely affected patients earlier, and enable therapeutic approach optimization. We envision that their use complements organ function assessments, improves consistency of care provision, and facilitates comparisons between clinical studies. Although evidence supporting structured illness severity scores in cats is more limited, it is neither practical nor conceptually appropriate to apply different sepsis definitions to dogs and cats. As such, given that our definition of sepsis encompasses organ dysfunction as a core principle, it is essential that the clinical criteria recommended to identify cats with sepsis are aligned with that.

In dogs and cats with infection, assessments of organ dysfunction are required for the diagnosis of sepsis. We recommend using a validated, structured illness severity assessment score (e.g., APPLE) to support the diagnosis of sepsis or the assessment of illness severity in clinical and research settings. We suggest using alternative scoring systems (e.g., SPI2, SAPS, SOFA) to support the diagnosis of sepsis, the assessment of illness severity, and the monitoring of patient status over time in clinical and research settings.

At the worksheet stage, two PICO questions were considered individually, and separate guidelines were drafted. During the drafting of this final guidance document, it became necessary to present these together, and the two sets of recommendations have been amalgamated and homogenized in this section. Some variables were relevant and, therefore, considered in both PICO questions (see the Supporting Information for additional information about parameters assessed). Here, the PICO questions, evidence summaries, and discussions are presented sequentially with the combined guidelines duplicated at the end of sections 4.3 and 4.4. Question 4.3 encompassed physiologic variables that could be objectively measured, such as arterial blood pressure, central venous oxygen saturation, cardiac output, stroke volume variation, or urine output.

In dogs and cats with infection ( P ), does measurement of objective physiological parameters ( I ), compared to not measuring objective physiological parameters ( C ), improve prediction of severity (e.g., development of sepsis or new organ dysfunction) or nonsurvival ( O )?

Thirty‐seven studies (nine high quality [, , , , , , , , ], 28 moderate quality [, , , , , , , , , , , , , , , , , , , , , , , , , , , ]) provided evidence that supports the use of objective physiological parameters in dogs or cats. Two studies (both moderate quality [, ]) were not supportive of the use of objective physiological parameters in dogs. These nonsupportive studies either did not evaluate relevant physiological markers or lacked structured physiological analysis.

Most reviewed studies provided evidence supporting the clinical utility of objective physiological parameters for illness severity assessment and outcome prediction in dogs and cats with infection. Frequently evaluated parameters included: Blood lactate concentration ( n = 14). Frequently associated with mortality, particularly when persistently increased. Arterial blood pressure ( n = 11). Hypotension (lower mean or systolic arterial pressure) is repeatedly linked to increased mortality, azotemia, shock progression, and organ dysfunction. Serum albumin. Hypoalbuminemia is associated with nonsurvival in studies of surgical sepsis (e.g., peritonitis, pyometra). Temperature, heart rate, and respiratory rate. Often correlated with nonsurvival, particularly when two or more parameters are deranged. Echocardiographic variables. Diastolic and systolic dysfunction markers are strongly associated with mortality in dogs and cats with sepsis. Blood biomarkers. Several have been associated with survival in cats and dogs with sepsis.

Although there was considerable heterogeneity in study design, parameter type, measurement method, and the diseases represented within the reviewed studies (e.g., CPV, babesiosis, pneumonia, surgical peritonitis, and pyometra), physiological derangements, particularly when interpreted serially or in combination, were consistently predictive of outcome. Frequently, serial measurements (e.g., lactate clearance, blood pressure recovery), structured scores (e.g., canine infection mortality score, CIMS), or combinations of scores with laboratory data had superior prognostic accuracy than single variables. The reviewed evidence supports integration of physiological parameters into scoring systems, protocols, and research methodologies to improve prediction, clinical decision‐making, and study comparability.

It should be noted, however, that while objective physiological parameters are useful for sepsis diagnosis in infected animals, they are insufficient in isolation. Hypotension, hyperlactatemia, abnormal core temperature, and altered mentation may aid clinical recognition of sepsis but are overly sensitive and too nonspecific to be used alone. Abnormalities in these parameters frequently occur in noninfectious critical illnesses and other systemic disease processes. While several studies confirmed the association between these parameters and sepsis‐related organ dysfunction or mortality, few evaluated their utility for the diagnosis of sepsis. Additional research in these areas and on the identification of the optimal combinations of parameters and biomarkers is still needed, combined with broad validation and standardization across diverse clinical settings. In summary, physiological parameters are valuable components of sepsis identification protocols but should be interpreted in combination with clinical context and confirmation of infection to maximize diagnostic accuracy.