This study examines approaches to assessing the risk of death and predicting the development of sepsis in dogs. Sepsis remains one of the key clinical problems in medicine and veterinary medicine, due to its widespread prevalence throughout the world, leading to high mortality. Sepsis is based on the reaction of systemic inflammation of the body, which is realized in the presence of signs of SIRS, bacteremia and/or an infectious focus of various nature in combination with acute signs of organ dysfunction and/or evidence of microbial dissemination. Since sepsis develops only progressively, it is impossible for the animal to recover independently without providing it with timely and appropriate medical care. Today, one of the key issues remains improving approaches to assessing the risk of death and predicting the development of sepsis in dogs. An attempt to combine the pathogenicity factors of an infectious agent and the failure of the host’s defense mechanisms is reflected in the concepts of PIRO, SAPS and MODS, which evaluate the links in the etiopathogenesis of sepsis. Thus, heuristic approaches will allow us to come closer to the development of pathogenetically based methods for assessing the risk of death of animals and predicting the development of sepsis in dogs. These approaches are based on criteria for assessing PIRO indicators: predisposition (P), including breed, sex, age of the animal; infection (I) associated with the type of pathogen detected; systemic response of the body to infection R, characterized by the main clinical indicators - body temperature, heart rate, respiratory rate and the number of leukocytes in the blood; organ dysfunction (O) identified in one or more organs of the animal. An important role in the objective prognosis is given to the evaluation criteria on the SAPS scale, taking into account changes in indicators towards hyperinflammation (SIRS) or immunosuppression (MODS), which determine the stage of development of sepsis and the likelihood of the risk of death of the sick animal. On average, the predictive accuracy of quantifying disease severity systems is 70–85%. An algorithm for translating clinical and laboratory data into a SAPS scale of indicators (scores) that determines the severity of the disease in leptospirosis in dogs is proposed. It has been shown that on the SAPS scale for canine leptospirosis, a score of 29