Predictors of clinically significant bleeding in dogs with severe thrombocytopenia: 159 cases (2012‐2021)

Objectives To evaluate risk factors for clinically significant bleeding in dogs with severe thrombocytopenia and to assess whether predictors of bleeding differ between immune thrombocytopenia and other causes, and whether platelet mass is more predictive than platelet count. Materials and Methods Medical records of dogs presenting to a referral institution between 2012 and 2021 with severe thrombocytopenia (platelet count <50 × 10 9 /L) were retrospectively reviewed. Dogs were classified by underlying cause and presence of clinically significant bleeding. Univariable and multivariable logistic regression analyses were performed. Results A total of 159 dogs were included, of which 74 developed clinically significant bleeding. Eighty‐four dogs were classified as having immune thrombocytopenia. Dogs with clinically significant bleeding had lower platelet counts and plateletcrit, and higher leucocyte and monocyte counts. On univariable analysis, immune thrombocytopenia was associated with increased odds of bleeding (odds ratio 3.49, 95% confidence interval 1.83 to 6.85). In the final multivariable model, increasing platelet count was associated with reduced odds of bleeding (odds ratio 0.72 per 10 × 10 9 /L increase, 95% confidence interval 0.55 to 0.95), while increasing monocyte count was associated with increased odds (odds ratio 1.09 per 0.1 × 10 9 /L increase, 95% confidence interval 1.05 to 1.13). Plateletcrit was not independently associated with bleeding. Within the immune thrombocytopenia subgroup, platelet count was not associated with bleeding, whereas monocyte count remained independently associated. Clinical Significance When considering all dogs with severe thrombocytopenia, the lower the platelet count the more likely the dogs are to develop clinically significant bleeding. However, for dogs with immune thrombocytopenia, the relationship between platelet count and the development of clinically significant bleeding is not clear‐cut, suggesting that factors beyond platelet number contribute to bleeding risk in this cohort of thrombocytopenic dogs.