Association between hyperglycemia and survival in 228 horses with acute gastrointestinal disease

• Published in: Journal of veterinary internal medicine 2009-11, Vol.23 (6), p.1261-1265
• Main Authors: Hassel, D.M, Hill, A.E, Rorabeck, R.A
• Format: Article
• Language: English
• Subjects: Animals; Colic; digestive system diseases; Equine; Female; Gastrointestinal Diseases - blood; Gastrointestinal Diseases - mortality; Gastrointestinal Diseases - veterinary; Glucose; Horse Diseases - blood; Horse Diseases - mortality; Horses; hyperglycemia; Hyperglycemia - blood; Hyperglycemia - veterinary; Male; mortality; Prognosis; Retrospective Studies; Risk factors
• ISSN: 0891-6640; 1939-1676
• DOI: 10.1111/j.1939-1676.2009.0395.x

Background: Glucose homeostasis is dysregulated in critically ill humans resulting in hyperglycemia and decreased survival. Hyperglycemia is common in horses presenting with abdominal crisis, and this might be associated with a worse prognosis for survival. Objective: To determine if hyperglycemia in horses with acute abdominal disease is associated with increased odds of failure to survive to hospital discharge. Animals: Two hundred and twenty‐eight adult horses with acute gastrointestinal disease. Methods: Observational retrospective study. Records of horses > 1 year of age presenting for treatment of colic over a 3‐year period were reviewed. Data collected included age, duration of colic, glucose, heart rate, PCV, total protein, anion gap, cost of hospitalization, breed, sex, pain at admission, diagnosis, whether surgery was performed, and life status at hospital discharge. Potential risk factors for nonsurvival were screened by univariable logistic regression and the best‐fitting univariable model was used as the basis for multivariable regression modeling. Results: Mean blood glucose was 155 mg/dL (8.5 mM) with 45% of the population above the reference range; 16.7% (38 of 228) of horses had severe hyperglycemia (>195 mg/dL; 10.7 mM). Factors associated with increased odds of failure to survive included glucose, severity of pain at admission, heart rate, PCV, anion gap, and diagnosis. The best‐fitting multivariable model included glucose and diagnosis, with age included as a confounding variable. The model correctly classified outcome for 92.5% of horses. Conclusions and Clinical Importance: This study has confirmed prior reports that hyperglycemia is common in horses with colic and is associated with a worse prognosis for survival to hospital discharge.