Prognostic indicators in a Danish hospital-based population of colic horses.

The research is about prognosis indicators in colic horses conducted in a Danish hospital, with the aim of formulating a predictive model useful in clinical analysis of horse colic prognosis. In the examination of 528 colic horses from 1994 to 1998 at the Royal Veterinary and Agricultural University of Copenhagen, Denmark, the researchers developed a clinical model that notably utilizes degree of pain, packed cell volume, capillary refill time, and rectal temperature deviations.

Study Methodology

• 528 colic horses from the Royal Veterinary and Agricultural University of Copenhagen, Denmark, were studied over a period from August 1994 to December 1998.
• The research undertook a multivariate logistic regression analysis using 357 of the colic cases to develop a clinical-practical model.
• The model was based on four primary factors including degree of pain, packed cell volume, capillary refill time, and rectal temperature deviation from 38 degrees Celsius.

Key Findings

• An unexpected U-shaped relation between rectal temperature and outcome (survival or death) was discovered, rather than the expected linear relationship. The researchers found this could be translated into a linear association when treated as a deviation from 38 degrees Celsius.
• Other clinical variables and laboratory measurements were found to have a significant correlation with the outcome.
• A recommended cut-off value for deciding between treatment or euthanasia was determined at the predicted probability of 86% mortality rate. This mark had the optimal accurate classification for both survival and death, at 87%. This implies the decision process became critical when the projected probability of death surpassed this cut-off.

Limitations And Recommendations

• The research also disclosed that the number of misclassified survivors (horses that would die despite treatment) increased when the cut-off value for euthanasia decision was moved upwards so as to not unnecessarily euthanize horses.
• For four horses, the outcome was notably poorly predicted, indicated by extreme deviance residuals. In two of these cases, this was due to sudden, severe changes in the disease’s progression.
• The researchers recommend repeated measurements to better the performance of the overall prognosis prediction.

Conclusion

• The study concludes with its main findings that the degree of pain, packed cell volume, capillary refill time, and rectal temperature deviations from 38 degrees Celsius, used in a logistic regression model, provide an effective model for clinical assessment of prognosis in colic horses.