A severity score for spontaneous canine acute pancreatitis

Objective To derive a severity score for spontaneous canine acute pancreati tis applicable to general practice. Design Cohort study of canine pancreatitis cases. Procedure Cases (n = 68) of spontaneous canine acute pancreatitis presented to general practitioners were identified among accessions to Veterinary Pa thology Services Brisbane. The primary veterinarian was surveyed by telepho ne to ascertain the outcome of each case. Scores were assigned for extent o f hyperamylasaemia, hyperlipasaemia and number of organ systems other than the pancreas compromised. The probability of mortality with each score of e ach analyte was calculated. The strength of interaction between scores for each analyte and mortality rate was assessed by chi-square analysis where a ppropriate. Relationships between the organ system score, other physiologic al variables and likelihood of euthanasia were analysed. Results Scores derived mathematically from analysis of enzyme activities ha d poor abilities to predict mortality. The score based upon the number of o rgan systems compromised showed good ability to predict mortality and the i nteraction between the organ system score and mortality rate was significan t by chi-square analysis (P < 0.01). Distribution of data within the amylas e and lipase scores was not compatible with chi-square analysis. Conclusion Assessment of severity of spontaneous canine acute pancreatitis using pancreatic enzyme activities is potentially inaccurate. The use of a severity score based upon organ system compromise was more accurate in dete rmining the likelihood of mortality in spontaneous canine acute pancreatiti s. This is compatible with the hypothesis that severe canine acute pancreat itis is a multiple organ failure syndrome.