Retrospective study of fever in dogs: laboratory testing, diagnoses and influence of prior treatment.

OBJECTIVES: To analyse the demographic information of dogs referred for investigation of fever, to determine the usefulness of various diagnostic investigations and to assess the effect of treatment before referral on the presence of fever at referral, the duration of the investigation and the ability to reach a final diagnosis.
METHODS: The clinical records of 66 dogs, in which fever was part of the clinical signs documented by the referring veterinary surgeon, were reviewed. The effects of treatment 24 hours before referral on temperature at initial consultation and on time to diagnosis were evaluated. The effect of body temperature at initial consultation on cost and on time to diagnosis was also determined. The effect of insurance on costs incurred was assessed. The utility of different diagnostic investigations was recorded, and cases were classified according to the final diagnosis.
RESULTS: Only 34.8 per cent of dogs were diagnosed with immune-mediated disease, with most frequent diagnoses being steroid-responsive meningitis and polyarthritis. Treatment 24 hours before referral significantly increased the time to diagnosis (P = 0.004) and affected the presence of fever at referral (P = 0.006). Insurance status did not significantly affect cost incurred by the owner. CLINICAL SIGNIFICANCE: This study documents a high incidence of immune-mediated disease in dogs referred for investigation of fever. It also documents a higher incidence of inflammatory central nervous system disease in febrile dogs than that reported previously. Of the diagnostic modalities employed in the majority of cases, radiography, cytology and bacteriological and fungal cultures (fluids/tissues) were the most useful. It is suggested that treatment is withdrawn or withheld before commencing diagnostic investigation of fever.