Pregnancy diagnosis in the dog: a comparison between abdominal palpation and linear-array real-time echography

This paper presents quantitative results of canine pregnancy diagnoses performed by abdominal palpation and/or linear-array real-time echography. Each animal was first palpated by an experienced clinician and then another investigator performed ultrasound scanning using a 3 MHz transducer. In other dogs echography was undertaken only when the result of abdominal palpation was negative or doubtful or when the owner requested ultrasound scanning. Using all tests performed between days 20 and 49 of pregnancy (day 0 = day of first mating or insemination), the results of abdominal palpation (n = 116) versus real-time scanning (n = 135) were: sensitivity: 89.0% v 92.9%; specificity: 93% v 96.0%; positive predictive value: 95.6% v 97.5%; negative predictive value: 83.3% v 88.9%. Results of ultrasound tests performed in 97 other dogs were: sensitivity: 92.4%; specificity: 97.7%; positive predictive value: 98.0% and negative predictive value: 91.4%. Almost identical figures were found when only the tests performed between days 25 and 35 of pregnancy were included. It was concluded that real-time ultrasound scanning is an accurate method for discriminating between pregnant and non-pregnant bitches; sensitivity and specificity of the abdominal palpation technique executed by an experienced clinician approach that of ultrasound scanning; small litter size and lack of knowledge of conception date are the main sources of false negative diagnoses.