Regla de predicción para determinar infestación canina por Echinococcus granulosus

Cystic hydatidosis is an endemic disease in Andean regions, our objective was to determine a score that helps to predict canine Echinococcus infestation in an endemic area with hydatidosis. A longitudinal study was carried out in 3 annexes of the province of Concepción located in the central Andes of Peru. Canine echinococcosis infection was defined by the presence of E. granulosus antigen identified in the stool sample. The predictive model was obtained through a logistic regression analysis based on the sociodemographic parameters, this rule was internally validated by bootstrap type resampling. Results: Of 152 dogs that underwent the stool study, 76 had an Echinococcus infection confirmed by reactive antigen. Four weighted factors were identified within the prediction rule that, in sum, gave the score: The dog that sleeps outside the home (3 points), feeding raw viscera to the dog (3 points), slaughtering the cattle inside the house (2 points) and the dog is fed with croquettes (-2 points). This prediction rule was found to have a value of the area under the ROC curve 0.78 (95% CI: 0.70 - 0.86). At a cutoff point of ≥ 3 points, the prediction rule was found to have a sensitivity of 75% and a specificity of 65.8%, with a positive predictive value of 85.3%. This rule will help health personnel to identify canes infested by Echinococcus granulosus, for their anti-parasitic and preventive intervention in their owners.