RESUMEN Introduccion : La prediccion de coledocolitiasis en pacientes con colelitiasis continua siendo debatible. El objetivo de este estudio es disenar un instrumento para la evaluacion preoperatoria de pacientes con colelitiasis en riesgo de coledocolitiasis y determinar su impacto en el uso de recursos diagnosticos. Materiales : Pacientes con colelitiasis- colecistitis aguda admitidos de urgencia. Analisis uni y multivariado de factores predictivos de coledocolitiasis. Determinacion de Odds Ratio (OR) con Intervalos de confianza 95% (IC95%) p7mm, ALT, bilirrubina total. Escala evaluacion 0-16 puntos con discriminacion de umbrales de riesgo bajo, medio y alto; area bajo curva ROC 0.88. Validacion prospectiva en cohorte 187 pacientes (area ROC: 0.87 p<0.001). Con el modelo se demuestra un impacto directo en uso de recursos segun diferentes umbrales. El modelo ofrece mejor discriminacion que otros identificados. Conclusion : La prediccion de coledocolitiasis debe involucrar una regla objetiva y la determinacion de umbrales de riesgo, especialmente entre el grupo de bajo y mediano riesgo. Este enfoque permite el uso racional y eficiente de recursos diagnosticos. Palabras clave : Riesgo, coledocolitiasis, umbrales, costos, prediccion. ABSTRACT Introduction: Prediction of choledocholithiasis continues been debatable. The aim of this study is to design an instrument for the evaluation of patients in risk of choledocholithiasis due cholelithiasis and to determinate the impact in the use of diagnostic resources. Materials : Patients with acute cholecystitis admitted in emergency setting. Single and multivariate analysis of predictive factors to determinate Odds Ratio (OR), with confidence Intervals of 95% (CI 95%), p7 mm, ALT and total bilirrubin. Development of a scale for the prediction of choledocholithiais, stratifying risk in low, intermediate and high groups. Area under ROC curve was 0.88. Prospective validation in a cohort of 187 patients (ROC area 0.87 p<0.001). Direct impact on cost savings. The model exhibit a better performance that other similar identified on medical databases. Conclusion: Prediction of choledocholithiasis should be involved an objective prediction rule and the evaluation of thresholds, mainly for patients in low and intermediate risk. This approach conduces to the rational and efficient use of diagnostic resources. Key words : Risk, prediction, choledocholithiasis, thresholds, cost.