P1393a System for Forecasting of Drug Dosage for Bone Mineral Disorders Correction in Patients on Chronic Dialysis

Abstract

Background and Aims Bone and mineral disorders (BMD) is a common complication of CKD in patients on chronic dialysis. Timely and adequate correction of BMD is the most important aspect of patient's treatment. This work presents a system for forecasting of phosphate-binding agents (PBA) dosage and vitamin D receptor activators (VDRA) dosage. The system consists of sequentially triggering artificial neural network forecasting models (separate model for each drug type). Method As an input dataset, system uses patient’s results of laboratory studies (blood calcium, phosphate and PTH) for the period of 6 months, information on previous drug therapy and data on adequacy of patient’s dialysis therapy. The output of the system are dosages of PBA an VDRA that have to be administered in order to bring the patient’s parameters as close as possible to target range of values (2.1-2.5 mmol/l for calcium, 0.87-1.5 mmol/l for phosphate and 150-300 pg/ml for PTH). The system consists of two sequentially triggering forecasting models (for PBA and for VDRA), where each model is an artificial neural network, that has been trained on a data, collected in more than 20 “Nefrosovet” hemodialysis clinics for the period of 3 years. The effect of system usage was examined for the group of 356 hemodialysis patients with median follow-up time of 3 month. The primary end-points were a number of patients in target range of values With respect to calcium (2.1-2.5 mmol/l), phosphate (0.87-1.5 mmol/l) and PTH (150-300 pg/ml). Results During the study we determined that as a result of using the dose forecasting system, number of patients in target range of values significantly increased with respect to calcium (from 178 to 209, p=.0196), phosphate (from 99 to 152, p=.0000), and PTH (from 83 to 109, p=.0281). Conclusion Employment of automated drug dosage forecasting system based on artificial neural network models, has a positive effect on BMD correction quality, which, in turn, reduces the risk of possible complications.