Determining the Effectiveness of Torasemide and Furosemide in Heart Failure

Abstract

A state-of-the-art clinical computing system was investigated to test the hypothesis of whether a completely and predictably absorbed loop diuretic (torasemide) offers clinical advantages over a diuretic that is poorly and erratically absorbed (furosemide). We enrolled 234 patients with heart failure into this randomised, prospective, open-label study of the effectiveness and cost effectiveness of torasemide and furosemide with a comprehensive electronic medical record called the Regenstrief Medical Record System (RMRS) serving as the primary data repository. The RMRS contained patient data for physician visits, emergency department visits, hospitalisations, laboratory tests, prescription medications, and in- and outpatient service charges. Because these patient data were captured during the course of routine care, there was much efficiency in study data collection and few associated costs. Before the study began, ready access to the RMRS data permitted characterisation of the target study population and an accurate sample size calculation. During the study, programs contained within the RMRS identified patients for recruitment, preserved randomisation, and alerted investigators to important patient events. Quality-of-life instruments programmed into notebook computers were loaded with available RMRS data and then supplemented with primary collection data, thereby permitting paperless data entry. The computing infrastructure used for this study has permitted us to examine an interface between clinical pharmacology and outcomes research within the realm of medical informatics.