Control groups in studies of contrast media adverse events

Autor: Judith A. W. Webb, Henrik S. Thomsen
Rok vydání: 2016
Předmět:
Zdroj: Acta radiologica (Stockholm, Sweden : 1987). 57(8)
ISSN: 1600-0455
Popis: For many years controlled trials have been used to evaluate both the effectiveness and the safety of drugs, with the patients who receive the drug under investigation being compared to a control group who receive either an inactive placebo or no treatment. The ideal study design is a prospective randomized controlled double-blinded trial, with the participants randomly allocated to either the group receiving the drug under investigation or the placebo control group, and double-blinding ensuring that neither the participants nor the researchers know which substance the patient received. More recently, studies have used control groups which receive an active comparator, a drug already approved for the treatment of the disease being studied, but this has some limitations. The purpose of this paper is to consider the use of control groups when contrast media are being evaluated for the risk of adverse effects. The problems of randomized controlled trials in contrast media research and some of the possible alternative methods of using control groups and attempting to obtain reliable information will be considered. Randomization of patients in contrast media studies is usually not possible. It would not be ethically acceptable in many clinical situations to randomize patients to receive either contrast medium or saline, since in many clinical situations the contrast medium is important for obtaining the most accurate diagnosis; for example, the use of contrast medium enhancement when evaluating a patient for metastases. Nor would it be acceptable in most circumstances to subject a patient to two examinations, one with and one without contrast medium. This would involve waiting for 7 days between the unenhanced and enhanced examination and would mean doubling the radiation exposure if iodine-based agents were being evaluated. Double blinding might also be difficult, since both the technical staff and the reporting radiologist can see from the images whether or not the patient received contrast medium. Prospective studies of acute general adverse reactions have nearly always been active comparator studies, involving a comparison of two iodine-based contrast media, rather than comparing the contrast medium to placebo. However, a recent prospective study of over 1300 patients having magnetic resonance imaging (MRI) or computed tomography (CT) by Azzouz et al. (1) highlighted a problem which arises when no placebo control group is used. They showed that patients undergoing unenhanced MRI or CT reported the same mild acute general adverse events as the patients who had enhanced MRI or CT, although the adverse events occurred less frequently in the unenhanced group. Thus mild adverse events experienced after contrast medium may not relate to the contrast medium, but may instead be caused by the patient’s underlying disease or other external factors. The use of patients having unenhanced scans as controls is a method for obtaining at least some of the information which a placebo controlled study would give. Because the majority of studies of acute general adverse reactions have not used control groups who did not receive contrast medium, contrast media may have been blamed for some mild reactions which they did not cause (2). If all minor adverse events after contrast medium are recorded, some patients may subsequently be denied enhanced scans, and the chance of the more accurate diagnosis which these may give. As with studies of acute general adverse reactions, most comparative studies of contrast medium induced nephropathy (CIN) published over the last 10 years have involved comparison of two iodine-based contrast agents, the non-ionic dimer and a non-ionic monomer, with the overwhelming majority of the studies being sponsored by the pharmaceutical industry. A prospective comparison between the various iodine-based monomers has not yet been undertaken, suggesting
Databáze: OpenAIRE
Externí odkaz: