The Role of Follow-up in Randomized Controlled Trials

Abstract

After completing this article, readers should be able to: 1. Describe useful tools for interpreting the randomized controlled trial literature. 2. Characterize the difference between a surrogate outcome and a clinically meaningful outcome. 3. Characterize the trend in the prevalence of neurodevelopmental impairment in very low-birthweight infants. 4. Describe the adverse effects that have been associated with postnatal steroid use and why they were first reported in 1998. 5. Justify neurodevelopment or death as the outcome of randomized controlled trials of neonatal interventions. 6. Describe the major methodologic problems in conducting such trials. Dramatic increases in basic and translational research have increased the number of therapies available to clinicians. Because randomized clinical trials are the gold standard for evaluating the effect of treatment, there have been parallel increases in the number of randomized controlled trials (RCTs) and the refinement of methods for their design, conduct, analysis, and reporting. For investigators designing a clinical trial, clinicians seeking an answer to a specific clinical problem, and clinicians seeking to remain abreast of the trial literature, efficient access to and use of the medical literature is a major challenge. A number of tools have been developed for evaluating the effectiveness of new therapies, including systems to rate the quality of evidence, such as The United States Preventive Task Force’s rating system (Table 1⇓ ). (1) View this table: Table 1. US Preventative Task Force’s Rating System Another major advance is the development of standardized methods for formal, systematic reviews of trials of an intervention (meta-analysis). The Cochrane Collaboration, an organization committed to the preparation and dissemination of systematic reviews of health care interventions, uses such methods to identify high-quality trials and to obtain a summary estimate of the overall treatment effect of a specific intervention in a defined population on each outcome of interest. In addition to providing a typical estimate of …