Journal of Clinical Epidemiology 56 (2003) 310–316 Subject’s Global Assessment of Relief: An appropriate method to assess the impact of treatment on irritable bowel syndrome-related symptoms in clinical trials Stefan Mu ¨ller-Lissner a , Gary Koch b , Nicholas J. Talley c , Douglas Drossman d , Peter Rueegg e , Cornelia Dunger-Baldauf e , Martin Lefkowitz f, * a Park-Klinik Berlin–Weissensee, Scho ¨nstrasse 80, D-13086 Berlin, Germany b Department of Biostatistics, University of North Carolina, 3106-B McGavran-Greenberg Hall, 238-CB #7420, Chapel Hill, NC 27599, USA c Department of Medicine, University of Sydney, Nepean Hospital, P.O. Box 63, Penrith, New South Wales, Australia d UNC Center for Functional GI and Motility Disorders, Division of Digestive Diseases, University of North Carolina, BI 1110 Trailer 49, Mason Farm Road, Chapel Hill, NC 27599, USA e Novartis Pharma AG, Postfach, CH-4002 Basel, Switzerland f Novartis Pharmaceuticals Corp., P.O. Box 11, One Health Plaza, East Hanover, NJ 07936, USA Received 5 January 2001; received in revised form 16 July 2002; accepted 9 December 2002 Abstract The lack of validated outcome measures is a key limitation for the evaluation of drug efficacy in the treatment of irritable bowel syndrome (IBS). In clinical trials with tegaserod, the Subject’s Global Assessment (SGA) of Relief (a global measure that includes overall wellbeing, abdominal pain/discomfort, and bowel function) was used to identify responders. A total of 1680 patients with IBS with constipation were included in two clinical studies comparing tegaserod with placebo. The SGA of Relief was obtained weekly by a single, self-administered question with five possible answers. Responders for the SGA of Relief reported statistically significant (P  .001) and clinically relevant improvements in multiple IBS-related symptoms compared with nonresponders. Response was also associated with a significant improvement in quality of life. The SGA of Relief is reliable as a new outcome measure for assessing response to therapy in IBS patients and has demonstrated responsiveness and reproducibility.  2003 Elsevier Inc. All rights reserved. Keywords: IBS; Tegaserod; Zelmac; Constipation; Subject’s Global Assessment of Relief; Quality of life; Visual analogue scale 1. Introduction Irritable bowel syndrome (IBS) is one of the most common noninfective gastrointestinal (GI) disorders in the Western world [1,2]. Although IBS can be defined by clinical criteria (e.g., those of Manning [3], Kruis [4], or Rome [5,6]), these criteria were developed primarily as diagnostic measures to identify homogenous groups of patients for epidemiologic studies or for inclusion in clinical trials. The criteria were not designed to be primary or secondary outcome measures to judge the efficacy of a pharmacologic treatment on IBS symptoms. Patients with IBS report numerous symptoms, including abdominal pain or discomfort, bloating, abnormal stool fre- quency, and abnormal stool consistency. Symptoms may vary over time and in duration [7,8]. Although the intensity * Corresponding author. Tel.: 001-862-778-2378; fax: 001-862-778- 2390. E-mail address: marty.lefkowitz@pharma.novartis.com (M. Lefkowitz). 0895-4356/03/$ – see front matter  2003 Elsevier Inc. All rights reserved. doi: 10.1016/S0895-4356(03)00027-1 of each symptom can be measured by specific question- naires using ordinal or visual analogue scales (VAS) [9], it remains difficult to evaluate the patient’s own opinion of the relative importance of each symptom [10]. To evaluate the patient’s response in therapeutic studies in IBS, a simple question or questionnaire could be used as a primary outcome measure. The validation of a primary outcome measure requires, as published by the Rome II working group recommendation [9], that (1) the primary outcome measure includes symptoms relevant to, and repre- sentative of, the disease, (2) the measure is reproducible (i.e., produces a similar result when administered to patients whose health status has not changed), (3) the measure is able to detect change (response) assuming a change took place, and (4) a change in the outcome measure should reflect a real change in the general health status. In addition, a primary outcome measure for treatment trials in GI disorders of function should attempt to capture an overall effect (i.e., should be a global assessment of key symptoms). However, no questions or questionnaires are sufficiently validated to