PURPOSE: To review pharmacological treatment of irritable bowel syndrome (IBS) in general. To determine whether treatment for IBS should be different in older patients.
BACKGROUND: IBS is a common clinical problem encountered by primary care physicians and gastroenterologists. Epidemiological studies indicate a high prevalence in the general population, 14% to 24% of women and 5% to 19% of men,1,2 with 2.4 to 3.5 million physician visits and 2.2 million medication prescriptions written annually in the United States alone.3,4 Although IBS is essentially a benign disorder, it is associated with considerable suffering.5,6 In a recent report,7 IBS patients experienced more-significant impairment in health-related quality of life than did those with gastroesophageal reflux disease and diabetes mellitus. In community-based studies, persons with IBS have greater disability, threefold higher absenteeism from work, and twofold higher average healthcare costs than do healthy controls.8,9
The prevalence of IBS in older people is lower10–12 and the incidence is much lower than in younger people. Traditionally, IBS is not diagnosed for the first time in patients presenting with the symptom complex after the age of 60 years.13 Nevertheless, when diagnosed in older people, IBS affects functional status and quality of life.14,15
Because symptomatology of IBS is complex and heterogeneous, its management is also complex. Current management of IBS1,16 focuses on relief of the predominant symptom. Management of IBS in older people requires basic knowledge of gastrointestinal physiology in older people; pharmacology, including drug-drug interactions; and pharmacological management of IBS in general. Before considering the management of IBS in older people, it will be also meaningful to evaluate pharmacological management of IBS reviewed in a well-designed meta-analysis.17 This meta-analysis is reviewed here.
DATA SOURCES: All published English-language studies were identified by a comprehensive electronic search of the following databases: MEDLINE (1966 to 1999), EMBASE (1980 to 1999), PsycINFO (1967 to 1999), and the Cochrane controlled trials registry.
STUDY SELECTION CRITERIA: Double-blind, randomized clinical trials in English had to satisfy the following six criteria for inclusion: (1) address treatment of IBS, (2) study adult patients, (3) administer a pharmacological intervention to more than 10 patients for at least 2 weeks; (4) include a placebo control group; (5) report an outcome measure of global status or individual symptoms (or both) of IBS; and (6) use a randomized, double-blind, parallel-group or crossover design. Three authors independently evaluated these inclusion criteria. Abstracts, studies not published in full, dissertations, and book chapters were excluded.
DATA EXTRACTION: From all included trials, qualitative data were extracted on criteria used for diagnosis of IBS, characteristics of the study sample, dose and length of the pharmacological intervention, methodological characteristics of the trial, reported outcomes, and statistical analysis.
The quality of each trial was determined by the following five criteria: (1) clear inclusion criteria, (2) baseline similarity of treatment and control groups, (3) similar compliance in treatment and control groups, (4) patient-rated outcome measures, (5) and rigorous analysis, preferably by the intention-to-treat principle. Any trial that satisfied at least four of the five criteria was judged as high quality. All three authors independently applied these quality criteria. Quantitative data on estimates of treatment effect and P-value for the reported outcomes were extracted when available. Each trial was classified simply as positive or negative in terms of efficacy.
A vote-counting approach was used and relied on two outcomes: global improvement and improvement in specific symptoms of IBS. Any trial that reported statistically significant improvement in global status or individual symptoms was classified as positive. The absolute difference in frequency of any global improvement between the intervention and control groups (along with the 95% confidence intervals) was calculated for each trial that provided sufficient data.18
MAIN RESULTS: Of 283 studies identified, 70 (25%) met the inclusion criteria. A single agent was evaluated in 66 trials, and a combination of two or more agents was examined in 4 trials. The most common medication classes were smooth-muscle relaxants (16 trials), bulking agents (13 trials), prokinetic agents (6 trials), psychotropic agents (7 trials), and loperamide (4 trials). The 70 trials spanned 1966 to 1999 and reported on 4,836 adult patients. The patients' mean age ranged from 24 to 51 (median 38), and 68% were women.