The evaluation of changes in functional health status in patients with abdominal complaints

Measuring functional status changes in various patient subgroups is important in stratifying risk, assessing disease severity, and predicting and defining clinically relevant outcomes. Data from a multi-centered study of 980 primary care patients presenting with nonspecific abdominal complaints were studied to demonstrate the importance of such an assessment procedure. Patients were prospectively followed for 6 months. Five diagnostic categories based on illness duration and seriousness were derived from the clinical course of these patients. The functional status of each patient was determined at baseline, 1 month, and 6 months using the Sickness Impact Profile (SIP). Intraclass correlation coefficients accounted for two aspects of the reliability of the SIP regarding the measurement of change over time: differences between patients which are stable over time (reproducibility) and different effects of treatment between subsets (responsiveness). A priori formulated expectations about the degree of health status change in patient subgroups were evaluated with the help of effect-size calculations. Patient impairment only partially depended on the final diagnosis and was also influenced by the presence of co-morbidity, psycho-social determinants, and other complaints. The health status change in the patient subgroups agreed with a priori formulated expectations. Standardized effect-size calculations revealed that the degree of change over time in SIP scores was in accordance with these expectations. We conclude: (a) the SIP appeared to be a reliable clinimetric instrument in detecting change over time resulting from different clinical courses, (b) clinical studies that use clinimetric instruments to assess the effects of clinical interventions must adequately control for the influence of baseline "functional status" as well as traditional demographic features such as gender and age, and (c) evaluating a priori formulated clinical expectations concerning functional change with statistics such as intraclass correlation and effect sizes can lead to a clearer understanding of the clinical relevance of statistically significant changes.