Choosing outcome variables: global assessment and diaries.

This report reviews the advantages and disadvantages of 2 types of outcome assessment regularly used in the literature on incontinence: (1) global assessments of symptoms or relief and (2) ratings of specific symptoms of incontinence. Global assessments are usually used when the target condition involves multiple and variable symptoms. They are easy for patients to understand and yield data that seem to correspond closely to the target of the treatment (e.g., symptom severity). The 2 major scaling approaches used in clinical symptom assessment, categorical scaling and cross-modality matching of symptom intensity to line length (visual analogue scales), are reviewed for their reliability and validity. Numerical scales provide the advantage of a large number of possible ratings (usually 9-100) and are intuitively easy to use. All methods of symptom assessment using patient recall are subject to potential reporting bias. Measures of current symptoms, although less biased, may not reflect ongoing symptom levels, and current symptom levels tend to influence patients' recall of both symptom severity and medication counts. However, patient-initiated ratings are relevant to assessing "real-time" symptoms in incontinence trials, and even single retrospective ratings of "usual" symptom intensity may actually be an adequate reflection of the average symptom severity (at least for chronic pain). Applying subjective, predetermined criteria for success, such as the provision of "adequate relief," allows researchers to closely mimic clinical situations and use a cutoff point ("threshold") specifically associated with the condition and treatment being investigated. Devices such as personal digital assistants that can cue subjects for information at set or random intervals, record multiple types of responses, and accurately assess compliance with data recording should help facilitate future research.