Choice of an outcome measure for a pilot randomised controlled trial into the treatment by herbal practitioners of symptoms associated with the menopause

Objective

Planning of clinical trials in CAM requires detailed consideration of methodology to reduce potential biases. Validated outcome measures are useful to demonstrate changes in health or well-being but choice of a suitable measure can be difficult. The choice of a measure is discussed.

Materials and methods

Menopause scales were reviewed in February 1999 and August 2001 and published reports of trials were obtained. Criteria used in the choice were: self-completion (to reduce practitioner bias), length and ease of completion (to improve compliance in treatment and control groups), construct validity, reliability and responsiveness to change.

Results

Measures compared table form: Greene Climacteric Scale, revised Kupperman Index, MENQOL, Menopause Rating Scale, Menopause Symptom List, Women’s Health Questionnaire.

The scores in all measures combine a range of physical and subjective symptoms. Quality of questionnaire design varied particularly in consideration of responsiveness to change and there were few published trials. The Greene Scale was chosen because construct validity has been shown and it proved simple to complete when piloted in practice.

Conclusion

Design of a clinical trial is an iterative process; the final protocol depends on the setting but much work on the methodology has to happen beforehand. Our experience is that it is important to review possible outcome measures early in the process to ensure that an appropriate validated measure is used.