Menstrual diary data and menopausal transition: methodologic issues.

OBJECTIVES

To identify problems arising in the analysis of menstrual diary data in a longitudinal study of the menopausal transition, and to propose solutions.

SUBJECTS AND METHODS

Menstrual diary information provided by 372 women over a period 7 years is the basis of the study. This paper considers the processes and decisions associated with transforming this information into a form suitable for answering research questions.

RESULTS

We decided to use 'cycle' as the unit of analysis and to regard spotting-only episodes as nonmenstrual bleeding. Difficulties in distinguishing between long cycles and gaps because of recording lapses are shown to be related to both the design of the menstrual diary cards and the procedures used by fieldworkers. A new design that should minimize such difficulties is presented. Procedural changes are also proposed.

CONCLUSIONS

It is important to minimize uncertainty about the meaning of gaps in menstrual diary data, especially in studies of the menopausal transition. Methodologic improvements that can help to achieve this aim include design of diary cards and fieldwork procedures to reduce uncertainty. The design of data entry interfaces to facilitate the checking of raw data and the use of graphs as a diagnostic tool are other important steps.