Objective To demonstrate similarities and differences between the various instruments used by clinical trialists for measurement of premenstrual syndrome to encourage development of consensus of measurement to aid future research. Design Computer and manual searches of all trials of drugs used for premenstrual syndrome supplemented by retrieval of all questionnaires and scales cited. Letters were sent to the main manufacturers to identify trials not in the public domain. Main outcome measures Number of scales in current use for classifying premenstrual syndrome; the symptoms included within them; their organisation of symptoms into clusters; the numerical values used in the scaling; the most frequently used items; and the extent of overlap in the criteria used. Results Three hundred and fifty clinical trials of 115 different drugs were identified. The most common symptoms or signs used in the scales were (frequency in square brackets) irritability [44], headache or migraine [40], depression [37], tearfulness [34], tension [31], anxiety [31], breast tenderness [31], mood swings or lability [30], back pain [30], weight gain [29], abdominal bloating [28], lack of concentration [28], food craving [28], and increased appetite [26]. Sixty-five different questionnaires or scales were used to classify premenstrual syndrome and for assessing entry eligibility and treatment outcomes; of these 65, 47 included sufficient details for a formal analysis and 18 scales grouped the symptoms or items into clusters or subgroups. Conclusion There is wide diversity in the scales used for assessing entry eligibility and treatment outcomes in premenstrual syndrome. This makes comparative evaluation of possible treatments extremely complex and difficult. It is recommended that authorities and organisations involved in conducting trials of treatments for premenstrual syndrome develop a consensus statement about the best instrument to be used. Regulatory authorities could perhaps provide guidance to which scales they would accept when clinical trials of drugs for premenstrual syndrome are submitted to them for assessment. This would ensure that industry-sponsored trials provide objective data. Consensus is required given the poor performance of most current treatments of premenstrual syndrome.
