To assess the sensitivity to within person change over time of an outcome measure for practitioners in primary care that is applicable to a wide range of illness. Comparison of a new patient generated instrument, the measure yourself medical outcome profile (MYMOP), with the SF-36 health profile and a five point change score; all scales were completed during the consultation with' practitioners and repeated after four weeks. 103 patients were followed up for 16 weeks and their results charted; seven practitioners were interviewed. Established practice of the four NHS general practitioners and four of the private complementary practitioners working in one medical centre. Systematic sample of 218 patients from general practice and all 47 patients of complementary practitioners; patients had had symptoms for more than seven days. Standardised response mean and index of responsiveness; view of practitioners. The index of responsiveness, relating to the minimal clinically important difference, was high for MYMOP: 1.4 for the first symptom, 1.33 for activity, and 0.85 for the profile compared with < 0.45 for SF-36. MYMOP's validity was supported by significant correlation between the change score and the change in the MYMOP score and the ability of this instrument to detect more improvement in acute than in chronic conditions. Practitioners found that MYMOP was practical and applicable to all patients with symptoms and that its use increased their awareness of patients' priorities. MYMOP shows promise as an outcome measure for primary care and for complementary treatment. It is more sensitive to change than the SF-36 and has the added bonus of improving patient-practitioner communication.
