OBJECTIVES: The Swedish National Diabetes Register (NDR) has since 1996 longitudinally recorded medical outcomes e.g. risk factors, comorbidities, and covers the majority (80%) of Swedish diabetes patients. In addition patient reported values were collected alongside the registry in a questionnaire. Our purpose was to evaluate a method for measuring patient reported abilities and ratings of quality in healthcare and evaluating them together with outcomes extracted from the NDR. METHODS: A questionnaire was developed to measure if diabetes care is percieved as patient focused and efficient, through questions on self management ability, worries, ability to carry out daily acitivites, and perception of service, access and involvement. The questionnaire was issued to 4,760 patients, 2,916 responded. Registry data on risk factors (HbA1c, blood pressure, cholesterol) were extracted for each patient and connected to the questionnaire. Item Response Theory (IRT) was used to estimate patient abilities and patient ratings of quality in healthcare (IRT scores) from the response patterns. For each patient, registry data and IRT scores were used to derive an overall Malmquist approach output quantity index, a healthcare related component and a patient ability component. The index is a measurement of how efficiently the patient leads his or her life with diabetes and its care, and provides a measure of the patient's state of health in relation to the patient's situation. RESULTS: We obtained IRT scale models with good fit, satisfactory validated in another population. The IRT scores provide basis for patient evaluation in a broader perspective than risk factors alone. The ability index component varies more than the healthcare component. CONCLUSIONS: The questionnaire provides estimates of abilities and ratings of quality. Our approach allows estimating patient benefit and health care production using combined registry and patient reported data, the procedure probably easier for patients than methods like time trade-off.