Background: Patient-Reported Outcomes (PROs) data are increasingly common and widely accepted in clinical investigations of symptoms.Agreements or differences regarding Doctor-Report Outcomes (DROs) and PROs data are unclear.Methods: In this study, we conducted a clinical trial and investigated the levels of agreement on DROs and PROs in symptoms of diabetics.This study had a parallel self-controlled and double blind design.In total, 90 diabetics who presented with the required symptoms were enrolled in this study, and 83 patients completed the study.The severity (none, mild, moderate, and severe) of each symptom was consistently compared with the Visual Analogue Scale (VAS) for the PROs measure and the numerical rating scales (NRS) for the DROs measure.The Kappa index was used to test the agreement in the variables.Results: Our results show that most of the agreement strength regarding DROs and PROs was moderate;some agreement strength results were modest, and some were substantial.However, the DROs data failed to agree with the PROs data in the symptoms-related effective evaluation criteria overlapping the inclusion criteria, which is caused by distortions of information regarding the doctor s decision process.Conclusions: We suggested that effective evaluation criteria should not be overlapped with the inclusion criteria in the DROs design research or otherwise be evaluated on the basis of the PROs data.