Using personal generated health data (PGHD) during medical consultations can be beneficial for both patients and clinicians. However, multiple acceptance barriers such as lack of PGHD reliability prevents a routine usage of this data. A clinical decision support system, called FullFlow, has been developed to address these acceptance barriers. The objective of this study was to determine if FullFlow was useful during consultations and to verify the hypothesis that the higher PGHD reliability, the more effective the system is. The assessment relied on a medical pilot during which clinicians and patients with diabetes used the FullFlow during medical consultations. The data collection relied on a post-consultation questionnaire in addition to system logs. This study showed that the PGHD reliability was low for an overwhelming majority of consultations. The information displayed was useful in half of the consultations according to the clinicians who answered the questionnaire. Despite this, the overwhelming majority of clinicians who answered the questionnaire found that the designed FullFlow system permitted to gain insights of the situation of the patients. The study showed the higher the PGHD reliability is, the more useful the system is for clinicians. PGHD usage in clinical settings can permit clinicians to gain valuable information regarding the situations of their patients. A clinical decision system can present useful information to clinicians. While the PGHD reliability is correlated to the usefulness of such system, it is not the only factor impacting it: context of the clinicians and patients such as novelty of usage and personal goals also plays a role in determining on how such system is useful for clinicians. However, due to a limited number of participants, a new medical pilot must be performed in order to confirm the results of this study.