IntroductionIn order to facilitate patient information, patient involvement, and to support patient-centered care, healthcare organizations are increasingly offering access to patient data that are stored in the institution-specific electronic health record (EHR). Patients can access these data, read, and print them, or download and integrate them into any type of patient-held record. This EHR access is typically web-based and called “patient portal” allowing the independent access via the Internet from everywhere. A patient portal may also offer additional features such as prescription requests, appointment booking, messaging, personal health-related reminders, individual therapeutic recommendations, personal diaries, and social networking with other patients. In a Cochrane review, we assessed the effects of providing access to EHR for adult patients on patient empowerment and health-related outcomes compared to usual care.MethodsAccording to the methods of evidence-based medicine, we developed a protocol for a Cochrane review, which is published in the Cochrane database.ResultsWe identified ten randomized controlled trials (RCTs) including 6,668 randomized participants. Seven RCTs took place in the USA, two in Canada, and one in Japan. Additional functionalities of interventions and disease conditions were heterogeneous. Three studies (n = 601) reported on patient empowerment. The risk differences reported were neither statistically significant nor clinically relevant. Eight studies (n = 2,070) reported on nine different risk factors (blood pressure, blood glucose, poor asthma control, 10-year Framingham risk score, cholesterol, body mass index, composite score of eight variables, intraocular pressure, composite score of three variables). The results were heterogeneous. Mostly there were no statistically significant risk differences between study groups.ConclusionsOverall, there is no evidence for a clear positive effect of patient portals on patient empowerment and health related outcomes (mainly risk factors). However, we identified only a small number of studies. The usage of portals was often low and several studies were older.