Background

To support patient‐centred care, healthcare organisations increasingly offer patients access to data stored in the institutional electronic health record (EHR).

Objectives

Primary objective

1. To assess the effects of providing adult patients with access to electronic health records (EHRs) alone or with additional functionalities on a range of patient, patient‐provider, and health resource consumption outcomes, including patient knowledge and understanding, patient empowerment, patient adherence, patient satisfaction with care, adverse events, health‐related quality of life, health‐related outcomes, psychosocial health outcomes, health resource consumption, and patient‐provider communication.

Secondary objective

1. To assess whether effects of providing adult patients with EHR access alone versus EHR access with additional functionalities differ among patient groups according to age, educational level, or different status of disease (chronic or acute).

Search methods

We searched CENTRAL, MEDLINE, Embase, PsycINFO, CINAHL, and Scopus in June 2017 and in April 2020.

Selection criteria

Randomised controlled trials and cluster‐randomised trials of EHR access with or without additional functionalities for adults with any medical condition.

Data collection and analysis

We used standard Cochrane methodological procedures.

Main results

We included 10 studies with 78 to 4500 participants and follow‐up from 3 to 24 months. Nine studies assessed the effects of EHR with additional functionalities, each addressing a subset of outcomes sought by this review. Five studies focused on patients with diabetes mellitus, four on patients with specific diseases, and one on all patients. All studies compared EHR access alone or with additional functionalities plus usual care versus usual care only. No studies assessing the effects of EHR access alone versus EHR access with additional functionalities were identified. Interventions required a variety of data within the EHR, such as patient history, problem list, medication, allergies, and lab results. In addition to EHR access, eight studies allowed patients to share self‐documented data, seven offered individualised disease management functions, seven offered educational disease‐related information, six supported secure communication, and one offered preventive reminders.

Only two studies were at low or unclear risk of bias across domains. Meta‐analysis could not be performed, as participants, interventions, and outcomes were too heterogeneous, and most studies presented results based on different adjustment methods or variables. The quality of evidence was rated as low or very low across outcomes. Overall differences between intervention and control groups, if any, were small. The relevance of any small effects remains unclear for most outcomes because in most cases, trial authors did not define a minimal clinically important difference. Overall, results suggest that the effects of EHR access alone and with additional functionalities are mostly uncertain when compared with usual care.

Patient knowledge and understanding: very low‐quality evidence is available from one study, so we are uncertain about effects of the intervention on patient knowledge about diabetes and blood glucose testing.

Patient empowerment: low‐quality evidence from three studies suggests that the intervention may have little or no effect on patient empowerment measures.

Patient adherence: low‐quality evidence from two studies suggests that the intervention may slightly improve adherence to the process of monitoring risk factors and preventive services. Effects on medication adherence are conflicting in two studies; this may or may not improve to a clinically relevant degree.

Patient satisfaction with care: low‐quality evidence from three studies suggests that the intervention may have little or no effect on patient satisfaction, with conflicting results.

Adverse events: two small studies reported on mortality; one of these also reported on serious and other adverse events, but sample sizes were too small for small differences to be detected. Therefore, low‐quality evidence suggests that the intervention may have little to no effect on mortality and other adverse events.

Health‐related quality of life: only very low‐quality evidence from one study is available. We are uncertain whether the intervention improves disease‐specific quality of life of patients with asthma.

Health‐related outcomes: low‐quality evidence from eight studies suggests that the intervention may have little to no effect on asthma control, glycosylated haemoglobin (HbA1c) levels, blood pressure, low‐density lipoprotein or total cholesterol levels, body mass index or weight, or 10‐year Framingham risk scores. Low‐quality evidence from one study suggests that the composite scores of risk factors for diabetes mellitus may improve slightly with the intervention, but there is uncertainty about effects on ophthalmic medications or intraocular pressure.

Psychosocial health outcomes: no study investigated psychosocial health outcomes in a more than anecdotal way.

Health resource consumption: low‐quality evidence for adult patients in three studies suggests that there may be little to no effect of the intervention on different measures of healthcare use.

Patient‐provider communication: very low‐quality evidence is available from a single small study, and we are uncertain whether the intervention improves communication measures, such as the number of messages sent.

Authors’ conclusions

The effects of EHR access with additional functionalities in comparison with usual care for the most part are uncertain. Only adherence to the process of monitoring risk factors and providing preventive services as well as a composite score of risk factors for diabetes mellitus may improve slightly with EHR access with additional functionalities. Due to inconsistent terminology in this area, our search may have missed relevant studies.

As the overall quality of evidence is very low to low, future research is likely to change these results. Further trials should investigate the impact of EHR access in a broader range of countries and clinical settings, including more patients over a longer period of follow‐up, as this may increase the likelihood of detecting effects of the intervention, should these exist. More studies should focus on assessing outcomes such as patient empowerment and behavioural outcomes, rather than concentrating on health‐related outcomes alone. Future studies should distinguish between effects of EHR access only and effects of additional functionalities, and investigate the impact of mobile EHR tools. Future studies should include information on usage patterns, and consider the potential for widening health inequalities with implementation of EHR access. A taxonomy for EHR access and additional functionalities should be developed to promote consistency and comparability of outcome measures, and facilitate future reviews by better enabling cross‐study comparisons.