Determinants and outcomes of patient access to medical records: Systematic review of systematic reviews

Background: Patient access to electronic health records (EHRs) is associated with several determinants and outcomes, which are interrelated. However, individual studies and the reviews summarizing them have only addressed particular aspects, such as policy, usability or health outcomes of adoption. Therefore, no comprehensive overview exists. Additionally, reviews used different theoretical frameworks, which makes results difficult to compare. Objective: We aimed to systematically review recent systematic reviews on determinants and outcomes of patient access to EHRs to create a comprehensive overview and inform policy-makers and EHR implementers about the available literature, and to identify knowledge gaps in the literature reviews. Methods: We searched MEDLINE, EMBASE and PsycINFO for systematic reviews on patient portals, personal health records, and patient access to records that addressed determinants and outcomes of adoption. We synthesized the results from these reviews into the Clinical Adoption Framework (CAF), by mapping quotes from the reviews to categories and dimensions of the CAF, starting with the most recent ones until saturation of the CAF had been reached. The risk of bias in the reviews was assessed using the AMSTAR2 checklist. Results: We included nineteen reviews from 8871 records that were retrieved until February 19th, 2018. The reviews had a median of 4 (IQR: 4-4) critical flaws according to the AMSTAR2 checklist. The reviews contained a total of 1054 quotes that were mapped to the CAF. All reviews reported on the dimension ‘People’ that can affect adoption (e.g. personal characteristics such as age) and the dimension ‘HIS use’ (health information system use). Most reviews reported the dimensions ‘Organisation’, ‘Implementation’, HIS ‘System quality’, and outcomes of HIS use. However, gaps in knowledge might exist on macro-level determinants and outcomes, such as healthcare standards, funding, and incentives, because few reviews addressed these aspects. Conclusions: No review covered all aspects of the CAF and there was a large variety in aspects that were addressed, but all dimensions of the CAF were addressed by at least two reviews. Although reviews had critical flaws according to the AMSTAR2 checklist, almost half of the reviews did use methods to assess bias in primary studies. Implementers can use the synthesized results from this study as a reference for implementation and development when taking quality restrictions into account. Researchers should address the risk of bias in primary studies in future reviews and use a framework such as CAF to make results more comparable and reusable