Information
The Need For RECORD?
The rapid entry of health care into the electronic age has led to a proliferation of large data repositories of health-related information. Such data are collected for administrative and clinical purposes, without specific a priori research questions. Examples of these “routinely collected health data” include health administrative data, electronic medical records, primary care surveillance data and disease registries. These data are increasingly being used for observational, comparative effectiveness and health services research. This expanding research area and the growing numbers of publications have generated increasing interest in specific methodological issues regarding the use of such data for research purposes.
At present, there is a lack of clarity regarding the strengths, limitations and biases of available electronic data sources. This confusion has been compounded by incomplete or inadequate reporting of research based on routinely collected health data in the literature.
The establishment of reporting guidelines is associated with improved quality of research reporting (1). An international collaboration recently produced the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement to improve reporting of observational research (2, 3). The STROBE statement has been endorsed and is being used by a wide range of journal editors, researchers and consumers of research (4). Since most research using routinely collected health data is observational in design, the STROBE guidelines should apply to these studies. However, due to the general nature of STROBE, specific issues related to reporting research using routinely collected data are not addressed. No reporting guidelines specifically addressing reporting of this type of research were identified in the Library for Health Research Reporting of the Enhancing the QUAlity and Transparency Of health Research (EQUATOR) Network (5). Three checklists have been published to aid in the conduct and critical-appraisal of studies using routinely collected data for comparative-effectiveness research (CER), one of which specifically framed retrospective database research in the context of the STROBE guidelines (6-8). However, the following issues limit the usefulness of these checklists: a) they focus primarily on the methods of such research and less on reporting, b) they are specific to CER and do not apply to most observational research, c) they have not been widely endorsed by journal editors and d) they were not developed with international stakeholder consensus or as part of a formal extension of STROBE.
Extensions of the STROBE guidelines have been produced for genetic association studies (9) and molecular epidemiology research (10). Formal reporting guidelines for studies using routinely collected health data, with multi-disciplinary expert stakeholder involvement and collaboration from STROBE methodologists, will ensure the transparency of the methods used in this growing field of research.
To gauge the interest and feasibility of producing such a guideline amongst the users of routinely collected health data, we conducted a workshop following the Infectious Disease Research Network (IDRN) Primary Care Database Symposium 2012 (January 27, 2012 in London, UK). Over 100 participants (including 5 conveners of the STROBE initiative) participated in the workshop. Strong interest was expressed in the possibility of developing a reporting guideline specific to research using routinely collected health data.
Participants at the workshop identified unique issues in database research which are frequently unreported in the literature, including (but not limited to): description of database characteristics, validation of disease identification algorithms, validation of codes to identify both exposures and outcomes and methodology of linkage of patient records between databases. As a result of the lack of specific guidance, both participants of the workshop and members of the STROBE steering committee have agreed that there is a need for development of the RECORD guideline and both groups will be involved in the formal process. Finally, the Journal of Clinical Epidemiology and Clinical Epidemiology published editorials from the RECORD Steering Committee to address the need for such a guideline (11,12) (see Publications Tab), indicating the interest of journal editors in complete and transparent reporting of this field of research.
We have therefore decided to proceed with a formal guideline development process, in close conjunction with members of the STROBE group, to ensure consistent methods and to make this a useful extension to the original STROBE statement.