NIH – Leveraging Electronic Health Records for Alcohol Services Research (R21/R33)

December 16, 2016 by School of Medicine Webmaster

Electronic health records are increasingly being utilized to support major research initiatives, as they bring opportunities to access (with appropriate permissions) the diagnostic, treatment, and outcome data of vast numbers of patients across diseases, clinical care settings, geographic areas, and time.  For example, the ability to access and analyze patient medical information via electronic health records is at the heart of the President’s Precision Medicine Initiative; likewise, access to data repositories and virtual data warehouses, powered by electronic health records, makes possible such initiatives as the NIH Health Care Systems Research Collaboratory, the PCORI-funded Patient Centered Outcomes Research Network (PCORnet), practice-based research networks, and various other national and health system disease registries, data libraries, and collaboratives.

Alcohol and other substance use disorder treatment lags far behind the rest of health care in both the deployment of electronic health records, and in leveraging existing records to realize efficiencies in clinical and health services research. Independent specialty treatment programs have historically lacked even basic information technology, while primary and general medical settings either have not reliably collected, or have not integrated, data on patients’ use of alcohol and illicit drugs or engagement in substance use disorder treatment. This segregation of records is in part due to the requirements of 42 CFR (confidentiality of alcohol and drug information) and stringent interpretations of HIPAA, despite the latter’s provisions for the use of substance use and other health records for research. In short, at present, the ability to access and utilize electronic health records for substance use disorder research, and the quality, consistency, and completeness of available data, remain in something of a proof-of-concept stage.

There has been recent and ongoing attention paid to building or improving EHR infrastructures throughout health care, which should yield new or expanded opportunities for research on addiction health services topics – including understanding patterns of service access, utilization, effectiveness, cost, and quality; implementation of evidence-based clinical interventions, programs, and quality improvement initiatives; and the identification and treatment of substance use disorders and co-occurring conditions in real-world clinical care settings. As requirements for meaningful use of EHR data have increased system capabilities, a number of health plans and health systems, large academic medical centers, and some established practice-based research networks are well-positioned to leverage their own data systems to conduct addiction health services research, but these activities have been limited. This FOA is intended to provide a mechanism to stimulate such research.

Specific Areas of Research Interest

NIAAA encourages investigators to conduct health services research related to the treatment of alcohol use disorders, for which existing electronic resources (electronic health records, patient registries, virtual data warehouses, medication orders, etc.) would serve as the principal source of data. To the extent possible, studies should utilize existing clinical infrastructure and data systems to conduct the research, with a focus on real-world usual care conditions. Pragmatic trials, if proposed, should utilize broad eligibility criteria and flexibility in the implementation of the intervention, with minimal disruption of current clinical practice. Together, these types of studies will ideally yield evidence that can be directly applied to improve clinical decision making, patient care, and health outcomes.

Although multisite studies are not required, research drawing on health systems, practice-based research networks, and existing health care collaboratives would provide access to larger data systems and thus maximize the generalizability of the findings.

Specific topics and study designs are open. The following examples are illustrative and not exhaustive:

Prospective studies to identify service utilization patterns and compare AUD or other health outcomes among different demographic, SES, and clinical sub-groups.

Studies that seek to replicate promising findings from smaller pilot trials in larger, more diverse, and real-world clinical care settings.

Studies that integrate Census or other geographic data to measure associations between crime rates and other indicators of community stressors on broad patterns of alcohol service utilization and outcome.

Studies using predictive modeling to anticipate and respond to treatment failure (i.e., better targeting interventions to patient status).

Studies that compare service utilization, cost, and outcomes for AUD to other chronic conditions.

Natural experiments at the healthcare center, system, or policy level (e.g., to assess the success or impact of program implementation, changes in formularies or payment structures, introduction of new clinical practice guidelines, etc., on service utilization and/or outcomes).

Retrospective or prospective cohort designs using linked data to assess the effectiveness of alternative “referral to treatment” procedures for patients with moderate to severe AUD.

Observational implementation trials of the adoption of evidence-based treatments or service delivery models.

Randomized registry trials.

Adaptive designs to match patients with the most appropriate AUD treatments.

Note: Applicants seeking to conduct traditional secondary analyses of administrative data should utilize PAR-16-234.

Deadline:  standard dates apply


Filed Under: Funding Opportunities