Health Affairs Paper Describes Three Innovations in Health Care Delivery to Improve the Value of Health Care

This posting was authored by David A. Dorr, M.D., M.S., vice chair and professor of medical informatics and clinical epidemiology and professor of medicine.

Recently, there was a patient on my primary care schedule in his early 50s with schizophrenia, a history of substance use, severe diabetes, and chronic hepatitis C infection.  He had recently resurfaced in the clinic after requesting refills of his medications for some time without an in-person visit, and came because he was only told to. He was reported as disheveled yet cogent, knowing all his medications, but didn’t have a stable place to live or the ability to afford his medications.  At my visit, he didn’t come and hasn’t responded to calls. People like him, with high needs across the behavioral, social, and physical domains, represent a fundamental challenge for our systems – his likelihood to have many hospitalizations and poor health outcomes is high while the quality of care we are able to give him in standard, visit-based care is low.  In some way, he represents a driver for us to try to change our fundamental models of care – by knowing who he is and other people like him, by acting proactively to find what will work for him, and to help him reduce his risks. To do so requires significant amounts of data, information, and knowledge.

Our paper in the February 2018 issue of Health Affairs describes three innovations in health care delivery intended to improve the value of health care – its benefit over its cost, especially for those most at risk, like my patient. These innovations vary substantially – from the banding together of large groups in Accountable Care Organizations (ACOs) to primary care transformation in small and rural clinics – but share some similar challenges. These challenges to the diffusion of innovations may be linked to an inability to use and apply data, information, and knowledge to change perceptions of current practice and motivate change.

Three authors – myself (David A. Dorr, M.D., M.S.), Deborah J. Cohen, Ph.D., and Julia Adler-Milstein, Ph.D. – gathered data about three different innovation models to understand the similarities and differences in the use of data and in health information technology (HIT) systems to drive these innovations in response to a Health Affairs special issue on Diffusion of Innovation.

Julia Adler-MilsteinDr. Adler-Milstein is an associate professor of medicine and director of the Clinical Informatics and Improvement Research Center, School of Medicine, University of California, San Francisco. She reported results of a national survey funded by the Commonwealth Fund that assessed ACOs adoption and intensified use of health IT and performance reporting functions, finding significant adoption of several types of HIT, including Electronic Health Records (EHRs), registries about patients with certain diseases, health information exchange, and the use of performance measurement at the physician level and for higher level dashboards. Over half of respondents reported health information exchange and enhancing EHRs for population management as very challenging.

David DorrI serve as vice chair and professor in medical informatics and clinical epidemiology at Oregon Health & Science University (OHSU), and reported on practices engaged in Advanced Primary Care programs, focusing on the nearly 500 primary care practices and over 2,800 practices engaged in the Comprehensive Primary Care (CPC) and subsequent CPC+ program.  Through funding from the Commonwealth Fund, we surveyed and interviewed practices; they reported significant investments in the use of HIT as they engaged in these programs. One primary example is the use of HIT to manage a process called risk stratification, where the entire group of patients seen within the practice were assigned a risk score for future adverse outcomes such as hospitalizations or ED visits. These risk scores could be assigned by clinical intuition, through an HIT algorithm, or both, and nearly all practices reported challenges in using HIT to create, adapt, update, and display these risk scores. The challenges in these efforts led to distrust and frustration with scores, and HIT changes were some of the top reported challenges in adopting the innovations proposed by the models. However, technology also provided a number of answers, including bringing diverse groups with similar problems together through on-line tools, enabling peer problem-solving.

Deborah CohenDr. Cohen, a professor in family medicine with a joint appointment in medical informatics and clinical epidemiology at OHSU, focused on EvidenceNOW, an effort where seven cooperatives across the country, funded by the Agency for Healthcare Research and Quality, helped 1,493 small-medium size primary care practices focus on improved quality of care for heart health in their patients.  Cooperatives helped practices extract data, use quality improvement techniques to increase adherence to the guidelines, and monitor the changes over time. Practices and the cooperatives reported significant barriers, including the variety of locations for data storage, lack of access to the data, and inconsistent implementation of measures across practices.

Overall, significant changes to HIT were made to adopt the innovations, yet were felt to be insufficient. Across all the efforts, engaged participants reported significant fatigue at the challenges related to HIT and the number of expected changes. This work was done in the setting of extensive policy changes to encourage the use of HIT, especially EHRs, and participants reported the number of requirements unrelated to their specific goals felt burdensome. Policy changes to tie the HIT adoption and support to more carefully drive the programs may improve response and success in using data to drive diffusions of innovations.

Funding for the research reported here comes from the Commonwealth Fund (Adler-Milstein, Dorr) and the Agency for Healthcare Research & Quality (Cohen).

Bookmark and Share

Add Comment

About the Author

About DMICE

The Department of Medical Informatics & Clinical Epidemiology (DMICE) is one of 27 academic departments in the School of Medicine at Oregon Health & Science University (OHSU). The mission of DMICE is to provide leadership, discovery and dissemination of knowledge in clinical informatics, clinical epidemiology, and bioinformatics / computational biology. This mission is fulfilled through programs of research, education, and service. For more information, visit http://www.ohsu.edu/informatics

Participation Guidelines

Remember: information you share here is public; it isn't medical advice. Need advice or treatment? Contact your healthcare provider directly. Read our Terms of Use and this disclaimer for details.