Health Information System Safeguards

P E R S P E C T I V E

815

From AIDS to Opioids — How to Combat an Epidemic

n engl j med 375;9 nejm.org September 1, 2016

grams or vouchers (covering MAT medications and the overdose- reversal agent naloxone), perhaps as a new mechanism under the Substance Abuse Prevention and Treatment Block Grant program or Medicaid demonstration waiv- ers, could provide access for many people with OUD, even in states that haven’t expanded Medicaid under the ACA. Although the mental health parity law of 2008 requires most managed-Medicaid and private insurance plans that cover substance-abuse treatment to do so at the same level as other medical care, violations abound.5 Despite the requirement that sub- stance-abuse treatment be consid- ered an essential health benefit, and despite the fact that the Na- tional Institute on Drug Abuse deems MAT the first-line treatment for OUD, the Centers for Medi- care and Medicaid Services has not yet made methadone or bupre- norphine maintenance treatment for OUD a mandated benefit.

Finally, another innovation of the response to AIDS was the creation of the Office of AIDS

Research within the National In- stitutes of Health (NIH) to coor- dinate HIV–AIDS research efforts across institutes and programs. Such an office overseeing a na- tional strategy for addressing the opioid epidemic could be devel- oped and housed within the NIH or an appropriate division of the Department of Health and Hu- man Services; it could emphasize that OUD is a chronic medical disorder, as Surgeon General Vivek Murthy has insisted, that should be managed according to stan- dards analogous to those for other chronic disorders.

The scope of reform needed to respond appropriately to this epi- demic is daunting. The response to AIDS, however, established a precedent for expanding access to lifesaving medications and sup- porting clinicians in implement- ing evidence-based treatment in marginalized populations. Current federal and state efforts have largely fallen short in addressing the opioid epidemic, as witnessed by ever-increasing mortality. We believe that federal funding should

be used to promote new and ef- fective models that provide patients with evidence-based treatment rather than supporting outdated treatment programs that are un- willing or unable to evolve.

Disclosure forms provided by the authors are available at NEJM.org.

From the Division on Substance Abuse, Columbia University Department of Psychia- try, New York State Psychiatric Institute, New York.

1. Nosyk B, Anglin MD, Brissette S, et al. A call for evidence-based medical treatment of opioid dependence in the United States and Canada. Health Aff (Millwood) 2013; 32: 1462-9. 2. Buck JA. The looming expansion and transformation of public substance abuse treatment under the Affordable Care Act. Health Aff (Millwood) 2011; 30: 1402-10. 3. Sigmon SC. The untapped potential of office-based buprenorphine treatment. JAMA Psychiatry 2015; 72: 395-6. 4. Bentzley BS, Barth KS, Back SE, Book SW. Discontinuation of buprenorphine main- tenance therapy: perspectives and outcomes. J Subst Abuse Treat 2015; 52: 48-57. 5. Wen H, Cummings JR, Hockenberry JM, Gaydos LM, Druss BG. State parity laws and access to treatment for substance use disor- der in the United States: implications for federal parity legislation. JAMA Psychiatry 2013; 70: 1355-62.

DOI: 10.1056/NEJMp1604223 Copyright © 2016 Massachusetts Medical Society.

From AIDS to Opioids — How to Combat an EpidemicAccelerating Innovation in Health IT

Accelerating Innovation in Health IT Robert S. Rudin, Ph.D., David W. Bates, M.D., and Calum MacRae, M.B., Ch.B., Ph.D.

Even as information technology (IT) transforms many indus- tries, the pace of innovation in health IT continues to lag. Elec- tronic health records (EHRs) re- ceive few accolades from providers and have been cited as a major source of professional dissatisfac- tion among physicians.1 Despite a proliferation of patient-facing health apps, few have been shown to produce health improvements and many are barely used. The

most common IT tools connecting patients to providers are patient portals that so far do little more than provide basic secure mes- saging and present unexplained clinical data. Though many start- ups and research programs exist and venture capital investment has been growing, health IT suc- cess stories remain rare.

A plan to accelerate innova- tion should begin with a diag- nosis of the problem. Some ob-

servers blame perverse financial incentives in health care that re- ward volume rather than quality and efficiency, regulations that restrict the flow of information ostensibly to protect patient pri- vacy, and technical integration challenges. Another factor has been the multiple demands of “meaningful use,” which have de- layed innovation in many areas of health IT. Though these issues are important, we believe there is

P E R S P E C T I V E

816

Accelerating Innovation in Health IT

n engl j med 375;9 nejm.org September 1, 2016

a more fundamental barrier that has not yet received due attention: the disconnect between health IT developers and users. Alternative provider-payment models should create incentives for innovation by rewarding health care provid- ers who use novel IT tools to control cost and improve quality, but the effect of these models will be attenuated unless the developer–user disconnect is ad- dressed.

Health IT developers typically work in one of three settings — established IT companies, start- ups, or academic research depart- ments — where they have little to no contact with patients and clinicians and therefore often lack a deep understanding of users’ needs. Established IT firms, most notably EHR companies, have adhered poorly to user-centered design principles, despite federal certification requirements that they apply such principles.2 In startups, developers are typically young and healthy, with little firsthand knowledge of clinicians or the chronically ill patients who consume most health care ser- vices. Much of venture capital is therefore clustered in wellness companies making products such as fitness trackers that cannot help the patients most in need and thus will have little effect on health care costs. Some health care incubators are producing startups that target clinicians and chronically ill patients, but we believe that these organizations generally underestimate the effort needed to understand such com- plex and diverse users. Some aca- demics have focused on under- standing users’ needs, but efforts tend to be small and fragmented and to involve multiple years of development. Rarely do findings

make their way into the design of novel functionalities, for which relatively few funding sources are available.

Users of health IT systems also face challenges in addressing this disconnect. Although clinicians may know what aspects of their system they dislike and may have ideas about how they might work better (e.g., perhaps notes could be made into a wiki, so that vari- ous clinicians could add to or revise them), few are trained to specify their ideas in a way that can be turned into workable software or understand IT capa- bilities well enough to propose technically feasible approaches. Experienced clinicians may also have difficulty imagining how their workf lows may be altered, especially in ways that relegate some of their revenue-generating activities to others.

We have observed myriad un- fortunate results of this developer– user disconnect. Tools are built on the basis of fundamental mis- conceptions about the clinical utility of new data sources (e.g., episodic blood-pressure readings or accelerometry). Developers make incorrect design assump- tions about when and how clini- cians are available to respond to data produced by monitoring de- vices and when such contact is appropriate and clinically useful. Developers incorrectly assume that the same features can be used for drastically different purposes — for example, for both individ- ual inpatients and large outpa- tient populations. Tools are highly customizable but require enor- mous effort from individual us- ers to tailor and configure them before they become practical. One- size-fits-all functionality does not accommodate differences in users’

technical proficiencies or in indi- vidual triggers of patients’ clini- cal events. Critical tasks that are important to users, such as care coordination, are not prioritized.

What can be done to bridge this chasm? One solution might be sustained innovation programs that could foster long-term collab- oration between developers and users, incubate ideas for new IT functionalities, and facilitate rapid- cycle testing and evaluation. Pro- grams will be most effective, in our view, if they include four key characteristics.

The first is involvement of multidisciplinary teams including both developers and users. The developers may include employees of established IT firms, entrepre- neurs, and researchers. Users may include relevant clinicians and pa- tients with various disease condi- tions. To plan for implementing sustainable ideas within complex incentive structures, teams may also involve public and private pay- ers and health services research- ers. Since these participants may not interact often, it’s helpful to get people who can “speak more than one language” to serve as liaisons, especially between users and developers.

The second essential trait is a focus on users’ needs. The two key stakeholders in health care — patients and clinicians — are diverse and have complex needs and expectations. The first step toward effective solutions is de- veloping a thorough understand- ing of these needs through direct interaction with users, including interviews and observations. Tech- nology is too often based on in- correct assumptions about users’ needs, and most organizations underinvest in this critical activity.

A third key is for health care

P E R S P E C T I V E

817

Accelerating Innovation in Health IT

n engl j med 375;9 nejm.org September 1, 2016

innovators to redesign care pro- cesses in parallel with IT tools. Research from other industries shows that most IT benefits do not result from “paving the cow path.” Instead, major transforma- tions occur after intensive pro- cess reengineering to leverage the technology’s potential.3 Major IT innovations can’t be bolted onto existing health care processes ei- ther. Changes in the work of pa- tients and clinicians will require not just knowledge of current user needs, but also the imagina- tion to address latent needs that users haven’t yet considered. Such work will require a deep and sus- tained relationship between devel- opers and users, as well as fun- damental understanding of the biology of diseases.

Fourth, developers can serve users’ needs better when they have the freedom to experiment and fail quickly. Innovation pro- grams can offer an environment in which ideas can be tested rap- idly in simulated or real clinical settings, allowing users to try out innovations and provide in-depth feedback in a systematic fashion. Such environments have been called “sandboxes.”

Programs with these charac- teristics can accelerate innovation through spinoff companies, open- source technology, care models

based on redesigned workflows, provision of implementation ser- vices, and published knowledge and best practices. An innovation program can help its developers and researchers decide which dis- semination channel is most ap- propriate.

Because such programs don’t currently exist, new funding mod- els will be needed. Funding op- tions to explore include public and private research sources, health plans, and private invest- ment. Traditional approaches to the management of intellectual property used by academic cen- ters may not be effective, because the primary financial returns will probably come from first-mover advantage and early discovery of best practices. Flexibility will al- low programs to support the full spectrum of research and devel- opment, from early-stage forma- tive research through prototype development and evaluation of clinical impact.

The transformative potential of IT is no less powerful in health care than in other industries. The essential missing ingredient is a forum for innovation. Dedi- cated programs that facilitate col- laboration among developers and users will help accelerate innova- tion so that health care can catch up with the modern world. As

other industries have demonstrat- ed, there’s an insatiable demand for new, useful, user-friendly IT functionality. As emerging pro- vider-payment models take hold and providers seek tools to help them reduce costs and improve quality, the demand for new health IT functionality will grow. With sustained commitment, the IT- enabled transformations that have revolutionized so many other industries might finally come to health care.

Disclosure forms provided by the authors are available at NEJM.org.

From RAND (R.S.R.), the Divisions of Gen- eral Medicine and Primary Care (D.W.B.) and Cardiovascular Medicine (C.M.), Brigham and Women’s Hospital and Har- vard Medical School, and the Harvard School of Public Health (D.W.B.), Boston, and the Broad Institute of Harvard and MIT, Cambridge (C.M.) — all in Massachusetts.

1. Friedberg MW, Chen PG, Van Busum KR, et al. Factors affecting physician profes- sional satisfaction and their implications for patient care, health systems, and health policy. Santa Monica, CA: RAND, 2013. 2. Ratwani RM, Fairbanks RJ, Hettinger AZ, Benda NC. Electronic health record us- ability: analysis of the user-centered design processes of eleven electronic health record vendors. J Am Med Inform Assoc 2015; 22: 1179-82. 3. Jones SS, Heaton PS, Rudin RS, Schnei- der EC. Unraveling the IT productivity para- dox — lessons for health care. N Engl J Med 2012; 366: 2243-5.

DOI: 10.1056/NEJMp1606884 Copyright © 2016 Massachusetts Medical Society.Accelerating Innovation in Health IT

Reproduced with permission of copyright owner. Further reproduction prohibited without permission.