For years, a major priority for researchers and health professionals has been working toward the triple aim of health care—increased access, lower costs, and better outcomes. A relatively new field, implementation science (collectively referred to as dissemination and implementation science) seeks to bring new tools to the table to help achieve those goals. At its most basic level, implementation science is used to evaluate the methods of influencing systematic changes to routine care using evidence-based practices (EBPs) (Eccles and Mittman, 2006). Evaluation involves looking at barriers and opportunities to provide solutions that maximize benefits across the system. Although the integration of research into practice is not a new concept, the need for carefully evaluated and constructed implementation methods is becoming increasingly more common. For example, in 2010 an estimated $550 billion could have been saved in the U.S. health system had effective implementation methods been used (IOM, 2010). New technologies such as genomic sequencing and the use of big data through electronic health records (EHRs) and mobile health applications have enormous potential to improve health outcomes and reduce cost if implemented successfully in the clinic, but they also have their own challenges and implications. For instance, certain tools that have been implemented, such as advanced imaging, have proven to be successful, but they are also raising questions regarding their effectiveness and economic value (IOM, 2010). As
1This background paper was prepared by Roundtable staff member, Meredith Hackmann, and shared with the participants in advance of the workshop.
genomics moves from the research to the clinical space, applying an implementation science approach may be considered if the goal is to reduce gaps among quality, cost, and health.
Terms and Methods
The main approaches or methods to implementation are diffusion and dissemination. Whereas diffusion is passive, dissemination uses active strategies based on how and what the intervention is aiming to achieve (Rabin et al., 2008). Although diffusion has been the dominant approach for implementing new EBPs, more and more implementation work is focusing on dissemination. Rogers (2003) illustrates diffusion as a bell curve with five different categories of adopters based on certain behavioral traits. Much like social media, there are “innovators” and “early adopters” at the beginning who tend to be thought leaders and open to new ideas, followed by “early majority” and “late majority,” who will adopt after there is evidence of success, and finally “laggards,” who tend to be more conservative and will not adopt until there is significant evidence or pressure from others (Rogers, 2003).
One challenge in implementation science is the lack of consensus concerning terminology (Damschroder et al., 2009; Tabak et al., 2012), which arguably makes identifying methods and research strategies difficult. Some of the most common terms used in implementation science are adaptation/reinvention, or how an intervention changes during adoption; feasibility, the probability of an intervention succeeding; and sustainability, or how well implementation is maintained over time (Proctor et al., 2011; Rabin et al., 2008; University of Colorado, 2015). Adaptation/reinvention, feasibility, and sustainability all depend heavily on the context, or setting, of an intervention. While there are numerous methods for dissemination and implementation research, Tabak et al. (2012) group the methods into three categories: construct flexibility, which scores the adaptability of an approach on a scale from 1 to 5 where 1 = broad and 5 = operational; dissemination and/or implementation, which defines a method by the extent to which it focuses on dissemination, implementation, or both equally; and a socio-ecological framework, which categorizes methods according to level at which they operate (e.g., individual, organization, community, or system).
In the context of the rapid learning health care concept discussed by Charles Friedman (IOM, 2015), the current system is ineffective at feeding the translation of data back into the system to improve future care and research. Rather than learning from what does not work, the system is sluggish and continues in the same cycle, with most of the data sharing occurring in the form of journal publications. Here, applying the principles of implementation science may have potential for improving the system.
Manojlovich et al. (2015) argue that implementation often places too much emphasis on evidence when more focus needs to be on how various groups can come to a collective understanding or, in other words, how they achieve knowledge translation. This is perhaps one of the biggest gaps and one of the most important to address as further funding is allocated to comparative effectiveness research. Too often, research stops after the evidence is generated, with little regard given to how best to roll out an intervention in a given setting (Glasgow et al., 2012).
Part of the issue may be a lack of research funding, instead of a lack of awareness. The National Institutes of Health (NIH) spends roughly $30 billion on basic research and discovery per year; by comparison, the Agency for Healthcare Research and Quality spent $270 million on research related to dissemination and implementation in 2010, or 0.9 percent of the total amount spent on discovery (Glasgow et al., 2012). To further complicate the matter, only 14 percent of research is ever fully put into practice and of that 14 percent it takes an average of 17 years for the research to be fully realized in practice (Balas and Boren, 2000). However, with the creation of the Patient-Centered Outcomes Research Institute in 2010 and the NIH Collaboratory in 2015, high-quality efficacy research and demonstration projects may now be moving the field forward.
Potential Opportunities for Implementation
Implementing EBPs into routine care is a significant challenge, with one of the most important factors in adoption success being organizational culture and behavior. When the implementation of EBPs is supported on multiple levels of an organization, there tend to be higher rates of success (Aarons et al., 2015). As expected, tailored planning based on the context of an intervention is an important consideration and offers
significant opportunities for leadership. Many interventions fail simply as a result of poor planning for an intended target setting (Glasgow and Emmons, 2007). On an organizational level, the interviews with institutional leadership completed by the Roundtable on Translating Genomic-Based Research for Health found, perhaps unsurprisingly, that centralized systems and institutional policies that required compliance had greater success in implementing new practices.
Health system leadership plays a significant role in the success of EBP adoption. While implementing more efficient methods and technologies has great potential for reducing costs and increasing value to the overall health care system, individual systems can face more constraints, particularly financial ones. For instance, it is estimated that $77 billion could be saved annually if 90 percent of health care providers adopted an EHR system, though the majority of providers investing in that system will not see those cost-savings (Balfour et al., 2009). On the other hand, research also suggests that cost-effectiveness does not always lead to an EBP being implemented into practice (Clark et al., 2013).
The role of regulatory agencies could also be considered. In the United States, basic EHR adoption by nonfederal acute hospitals was 9.4 percent in 2008, but the major uptake after 2009 (from 15.6 percent to 59.4 percent) was due in part to the Health Information Technology for Economic and Clinical Health (HITECH) Act passed by Congress, which encouraged EHR implementation through incentives and grants (Charles et al., 2014). In that way, non-academic health centers and smaller systems have been able to participate as well. In comparison with the United States, many countries with more centralized health systems tend to have higher rates of EHR adoption (Balfour et al. 2009).
Implementing New Technologies
Implementing new technology shares many of the hurdles of implementing EBPs. Across industry, it takes an average of 15 years for a new technology to be fully implemented (RAND, 2005). In the case of laparoscopic surgery,2 for example, adoption has become more widespread for certain procedures, such as cholecystectomies and colectomies. A large part of the success of laparoscopy in these areas has come as a re-
2Other examples of technology for comparison not explored here may be positron emission tomography (PET) scanning, robotic surgery, smart infusion pumps, bladder scanners, VeinVue, wound vacuum-assisted closure (VAC), the use of Extracorporeal Membrane Oxygenation (ECMO), and telehealth.
sult of the medical benefits realized by patients and the financial benefits realized by health systems (NIH, 1992).
Laparoscopic cholecystectomy, a camera-guided removal of the gallbladder performed through a small incision (as opposed to open cholecystectomy, an open abdominal surgery to remove the gallbladder) began being implemented in the late 1980s. It was estimated that gallstones cost the health system $5 billion annually, with the majority of the cost coming from the long length of hospitalization required for an open cholecystectomy (NIH, 1992). In this particular case, the technology was quickly adopted, and implementation occurred very quickly because of the lack of perceived deterrents and a surplus of patient support. By 1992, the adoption rate was estimated to be 80 percent, and an NIH consensus panel had placed its seal of approval on the procedure, citing decreased pain and disability as two major benefits for patients (Allori et al., 2010; NIH, 1992). Furthermore, the cost-effectiveness of the surgical procedure was touted since it reduced post-operative complications. One of the interesting aspects of implementation was the limited research that supported the operation’s use. Clinicians eagerly adopted the procedure, seeing only benefits for patients. Because of high patient demand, research data supporting the comparative effectiveness was limited and became difficult to justify after widespread use (Allori et al., 2010; NIH, 1992).
Laparoscopic colectomy, which removes all or part of the colon through a small incision using a camera, has not had such high rates of adoption, but its use has nonetheless been growing steadily since its first implementation in the early 1990s (Bardakcioglu et al., 2013). Despite the fact that evidence from randomized trials has in this case shown benefits for patients, the adoption rate was only 31.4 percent in 2009, up from 5 percent in 2004 (Bardakcioglu et al., 2013). Of course there are other considerations with a laparoscopic colectomy—mainly that it is a more difficult procedure and has other complications that must be considered. Surgeons cite the lengthy learning process as one of the biggest impediments to its adoption (Luglio et al., 2015; Moloo et al., 2009). In addition, certain financial and socioeconomic factors have played into adoption. For instance, Bardakcioglu et al. (2013) found that private insurance was a positive factor for use of the laparoscopic procedure, while factors such as being a minority and having a low economic status were negative factors in adoption, raising important questions about the potential for new and beneficial technology to create health disparities.
Possible Prospects for the Future
With genomic medicine in its early stages, there may be opportunities to apply the principles of implementation science to inform best practices and facilitate adoption. In the context of genomics, some of the challenges may be different, but certain evidence constraints and complexities faced by large health systems have been encountered in other implementation efforts and offer learning potential for the field. Perhaps one of the lessons learned from the implementation of laparoscopic technology is that one of the key areas for engagement and effective communication could be in working with patients and consumers. As health care increasingly moves toward a patient-centric model with patients having more of a voice in their treatment options, the demand for new health technologies will likely grow. Nilsen (2015) suggests that more research should focus on how these end users impact implementation outcomes. Quality improvement for patients also means looking at the possibility that these new technologies might exacerbate health care disparities, which will require learning how to mitigate them. From the gaps explored in implementation science, it seems that a multi-stakeholder approach may provide a unique opportunity to bring about improved health and lasting change for the health care system.
Aarons, G. A., M. G. Ehrhart, L. R. Farahnak, and M. S. Hurlburt. 2015. Leadership and organizational change for implementation (LOCI): A randomized mixed method pilot study of a leadership and organization development intervention for evidence-based practice implementation. Implementation Science 10(1):11.
Allori, A. C., I. M. Leitman, and E. Heitman. 2010. Delayed assessment and eager adoption of laparoscopic cholecystectomy: Implications for developing surgical technologies. World Journal of Gastroenterology 16(33):4115–4122.
Balas, E., and S. Boren. 2000. Managing clinical knowledge for health care improvement. In Yearbook of Medical Informatics, J. Bemmel and A. McCray (eds.). Stuttgart: Schattauer Verlagsgesellschaft mbH. Pp. 65–70.
Balfour, D. C., 3rd, S. Evans, J. Januska, H. Y. Lee, S. J. Lewis, S. R. Nolan, M. Noga, C. Stemple, and K. Thapar. 2009. Health information technology—Results from a roundtable discussion. Journal of Managed Care Pharmacy 15(1 Suppl A):10–17.
Bardakcioglu, O., A. Khan, C. Aldridge, and J. Chen. 2013. Growth of laparoscopic colectomy in the United States: Analysis of regional and socioeconomic factors over time. Annals of Surgery 258(2):270–274.
Charles, D., M. Gabriel, and M. Furukawa. 2014. Adoption of electronic health record systems among U.S. non-federal acute care hospitals: 2008–2013. Office of the National Coordinator Data Brief, no. 16. https://www.healthit.gov/sites/default/files/oncdatabrief16.pdf (accessed June 13, 2016).
Clark, F., D. J. Park, and J. P. Burke. 2013. Dissemination: Bringing translational research to completion. American Journal of Occupational Therapy 67(2):185–193.
Damschroder, L., D. Aron, R. Keith, S. Kirsh, J. Alexander, and J. Lowery. 2009. Fostering implementation of health services research findings into practice: A consolidated framework for advancing implementation science. Implementation Science 4:50.
Eccles, M. P., and B. S. Mittman. 2006. Welcome to implementation science. Implementation Science 1:1.
Glasgow, R. E., and K. M. Emmons. 2007. How can we increase translation of research into practice? Types of evidence needed. Annual Review of Public Health 28:413–433.
Glasgow, R. E., C. Vinson, D. Chambers, M. J. Khoury, R. M. Kaplan, and C. Hunter. 2012. National Institutes of Health approaches to dissemination and implementation science: Current and future directions. American Journal of Public Health 102(7):1274–1281.
IOM (Institute of Medicine). 2010. The healthcare imperative: Lowering costs and improving outcomes: Workshop series summary. Washington, DC: The National Academies Press.
IOM. 2015. Genomics-enabled learning health care systems: Gathering and using genomic information to improve patient care and research: Workshop summary. Washington, DC: The National Academies Press.
Luglio, G., G. D. De Palma, R. Tarquini, M. C. Giglio, V. Sollazzo, E. Esposito, E. Spadarella, R. Peltrini, F. Liccardo, and L. Bucci. 2015. Laparoscopic colorectal surgery in learning curve: Role of implementation of a standardized technique and recovery protocol. A cohort study. Annals of Medicine and Surgery (London) 4(2):89–94.
Manojlovich, M., J. E. Squires, B. Davies, and I. D. Graham. 2015. Hiding in plain sight: Communication theory in implementation science. Implementation Science 10(1):58.
Moloo, H., F. Haggar, G. Martel, J. Grimshaw, D. Coyle, I. D. Graham, E. Sabri, E. C. Poulin, J. Mamazza, F. K. Balaa, and R. P. Boushey. 2009. The adoption of laparoscopic colorectal surgery: A national survey of general surgeons. Canadian Journal of Surgery 52(6):455–462.
NIH (National Institutes of Health). 1992. Gallstones and laparoscopic cholecystectomy. NIH Consensus Statement10(3):1–20. https://consensus.nih.gov/1992/1992gallstoneslaparoscopy090html.htm (accessed June 13, 2016).
Nilsen, P. 2015. Making sense of implementation theories, models and frameworks. Implementation Science 10(1):53.
Proctor, E., H. Silmere, R. Raghavan, P. Hovmand, G. Aarons, A. Bunger, R. Griffey, and M. Hensley. 2011. Outcomes for implementation research: Conceptual distinctions, measurement challenges, and research agenda. Administration and Policy in Mental Health 38(2):65–76.
Rabin, B. A., R. C. Brownson, D. Haire-Joshu, M. W. Kreuter, and N. L. Weaver. 2008. A glossary for dissemination and implementation research in health. Journal of Public Health Management and Practice 14(2):117–123.
RAND. 2005. Health information technology: Can HIT lower costs and improve quality? RAND Health Document No. RB-9136-HLTH. http://www.rand.org/pubs/research_briefs/RB9136.html (accessed June 13, 2016).
Rogers, E. 2003. Diffusion of innovations. 5th ed. New York: Free Press.
Tabak, R. G., E. C. Khoong, D. A. Chambers, and R. C. Brownson. 2012. Bridging research and practice: Models for dissemination and implementation research. American Journal of Preventative Medicine 43(3):337–350.
University of Colorado. 2015. Users’ guide to dissemination and implementation in health for researchers and practitioners. Denver. http://www.crispeboks.org/National/workbook-7VT-77CM.html (accessed June 13, 2016).