It has been proposed that innovation will play a crucial role in the viability of any organization in today’s high-tech and constantly changing commercial world.1,2 Policies of professional organizations have reflected this interest in fostering innovation in medicine. For example, the Association of American Medical Colleges (AAMC) has launched two new challenge programs designed to recognize innovations in medical education, care delivery, research, and diversity inclusion: the Learning Health System Challenge and Planning award and the Clinical Care Innovation Challenge.3 Although the exact number is not known, many medical centers have also established innovation offices or innovation centers in response to this need for innovation.
Ness4 defines innovation as creativity with a purpose. The critical aspect of a creative accomplishment is that the outcome is new; creativity can be defined as the production of goal-directed novelty.5 Innovations occur when the creative process results in novel outcomes, and people then use those outcomes in some way. The Agency for Healthcare Quality and Research Health Care Innovations Exchange defines innovation as the implementation of new or altered products, services, processes, systems, policies, organizational structures, or business models that aim to improve one or more domains of health care quality or reduce health care disparities.6
Enthusiasm for fostering innovation is laudable, but raises several important issues. We have developed a five-stage framework (see Figure 1) for examining factors involved when medical centers attempt to foster innovation, described in detail elsewhere, and which we will briefly review.7 First, the framework highlights possible barriers to adoption of an innovation strategy (stage 1). The framework also outlines the innovation strategies an institution might adopt, with philosophical assumptions underlying each. For example, an institution might design a top-down or bottom-up strategy (stage 2). In the former, institutional leadership sets the direction for innovative activities, while with the latter, staff have the freedom to innovate as they choose.
Selection of a general innovation strategy leads to decisions about the specific strategy to employ (stage 3). The first decision involves whether to develop innovations in-house or through outsourcing (stage 3a). If outsourcing innovation, through the acquisition of companies or products, little more need be done beyond deciding whether a product will be a valuable acquisition. If fostering in-house innovation, there are two strategies that can be adopted (stage 3b), defined based on an assumption concerning whether innovators are “born” or “made.” On the former assumption (innovators are born), it is assumed that some individuals are capable of innovating, whereas others are not, and therefore innovation is special. The innovation-is-special strategy centers on motivating the innovators to work toward innovation and supplying them with whatever resources they need. An example of such a strategy is the innovation tournament,8 in which employees are encouraged to bring forth ideas that are judged through a series of elimination rounds, until a small number are chosen for support.
In contrast, the innovation strategy could be based on the assumption that innovation is universal—that everyone has the potential to innovate if motivated to do so. In this strategy, efforts and resources are aimed at fostering the actual creation of ideas. Examples of such methods include Ness’s4 “ innovative thinking” program; Dyer and colleagues’9 notion of the “innovator’s DNA”; and “design thinking,”10 based on methods developed at IDEO, perhaps the world’s leading industrial design firm. These methods, which have some general similarities, are based on the assumption that anyone can innovate. Innovation does not depend on an innate set of skills or psychological characteristics but, rather, a set of skills that can be learned.
In our framework, we also discus possible barriers that may arise at the individual level, involving the potential innovators who are the targets of innovation center initiatives and programs (stage 4). Finally, we consider ways to evaluate and measure the effectiveness of a program designed to foster innovation (stage 5).
Using this framework, we designed a semistructured qualitative interview study to collect information about how a number of academic health systems have implemented strategies for supporting innovation. The objective was to describe the processes and practices used by a select sample of academic health systems to foster innovation of their physicians and staff. In the present study, we concentrated on the framework’s first, second, third, and fifth stages; for reasons of space and focus, data relevant to stage 4 will be discussed separately.
The purpose of this pilot study was to better understand the processes and practices used by a select sample of academic health systems to foster innovation among their physicians and staff. Consequently, recruitment was initiated with purposive sampling. Purposive sampling is a nonprobability sampling strategy involving the deliberate selection of participants due to some individual or group characteristic,11,12 allowing for a very specific individual, characteristic, or group to be studied.13 Our goal was to identify three academic health systems that have approached the development of innovation from different perspectives.
Sampling began with identification of an expert with a national reputation in innovation center programming: the AAMC’s chief health care officer, who was also its director of health care innovation. The identified expert then served as a key informant, providing a purposive sample of 12 potential interview sites and, if known, the first and last names of innovation center directors. An online search was completed for each sample site to gain more information about the innovation centers’ reputation, approach, and services. We specifically noted any press received, AAMC Innovation Challenge award winners, and existence of an innovation center Web site. From the information available, we identified three innovation centers using different approaches. We completed our recruitment of subjects using snowball sampling, in which the initial interview subjects recommended by the key informant (innovation center directors) suggested one to three additional individuals to participate as interview subjects (faculty or staff within their academic health system that the director knew to have a reputation for innovation initiatives).
Subjects were invited to participate in an interview via electronic mail. For each subject that responded with willingness to participate, a study information sheet was e-mailed. The University of Pennsylvania institutional review board waived written informed consent. Subjects did not receive incentives for participation.
Primary data collection included phone-based semistructured interviews. An interview guide was developed by the investigative team based on our conceptual framework of innovation in medicine,7 review of the literature, expertise in creativity and cognition, and experience in qualitative methodology. For the full interview guide, see Supplemental Digital Appendix 1 at http://links.lww.com/ACADMED/A263. Questions were piloted with the AAMC’s chief health care officer, modified, and then tested with the AAMC’s full Council of Academic Societies Innovation Task Force. All interviews were audio taped with a digital recording device and conducted between July 2012 and May 2013.
As noted above, our interests were directed by stages 1, 2, 3, and 5 of our framework. Accordingly, we questioned the respondents about (1) the barriers that arose in their institution to the general idea of adopting a policy for supporting innovation; (2) which strategy each institution adopted (bottom-up versus top-down innovation) and how that strategy was implemented; (3) whether the strategy was based on the innovation-is-special or innovation-is-universal perspective7; and (4) if the strategy incorporated methods of fostering idea generation.4,9,10 Finally, we wanted to better understand the outcomes of the programs toward the goal of evaluating their success in the future.
Data management and analysis
Interviews were transcribed verbatim by one of the authors (R.M.S.), and the audio files and transcripts were reviewed for accuracy. Transcripts were uploaded as Microsoft Word documents into NVivo Software (Version 9, 2010) for management and analysis (QSR International Pty Ltd., Melbourne, Australia). NVivo allows words, lines, and sections of text to be coded, sorted, and retrieved, both within an interview and across interviews.
The analytic approach for interview data involved the constant comparative method, an iterative coding and review process.14 The constant comparative method involves the revision of coding by taking themes that emerge from the data and rechecking them against other transcripts to investigate similarities and differences in themes across participants. Memo writing and logging code modifications and definitions were ongoing to document the process. Two authors (R.M.S. and R.W.W.) coded all transcripts. Discrepancies in coding were discussed by the team, and a consensus resolution was reached. Ethical approval for this study was granted by the University of Pennsylvania institutional review board.
Three academic health systems with a national reputation in innovation center programming agreed to participate. The specific site identities are not divulged because of the confidentiality of our seven interview subjects (three directors, three faculty members, and one staff member). Site 1, located on the East Coast, is one of the largest comprehensive pediatric medical centers in the United Sates. Site 2 is a leading public, research-oriented medical school in the Midwest. Site 3 is on the West Coast and is one of the nation’s largest not-for-profit health care providers.
All three sites carried out some form of in-house innovation strategy. Site 1 adopted a bottom-up strategy of innovation. A Site 1 interview subject described the innovation center as having several programs to support grassroots innovation. “I think of grassroots innovation as anything that is coming from an innovator as opposed to a strategic program that leadership says we’re going to focus on or innovate around,” the interview subject said. Some features of this innovation center that support grassroots efforts are a monthly forum in which anybody can present an idea or concept, an internal Web site where there can be constant dialogue and idea exchange among innovators from multidisciplinary backgrounds, and a competitive seed funding program for interdisciplinary projects.
Site 2 used a combined top-down and bottom-up strategy. The site runs an innovation fellowship program for postgraduate professionals, which includes a structured process of skill and knowledge acquisition. In the program, fellows assemble a multidisciplinary innovation team tasked with creating a medical product or service that addresses an unmet need identified through extensive observation in medical facilities and literature research, with the goal of advancing to commercialization. The site also provides resources and services, including help in finding collaborators, project management support, funding opportunities, prototyping and testing, and navigating the regulatory processes, to assist innovators at any stage of a project that has a focus on health.
Site 3 also practices a bottom-up strategy in the early stage of innovation, and ideas can come forward from any employee of the health system. However, once an idea is brought forward, there are a large number of structured processes in place to ensure that support for any given innovative idea is in the best interest of the health system. When ideas come in, they are filtered and evaluated against criteria that link to the health system’s strategic plan. An interview subject at this site stated, though, that the criteria “give us a bit of flexibility for things that we instinctually know are going to be interesting and are not yet on the radar.”
Are innovators born or made? Is innovation special or universal?
From the interviews, we found that the in-house strategies at all three sites all are based on the innovation-is-special perspective (i.e., people with ideas will bring them forward if they are motivated to do so). With the exception of Site 2, programming within the innovation centers that incorporated or fostered the “innovator’s DNA” or similar ideas toward the goal of developing idea-generating skills was limited or nonexistent.
Although it was not specifically cited by interview subjects, the Site 2 innovation center’s fellowship program has a curriculum and experiential learning component that is consistent with the behavioral and cognitive discovery skills (associating, questioning, observing, networking, and experimenting) that are typically discussed in those innovation-is-universal skill-development programs cited earlier.4,9,10 The fellowship program teaches the skills for becoming conversant in multiple domains of the health care innovation process (e.g., clinical, engineering, business, legal). Through the program, fellows are immersed in a particular clinical space, where they then observe and question any element of care delivery and find opportunities to innovate including areas outside their clinical expertise. The program also brings in people from different disciplines with training and expertise in areas beyond the backgrounds and skill sets of the fellows, creating an opportunity for networking. As compared with the other sites, Site 2 was unique in targeting some aspects of idea generation by virtue of the fellowship program. The innovation centers at both Sites 1 and 3 do have internal Web sites or message boards where innovators can network and engage in dialogue that questions the status quo, potentially making surprising connections across different areas of knowledge.
Barriers to innovation
Across all three sites, institution-level barriers to innovation were present. A common institution-level barrier was having the right resources and infrastructure in place to support the various stages of innovation development, particularly at the transition from prototype and pilot to operations. One interview subject articulated this issue particularly well, saying: “We can come up with all the greatest ideas in the world, but if we can’t implement them and scale them up then we’re not contributing a lot of value to the organization.”
Another institution-level barrier is the need to appease leadership. Innovation is an iterative, tedious process, which may not meet leaders’ need for return on their investment in a timely manner. Leaders also need to maintain the institution’s mission and agenda. One interview subject said:
You want to give people freedom to pursue projects that might not be on the leadership radar, but you also want to support things that are on the leadership agenda, otherwise innovation is not viewed as a core competency, it’s viewed as something nice and fun. Until you can prove value you’re going to be at risk. We need to have one foot in to maintain credibility and be connected to the leadership priorities, and we need to have one foot out so that we’re not caught up with doing what is always on the strategic plan.
Another institutional barrier to innovation can be the lack of physical space for people of various disciplines and backgrounds to come together. Health systems can be very siloed, and supporting and promoting multidisciplinary collaborations can be difficult without a physical space for people to congregate, or have the “water cooler effect,” as one interview subject described. Funding is often a challenge to creating the appropriate physical space, but finding a comfortable level of institutional presence is also important. “The physical entity has to have an expression … it has to be separate from the institution, but also part of the institution where input can come in.”
Outcomes for the institution
We found that the three sites had a range of metrics and criteria by which to measure an innovation’s outcomes and success. The innovation center at Site 1 has formal metrics tied to its competitive seed-funding program. To receive funding, projects need to occur fairly quickly (3–6 months), they need to involve more than one department and have cross-departmental value, they should be interdisciplinary in nature, and if the project is perceived as disruptive or really paradigm shifting it receives highest priority. The innovation center expects the grant recipients to deliver something in return, so on grant applications the applicants must state their project metrics and be explicit about what outcomes they are going to measure.
The success metrics discussed by Site 2 interview subjects were specific to the innovation center’s fellowship program. The goal of the program is to create in people unique skills as health entrepreneurs and innovators, with the hope that they will generate innovations that move off the campus and impact the world. One measure of that is the formation of a company. Another measure is for program graduates to receive traditional and nontraditional funding for the development and testing of their ideas. Ending up with a product with another company’s portfolio is also a meaningful potential outcome. Fellows also compete in business competitions while in the program, so any success in those is also significant.
An interview subject from Site 3 described her definition of an innovation success as being broad: “If we are trying something in a safe, controlled manner, where we design a test we can learn from, that to me is an innovation success.” The same interview subject also noted that “If you try something out quickly and realize there are 10 things wrong with it, that is a very successful failure in my eyes.” She commented on how this position is counter to the culture of health care, which is very analytic and wants to think through every issue in order to solve problems, which can be a slow and lengthy process.
We have used our framework as the basis to systematically review a purposive sample of three different innovation centers with three distinct approaches to supporting and fostering innovation. Among the three sites, we found several institution-level barriers to innovation. The institutions sometimes had difficulties with the limited resources available and infrastructure in place to support the development of innovative products, including simply the lack of physical space in which to bring potential innovators together. Based on this sample, institutions must realize that making a commitment to innovation may require reorganization of space and other resources, going beyond designating a part of the existing organizational structure as an innovation center.
We found that the sites differed in the top-down versus bottom-up strategies adopted, which ranged from a straightforward bottom-up strategy to mixtures of bottom-up and top-down strategies, where leadership shaped the direction of innovation. Given the significant investment of time and resources needed to support an innovation program, it is not surprising that leadership might exert direction over the process. An important question for later research, beyond the scope of this current study, is whether these variations on the top-down versus bottom-up strategies are differentially successful.
Regarding the competing philosophies of whether innovation is special or universal, we found little evidence in this sample that leadership believed innovation could derive universally from a wide range of employees. Only one of the sites attempted to focus on training a group of select innovators, while the other two focused on the development of innovative ideas brought to them. None of the individuals we interviewed at any of the innovation centers mentioned the possibility of structuring a program to develop skills for innovating and exposing most or all staff members to that program. Thus, to the degree that the innovation-is-universal-focused programs discussed earlier have been successful,15 and to the degree that our small sample allows us to generalize about innovation centers, the sites’ centers may be missing an opportunity to involve more of their staff in the innovation process.
We also found that the three sites had varying criteria for defining and assessing success, ranging from seed funding recipients laying out criteria for assessing the success of their project, to program participants forming a start-up or receiving external funding, to acquiring information from unsuccessful projects. Thus, though all sites clearly had the support of senior leadership in terms of funding, innovation center success was conceptualized in different ways. At this point we are unable to assess the relative value of those various measures of success because of the small sample and short time frame of our study and the innovation centers’ existence.
In determining an innovation’s success, the health care industry must acknowledge the complex problem of its multiple missions. In other industries, success frequently can be measured as increased profits from delivery of a product or service at an increased price or volume or with a lower cost. The mission of health care can be focused on the individual level—aimed toward delivering better health to the individual—or more globally, as delivering better health to a population. From an economic perspective, price for the product may be fixed, but profit margins are local market specific and related to reimbursement for a given health care provider. Therefore, an innovation that results in improved profit for one entity may not be generalizable to another location. Finally, innovation in the educational and research domains may be specific to entities with those missions. Therefore, any outcome assessment of health care innovation centers must include a multimission assessment.
This study demonstrated the usefulness of our proposed framework in conceptualizing innovation in medicine,7 providing valuable preliminary information about the current status of efforts to foster innovation in medicine. In the three innovation centers studied, we found a range of strategies employed to foster innovation, and a range of criteria used to assess a program’s success. Although based on a limited sample, this study provides useful information concerning the current status of innovation efforts in present-day academic health systems and suggests the need for future research. Separately, we will present information concerning staff members’ responses to the initiation of innovation strategies. Further study should examine a wider range of institutions and be carried out over a longer period of time to allow for assessment of success. The present investigation, though limited, suggests the potential of collecting data to support the assessment of innovation programs.
Acknowledgments: The authors wish to thank Joanne Conroy, MD, formerly chief health care officer of the Association of American Medical Colleges (AAMC) and currently chief executive officer of the Lahey Hospital & Medical Center, as well as members of the AAMC Council of Academic Societies Innovation Task Force, for their input on the development of the authors’ semistructured interview guide.