Recent industry analysis indicates notable acceleration in the adoption of AI over the next 24 months. CIOs can use insights from these peer activities to gauge the maturity and alignment of their HDO against those of their industry peers.
- The majority of healthcare executives realize the game-changing nature of artificial intelligence, (AI) and are increasing investments in AI and advanced analytics.
- Improving clinical outcomes, advancing digital transformation and improving operational and clinical performance are the priorities driving investment in AI.
- The lack of healthcare delivery organization (HDO) IT funding is one of the biggest impediments to moving AI initiatives forward more rapidly. However, cultural impediments, including lack of trust, skills and knowledge, also hamper adoption. And AI governance is woefully lagging in projected adoption rates.
- All types and sizes of HDOs are delivering AI — not just larger or research-focused organizations. Many smaller HDOs realize the value AI offers and are actively pursuing it.
Provider CIOs responsible for healthcare analytics strategy and innovation should:
- Compare your HDO to your peers by selecting what would have been your responses to the AI survey questions, and prepare an action plan for your AI strategy and investments.
- Get a jump on addressing changes in this rapidly developing space by identifying business opportunities for your HDO. Budget for increased investments in AI by using industry evidence presented in this research note to create the imperative.
- Prepare for AI impediments by augmenting and complementing current analytic governance with strong AI governance that safeguards the benefits and protects against the risks and vulnerabilities when implementing AI capabilities.
- Invest in skills and talent by creating a center of excellence responsible for education, training, and executing governance best practices.
In November 2018, Gartner collaborated with the College of Healthcare Information Management Executives (CHIME) to survey healthcare provider CIOs in their adoption of AI. The survey was conducted to understand where healthcare providers are in their perception of the value AI offers, which priorities are driving investments, and how prepared CIOs believe their organizations are to embrace and deliver AI.
The need for advanced analytics and AI as a center of excellence (COE) does not discriminate — all HDOs, regardless of size or nature, will need to drive change with insight that only analytics can offer. Even smaller, specialized HDOs wishing to survive acquisitions, consolidations or closure cannot afford not to invest in a foundation of analytics and AI excellence. Advanced and real-time analytics and AI enablement offer the potential to decrease medical errors, improve diagnostic accuracy and outcomes, help optimize revenue, and streamline operations. AI lies at the heart of every digital era initiative for healthcare improvement and transformation.In Gartner’s 2019 CIO Survey (see “2019 CIO Agenda: Healthcare Provider Industry Insights”), healthcare provider CIOs listed data and analytics (including predictive analytics) and AI as the top game-changing technologies. These findings align with Gartner’s 2018 predictions about the relationship of advanced analytics and the very essence of creating growth and health value. In “Predicts 2018: Healthcare Providers,” Gartner predicted that “by 2020, 70% of healthcare provider CIOs will cite advanced analytics as their top priority” based on three hypotheses:
- Unprecedented industry transformation continues to demand cost and operational efficiencies, and the improved health status of citizens.
- More-complex economic models are changing the way outcomes get measured and value gets assessed.
- The momentum of AI and the rapidly developing space will quickly demand new strategies, talent and technologies.
All indicators point to the end of the era of AI simply being the bright shiny object blinding the floors at vendor showcases. Business and IT leaders who have watched with caution, and cynically snickered at early public flops (or who maybe had a failed project of their own), need to reconsider. The survey findings help substantiate the conclusion that AI has advanced beyond the hype and is being applied to solve very real problems.A total of 50 CIOs participated in the CHIME survey. The mix in size and type of HDOs reflects a diverse base. Thirty-five percent of the CIOs represented HDOs had fewer than 250 beds, 10% had between 251 and 500 beds, 31% had between 510 and 999 beds, and 24% had more than 1,000 beds. Additionally:
- Twenty-seven percent identified themselves as an academic medical center; these belong primarily to multihospital systems. An additional 27% identified their HDO as part of a multihospital system, but with no affiliation to an academic medication center.
- Twenty-four percent identified themselves as a single hospital.
- Twelve percent indicated an ownership relationship with a payer (either owned a payer organization or was under the same ownership as a payer organization).
Overall results of the survey indicate increased adoption of AI, particularly to help improve clinical outcomes, advance digital transformation and improve performance. This will gain momentum in the next 24 months. Significantly, CIOs cited a lack of IT funding (58%) as the biggest impediment to moving AI forward more quickly.Some survey questions were further vetted during a CHIME AI focus group conducted by Gartner at the Health Information Management Systems Society (HIMSS) annual conference in February 2019. Results of the CHIME survey were compelling enough to want to validate, and the CHIME focus group presented the perfect opportunity. Twenty-three CIOs/other executive leaders participated in an informal discussion and polling of the survey questions. The results from the focus group largely concurred with the survey results and yield to conclude that the industry is moving, almost at an accelerated pace, toward much broader adoption of AI. However, the group also recognized the many challenges regarding skills, readiness and data.Results of the survey and focus group are further explored below.
Value Perception and Priorities
To gauge executive leadership perception of the value AI offers, the survey asked, “How does your executive leadership view AI in meeting critical business challenges?” Respondents were asked to select from the following answers:
- A top priority for funding
- One of many overhyped technologies
- A necessary technology to improve cost-effectiveness
- A strategic capability to improve performance
- An innovation to be experimented with and considered cautiously
Fifty percent report AI as a strategic priority to improve performance and cost-effectiveness — although only 4% consider it a top priority for funding (see Figure 1). Eight percent believe it is still overhyped, while the majority (38%) believe it needs to be better understood through cautious experimentation and understanding how best to apply the various AI capabilities to achieve value. The results were further confirmed during the CHIME AI focus group where, of the 23 participants, close to 54% agreed that AI will be a strategic capability needed to remain competitive and meet ongoing industry challenges. Only one person in the focus group responded that their senior leadership still perceives AI as being overhyped.Figure 1. Executive Leadership Perception of AIn = 50
Source: Gartner (March 2019)
A deeper analysis of the survey data shows an even distribution across all types and sizes of organizations regarding how this question was answered. For example, contrary to instinct, the majority of CIOs who reported that their leadership viewed AI as overhyped were not the smaller, under-250-bed hospitals, but rather the midsize hospitals in the 501-to-999-bed range. And academic medical centers are not necessarily leading the charge — their answers were distributed across the choices. The 4% of respondents that said it was a top priority for funding were from institutions with more than 250 beds and those with few than 1,000 beds.To understand investment decisions, the survey asked, “What strategic priorities are driving your executive leadership team’s ambition to pursue AI?” Respondents were asked to select three of the following:
- Digital transformation
- Financial performance improvement
- Operational performance improvement
- Clinical: improved clinical outcomes
- Clinical: improved satisfaction and clinical workflow efficiencies
- Patient and consumer: engagement and satisfaction
- Population health management and value-based care
- New value delivery within one or more service lines (example: new AI-guided imaging diagnostics)
- Other (please specify)
Improved clinical outcomes came out the overwhelming top priority driving investment in AI in the survey results (see Figure 2) as well as the CHIME/Gartner focus group, where 82% selected this category. These priorities align with top priorities common to the industry (see “Business Drivers of Technology Decisions for Healthcare Providers, 2019”). Foundational priorities of managing the finances of the organization and delivering quality care to patients are pervasive, and they are reflected in CEO and ministry of health statements worldwide. The focus group also identified research and provider satisfaction as other strategic priorities driving AI investments.Figure 2. Strategic Priorities Driving AI Investmentn = 50
Source: Gartner (March 2019)
These strategic priorities can be directly translated into AI use cases that reflect emerging applications and vendor solutions. Investments in AI to help improved clinical outcomes are often via algorithmic medicine, which is an umbrella term for the use of complex clinical algorithms to aid and drive clinical decisions to streamline, improve and standardize medical care. Neural nets are often trained over time to increase the accuracy and specificity of the algorithms. To help manage the financial viability of the organization, AI is being applied to support enterprise revenue cycle initiatives that capture opportunities such as reducing leakage, improving billing and reducing claims denials (see Table 1 for use case definitions).
The CHIME/Gartner survey asked respondents to indicate which AI use cases their organization has implemented, has plans to implement within the next 24 months, or has no plans to implement at all. The results were striking (see Figure 3). While current adoption rates are marginal, in almost every category, projected adoption over the next 24 months is impressive, leading to the conclusion that most HDOs will find themselves implementing one or more AI capabilities in the near future. The participants in the CHIME focus group agreed. Algorithmic medicine is being used by most organizations in some capacity, with some homegrown development by via vendor partnerships. Adoption patterns are also well-aligned with capability maturity. For example, virtual personal assistants are still nascent, so it’s not unexpected that they will have the lowest adoption within 24 months, whereas the category of virtual care monitoring and real-time operations management is significantly more mature and mainstream (see “Hype Cycle for Healthcare Providers, 2018”).Figure 3. Indicate Which of the Following Use Cases Your Organization Has Implemented or Plans to Implementn = 50. Totals may not equal 100% due to rounding.
Source: Gartner (March 2019)
Assess your executive leadership team’s position on AI and strategic imperative for funding and ask yourself how you would score your HDO against the survey questions. Bring a common understanding of definition and value to the executive team so a universal language is being used and more productive conversations can be pursued. Strong momentum and ample evidence suggest that the time of being a passive observer has passed. Use the survey insights to encourage deeper awareness around AI, cite industry adoption and translate that into competitive positioning — especially if you feel your leadership still views AI as an overhyped technology.CIOs can help executive leadership find the right opportunities for their HDO. The survey results indicate a significant uptick in activity in AI over the next 24 months that will be game-changing and differentiating. The ambitions need not be a moonshot — keep it simple — but look for ROI. Consider the following:
- Revenue Cycle Optimization: Natural language processing and machine learning are being used to increase coding efficiencies and optimize submission of the claim first time, resulting in decreased underpayments and denials. Revenue cycle optimization offers both a clear financial ROI as well as administration efficiencies. CIOs can work with the CFO to understand what is offered through their existing revenue cycle vendor; which capabilities might need to be augmented, and the right timeline and approach for introducing AI into the revenue cycle workflow.
- Diagnostic Imaging Interpretation: This field of AI is rapidly maturing, offering multiple tiers of benefits (from screening efficiencies to triage to diagnostics). CIOs can track the evolution of these vendors through Gartner’s Hype Cycles (see “Hype Cycle for Healthcare Providers, 2018”), and should be engaging radiology and pathology leadership in a discussion and evaluation of the market.
- AI for Virtual Care Monitoring and Real-Time Operations Management: This AI capability is projected to have the highest penetration rate within the next 24 months. Adoption is being driven largely by the acceleration in virtual care capabilities. However, the movement to implement command-and-control center dashboards for real-time operational efficiencies, especially throughput and capacity management, is also gaining momentum. CIOs should be working with clinical and operational leadership to determine opportunities for complete re-engineering of how clinical care and hospital efficiencies are monitored and managed.
Progress in organizational alignment around AI adoption is evident (for example, 30% of CIOs reported that they have appointed a single leadership role for accountability), yet challenges around skill sets and trust are relatively universal. In Figure 4, the survey asked how respondents organized around evaluating, selecting and deploying AI. The following options were offered:
- Established a center of excellence (COE) for concentrating AI skills and activity
- Established a separate governance process for AI projects
- Appointed a single leadership role for accountability
- Developed a verification methodology for evaluating the reliability, accuracy and performance of AI algorithms
- Established incremental responsibilities and accountabilities for clinical applications
- Established formal procurement processes for evaluating AI vendors and/or products
The results show that many organizations still operate on a need-by-need basis, with little formal structure. While a deeper data analysis shows it is typically the systems with more than 1,000 beds that employ more rigor around verification, governance and more formal processes, once again the distribution across size and type of hospital is strikingly dispersed. And only 16% percent of the respondents indicate formal governance for AI.Figure 4. How Have You Organized Your AI Efforts for Evaluating, Selecting and Deploying AI?n = 50.
Source: Gartner (March 2019)
The absence of more formal structure and process for AI may also be a factor in some of the reasons why AI isn’t being more aggressively implemented. Aside from the top impediment shown in Figure 5 (lack of IT funding), the overwhelming theme is the lack of skills, trust and business case (or value articulation). Surprisingly, data issues, while noted, were not at the top of concerns for the survey respondents, but were noted as a top barrier by the focus group. And while governance was also low on the list of inhibitors, it could be argued that proper and effective governance in fact would also help to:
- Mitigate many of the cultural barriers causing friction and lack of trust
- Create solid business cases and articulate value
- Enable better investment and funding
- Increase literacy and skills
Figure 5. Barriers to AI Implementation = 50
Source: Gartner (March 2019)
Most CIOs are confident that their staffs are reasonably well-equipped to support, maintain and improve AI solutions (see Figure 6). However, being reasonably well-prepared is not sufficient. After all, the IT team, including data scientists and engineers, analysts and database managers, will have a role — and be accountable for — sustaining the accuracy, integrity and reliability of the algorithm. This survey finding, the marginal confidence that IT staff is prepared, is a direct corollary to the finding that 58% of CIOs indicated a lack of IT funding as being an impediment to implementing AI more aggressively. The inability to invest in the right resources will hamper even the best intentions. Figure 6. IT Readiness = 50
Source: Gartner (Gartner 2019)
- Implement AI Governance: The need to implement AI governance has been noted as a top action for CIOs in 2019 (see “Healthcare Provider CIOs: Get Ahead of AI Innovation With Strong AI Governance”). The survey and focus group substantiated the observation that a notable ramp up in adoption will occur across the revenue cycle, including AI for diagnostic imaging interpretation, Al for automation of repetitive tasks, and AI for remote clinical and operational monitoring (see Figure 3). This is why AI governance is necessary. AI efforts will be challenged to be successful and realize ROI without organizational governance. Industry adoption of AI is very real and it is vital for you as CIO to take a lead role in ensuring discipline and accountability around the use of AI in your organization. Use the AI governance council to facilitate discussions and formally adopt an enterprise perspective, and to establish rigor around AI opportunity, identification and selection. Establishing governance for AI sends a clear signal to the organization that AI is considered strategic and has the attention and interest of senior executives.
- Invest in Skills and Talent: In Gartner’s 2019 CIO Survey, healthcare providers overwhelmingly noted data and analytics as the top area for increased investment in 2019; AI was No. 6 (see Figure 7).
Figure 7. Funding Outlookn = 106. Multiple responses allowed.
Source: Gartner (March 2019)
Investment funding must go beyond tools. The lack of clinical and business skills was noted as the top barrier to more aggressively implementing AI. You must invest in data scientists and engineers. Whether your AI investments are spent on vendor partnerships (which a large portion will be) or used internally, data science expertise is required to verify the algorithms, provide the datasets and make sure integrity is sustained.End-user data literacy skills are needed. These skills are not associated with navigating through an end-user tool, they are associated with analytic and data literacy. It’s about understanding how to ask different questions of the data and be data-driven in solving problems. Gartner calls this “information as a second language.” Whether translating to the board how data and analytics can manifest in digital transformation and industry use cases, explaining how to blend internal and external datasets, or describing advanced analytics techniques, data and analytics represent the new language of the digital economy (see “Information as a Second Language: Enabling Data Literacy for Digital Society”). CIOs can establish a center of excellence to bring ongoing training and education to the enterprise and to help democratize analytic expertise.
Gartner’s scan of the healthcare provider industry produced compelling results. The survey and the additional validation via the focus group reflect an industry mobilizing around AI. As a CIO, you have choices. You can either lead and get a grasp on how to utilize AI for improved performance and outcomes, or sit back and passively observe the capabilities (and industry) mature around you. Your approach will probably be a hybrid of both. Early opportunities will emerge and be tested, longer-term opportunities will be identified and watched. Use insights from the industry survey to build up governance, skills and talent. Make sure senior leadership is engaged and views AI as a strategic necessity in the digital transformation journey.
The CHIME/Gartner CIO survey was conducted via the CHIME survey website in November 2018. A total of 50 responses are included in the summary statistics. The CHIME/Gartner focus group was conducted during the annual Healthcare Information and Management Systems Society (HIMSS) conference in Orlando, Florida, in February 2019.