This article was co-written with Tom Lawry, managing director of Second Century Tech, and Dr. Spencer Dorn, vice chair and professor at the University of North Carolina School of Medicine.
The recent surge in AI enthusiasm has encouraged big thinking. LinkedIn feeds are brimming with visionary AI solutions promising to rescue healthcare.
Indeed, these are “exciting times,” as numerous posts often conclude. But it’s time now for leaders to get down to business and focus on refining their existing practices while innovating new ways of providing healthcare.
In other words, thriving in today’s complex and dynamic environment requires “organizational ambidexterity.” That refers to an organization’s ability to simultaneously exploit its current capabilities while exploring new opportunities.
In healthcare, this “exploration-exploitation trade-off” translates to leveraging established practices and protocols that have stood the test of time, while also embracing innovative technologies, treatments, and management strategies that could potentially revolutionize patient care and operational efficiency.
To help guide leaders on this journey, we’re exploring six steps for optimizing the value of both tradition and transformation in healthcare.
Organizational ambidexterity in practice
On one side, there’s a rich heritage of medical knowledge and practices honed over decades, forming the backbone of healthcare organizations. On the other side, a relentless wave of innovation – digital health technologies, cutting-edge treatments, and novel patient care models – drives healthcare into new frontiers.
Caught in the middle of this tension between tradition and transformation are healthcare leaders.
They face the challenge of balancing the need to maintain and incrementally improve their organization’s existing processes while simultaneously pursuing radical change. If they over-rely on traditions, their organization can get left behind, while over-emphasizing transformation can lead to reckless investments and potential threats to patient care.
To achieve organizational ambidexterity, the following process works to create a competitive equilibrium between short-term performance and long-term success.
Step 1: Define clear objectives
This is key. For example, UNC Health identified a clear need to revamp the patient messaging process to ease clinician workload. By adapting their policies, educating patients and staff, adjusting EHR configuration, and developing team-based workflows, they reduced the volume of patient messages reaching clinicians’ inboxes by 25% and cut the time staff spent on messages by up to 60%. Rather than exploring new digital solutions, UNC optimized its existing tools to improve outcomes. In other words, a clear objective, rather than the hype surrounding emerging tech, drove their decision-making.
Leaders should first optimize existing systems before jumping into new technologies. Start by identifying specific pain points and measurable objectives. Audit your current workflows and systems to identify inefficiencies. Train staff on better practices and test incremental changes. This approach maximizes current tools to quickly deliver results without large-scale investments.
- Ask: What processes can be (and need to be) improved with existing resources and minimal disruption?
From an exploration perspective, The Permanente Medical Group (TPMG) harnessed new technology to reduce clinician documentation burden. Over a ten-week pilot period, one in three eligible clinicians used “ambient intelligence” to automatically generate clinical notes. Participating doctors spent significantly less time documenting without degrading note quality. In this example, the objective exceeded the capabilities of existing tools, prompting TPMG to embrace innovative tech to meet clinicians’ needs. Again, it was the objective, not the hype, driving decision-making.
When current systems fall short, look to new technology that directly solves the problem. Identify gaps and then explore novel solutions like GenAI. Leaders should pilot emerging technologies to ensure they meet clearly defined objectives before wide adoption. The key is focusing on technology as a solution to an existing, unmet need, not as a trendy innovation.
- Ask: Does this new tool directly address a defined objective that our current system can’t handle?
Step 2: Build cross-functional teams
Diverse perspectives remove bias blind spots and enhance effectiveness. For example, Texas Children’s refined existing practices by forming a cross-functional workgroup (clinicians, research nurses, data specialists, and others) to implement a comprehensive “Three Systems Approach” – analytics, content, and deployment systems – to improve appendectomy procedures.
Using real-time analytics and deploying evidence-based content into daily operations, the team co-created improved care practices, increased clinician trust, and ultimately boosted performance.
Done right, these cross-functional teams bring unmatched clarity to addressing inefficiencies through a culture of cooperation.
- Ask: What mix of stakeholders is needed to address current challenges in our processes using existing resources?
When it comes to exploring new ways of doing things with cross-functional teams, Mayo Clinic partnered with Google to develop the “AI Factory,” a hub of consultants, educators, clinical staff, and legal experts established to remove silos and barriers, allowing diverse groups to collaborate on cutting-edge projects. Here, Mayo’s cross-functional approach helps orchestrate the cohesive and safe exploration of new technologies, avoiding duplication of efforts and risky initiatives.
Put simply, in addition to leveraging cross-functional teams to refine existing practices, leaders must utilize diverse perspectives to scale emerging solutions safely.
- Ask: What mix of stakeholders is needed to safely pursue innovations aligned with our strategic goals?
Step 3: Assess readiness for change
The next step – readiness for change – isn’t solely about large-scale innovation. It’s equally about refining existing business practices. For example, one of the authors worked with Omada Health to develop 10 KPIs for assessing readiness to leverage data. By focusing on the fundamentals of change and introducing simple, data-driven adjustments to existing practices – such as setting alerts for providers to follow up with patients within seven days of their initial encounter – they increased patient Net Promoter Scores by six points in just six months.
The key here is assessing whether your organization is fully utilizing existing tools. Conduct regular audits of current processes and systems to ensure they are being used efficiently and introduce incremental changes that align with organizational goals.
- Ask: Are we ready to make the most of what we already have?
On the exploration side of readiness, Intermountain Healthcare established the Data Science and Artificial Intelligence Center of Excellence to establish standards, set goals, and guide change. A part of that initiative was implementing an “AI Playbook” to “deploy and scale human-centered AI that is transparent, equitable, ethical, and, above all, ensures patient privacy.”
The playbook “outlines goals to establish appropriate AI governance, set validation and documentation standards, detect inherent bias, ensure data integrity and promote AI literacy among caregivers.” It’s an excellent example of developing a strategic innovation agenda to assess and boost readiness. This groundwork promotes safe and effective adoption of, and satisfaction with, novel technologies.
If new solutions are needed, readiness for change should focus on ensuring that your organization is prepared for the risks and opportunities of innovation. Develop governance structures, clear guidelines, and processes to manage the safe adoption of new technologies. Use pilot initiatives to assess how well your team can integrate new tools without compromising ethical standards.
- Ask: Is our organization prepared to adopt and scale new technologies responsibly?
Step 4: Allocate resources appropriately
Ambidextrous organizations also critically examine how to balance tradition and transformation from a financial perspective. To accomplish this goal, the Mayo Clinic Platform outlined a rubric for tech-specific budgeting, which includes the following questions:
● Is it strategic? It may be really cool tech, but what is the imperative, and is it addressing a current need?
● Is there a quality or safety imperative? Don’t become a healthcare institution of lesser quality by standing still.
● Is there a compliance or regulatory mandate? If something could result in an action against you, pay attention.
● What’s the impact factor? Do what impacts more patients, providers, social workers, and other stakeholders.
● Is there a revenue impact? Is there a cost reduction or significant impact to profit and loss?
Leaders must ensure they dedicate resources to clearly defined strategic, patient-centered goals. Prioritize investments in solutions that improve patient outcomes, ensure quality, and meet regulatory standards.
- Ask: Are we investing in innovation that aligns with our strategic goals and delivers real value or are we investing in new technology because it’s flashy and exciting?
Ambidextrous organizations also adjust their investments over time. For example, UW Health puts its tech budget into three buckets: running the business, growing the business, and transforming the business. UW initially allocates most dollars to run the business, and as efficiency increases and costs go down, it redirects more funds to the growth and transformation buckets. This staged approach prioritizes the core business while working towards exploration.
The key here is focusing resources on improving efficiency within existing systems before shifting funds toward innovation. Develop a clear plan for managing operational costs and reinvest savings into growth or transformation initiatives.
- Ask: Are we allocating enough resources to sustain and optimize our operations?
Step 5: Foster a culture of experimentation
Next, experimentation is core to both optimization and innovation. On the optimization side, MD Anderson Cancer Center researchers used patient flow analysis (PFA) to surface workflow inefficiencies and then developed and tested various tactics to improve flow. Establishing such teams, with deep knowledge of existing practices and a mandate to experiment safely, optimizes existing resources and enhances efficiency.
Encourage people to challenge the status quo. Then provide them with autonomy and the tools to run small experiments, driving teams to optimize existing processes based on real-time data. Reward these efforts, even those that do not achieve immediate results, to foster a mindset that values learning.
- Ask: Where can we empower teams to test incremental improvements in our current processes without disrupting overall operations?
When it comes to experimenting with new solutions, researchers at NYU Langone Health developed NYUTron, a large language model-based system that integrates in real-time with clinical workflows centered around documentation and order entry. This model outperformed traditional methods for predicting 30-day readmission rates and other outcomes such as insurance denial and length of stay. NYU’s experience highlights the importance of building lean, agile teams and equipping them with the technical skills to rapidly explore and innovate.
When introducing new technologies, create controlled environments to experiment without affecting core functions. Ensure that novel solutions are rigorously evaluated and aligned with clear objectives (see Step 1), then scale successful experiments across the organization.
- Ask: How can we create safe spaces to experiment with emerging technologies that align with our objectives while ensuring measurable outcomes?
Step 6: Integrate feedback loops
When it comes to feedback, Cleveland Clinic developed its Severity Scale to categorize, assess, and triage patient complaints and provide real-time insights into specific processes and practices needing attention. After finding that 25% of grievances were related to communication breakdowns, their leadership teams, armed with this granular knowledge, could rapidly create tailored patient experience strategies, set precise targets to improve performance, and prioritize resources with greater accuracy.
Leaders should establish feedback mechanisms that allow for continuous improvement. Use feedback collected through surveys or focus groups to identify inefficiencies and refine existing processes.
- Ask: How are we using feedback to enhance our current operations?
Establishing formal feedback is also crucial for exploring emerging technologies. For example, during the development and deployment of an AI tool called SepsisWatch, Duke Health researchers used end-user feedback to continually refine the tool to the point clinicians wanted to use it. The key was drawing on established design principles, utilizing subject matter experts, and investing in talent such as user experience and user interface professionals to elicit and leverage user feedback.
When adopting new ways of working, continuous feedback is essential. Focus obsessively on stakeholder input to refine tools and ensure they drive outcomes aligned with clearly defined objectives (see Step 1).
- Ask: Are we incorporating feedback effectively to adapt new technologies to our team’s needs?
Going forward
Healthcare leaders who follow these six steps can simultaneously exploit their organizations’ existing strengths as they explore developing new capabilities.
This ambidexterity enables them to navigate the uncertainties of healthcare’s evolution and best ensure their organizations effectively balance tradition with transformation.
Brian R. Spisak, PhD is the chief people and communication officer and senior partner at Csuite Growth Advisors, which focuses on innovation and digital transformation in healthcare. He’s also the Program Director of AI and Leadership at the National Preparedness Leadership Initiative at Harvard University, a faculty member at the American College of Healthcare Executives, and the author of the book, Computational Leadership: Connecting Behavioral Science and Technology to Optimize Decision-Making and Increase Profits.
Tom Lawry is managing director of Second Century Tech LLC, and a former Microsoft exec who served as National Director for AI for Health and Life Sciences, Director of Worldwide Health, and Director of Organizational Performance for the company’s first health incubator. He’s also a former health system executive and board member and the author of the best-selling book, Hacking Healthcare – How AI and the Intelligence Revolution will Reboot an Ailing System.
Spencer Dorn, MD, a practicing gastroenterologist, is vice chair and professor of medicine at the University of North Carolina School of Medicine, and a Lead Informatics Physician at UNC Health.
