Stop Chasing Consensus: Why Health Systems Need Consultation to Translate AI into Impact
At this year’s Becker’s Health IT + RCM Conference, I listened as a health system CFO described how their organization comes to consensus on AI use cases and budgets. The process sounded familiar: gather the committee, review updates, weigh objections, and find the least controversial way forward.
This is the culture of health care today. Conflict avoidance is dressed up as due diligence. “Death by a thousand cuts” becomes the unspoken norm. Decisions advance only when no one strongly objects. On the surface, this feels safe and inclusive. In reality, it is one of the biggest obstacles to transforming care, especially with tools like artificial intelligence.
Consensus is negotiation. It is about smoothing rough edges, compromise, and finding a lowest common denominator. That might preserve peace, but it rarely produces bold or transformative outcomes.
Consultation, by contrast, is a discipline. It is grounded in humility and the search for truth. Participants put aside personal agendas, surface evidence openly, and test ideas rigorously. Once a decision is reached, even those who initially disagreed commit fully, knowing that unity in implementation matters more than unanimity in opinion.
History shows us what happens when consensus overrides consultation.
On January 28, 1986, the space shuttle Challenger broke apart 73 seconds after launch, killing all seven astronauts on board. The technical cause was clear: O-ring seals in the rocket boosters failed in the unusually cold weather. But the deeper cause was cultural.
Engineers at Morton Thiokol, the contractor responsible for the boosters, raised urgent warnings. Data showed the O-rings became brittle in cold weather and could not guarantee a safe seal. They recommended postponing the launch.
Instead of asking, “What is the truth, and what does the evidence show?” NASA managers reframed the discussion: “Can you prove it is unsafe?” The focus shifted from truth-seeking to risk-avoidance. Under pressure to keep the launch on schedule, Thiokol executives overruled their own engineers. Consensus was reached not because everyone believed the launch was safe, but because dissent had been suppressed.
This is what consensus does. It creates silence where candor is needed. It prioritizes harmony over truth.
Now imagine the same situation approached with consultation. Engineers would not have been worried about their careers or reputations if they spoke up. Their concerns would have been heard as contributions to a collective search for truth, not as challenges to authority. The questions would have been different. Instead of “Can we prove it is unsafe?” the group would have asked:
“What do we know with certainty about the O-rings in cold conditions?”
“What are the gaps in our evidence?”
“What is the worst-case scenario if we launch today?”
“What principles should guide our decision when lives are at stake?”
Consultation would have shifted the culture from pressure and self-preservation to openness and shared responsibility. Once the decision was made to delay, all parties would have united behind it, presenting a single voice to the public and political leaders. In that process, truth would not have been sacrificed for expediency, and lives would have been saved.
Health systems face a similar cultural challenge. AI is not a technical add-on. It is part of a larger transformation process that affects workflows, finances, and patient outcomes. Yet too often, AI projects are handed to committees where consensus-seeking dominates. Each department voices concerns such as compliance, IT, finances, and clinical workflow. The energy quickly shifts from truth-seeking to compromise. What emerges is a watered-down pilot that inspires no one, consumes resources, and ultimately changes little.
Consultation offers a different path. In consultation, participants put aside personal or departmental agendas, are not cynical about their impact, nor are they calculated a bigger strategy for career advancement. A nurse’s frontline experience, a physician’s clinical perspective, a financial officer’s realities, and a data scientist’s models are all tested together. Biases are surfaced. No plan is perfect, but once the wisest path forward is identified, all commit to it. That unity of action is what drives results.
To move away from consensus-seeking and toward consultation takes practice. It requires building a sustainable process at the organizational level, with a health systems perspective. Through my work at the School of Public Health with Professor Rifat Atun and with applied field experience, we are focusing on exactly this: the questions leaders must ask to create transformation processes that adapt not only to emerging technologies like AI, but also to external shocks and systemic challenges they are facing.
Ultimately, this is a change in consciousness from all involved, moving away from pure tactical execution to a process that starts with a systems view that produces clarity and unity in vision, thoughts and ultimately action.
Finally, tools like Six Sigma, Lean, Total Quality Management, Agile, and Kaizen are all valuable, but they are subsets of a higher order of thinking, which is consultation. Without consultation, they risk becoming operational checklists rather than guides to the right decision. In fact, consensus with these tools can be more dangerous than consultation without them.
Consensus will keep us busy. Consultation will make us better.
If you are interested in learning more about how to make this shift in your organization, or what we are doing to enable leaders at the HSPH, feel free to reach out.
Linkedin: https://www.linkedin.com/in/salimafshar/