Healthcare leaders rarely suffer from a shortage of problems. Falling volumes, workforce shortages, access constraints, financial pressure, capacity limits, new competitors, and changing patient expectations may all be real at the same time. The harder strategic question is: which problem is actually governing the outcome?

In Episode 2 of Leading Healthcare AI, we examine how healthcare leaders can identify the crux: the consequential and addressable strategic challenge where concentrated action could materially improve an enterprise outcome.

Using the case of a regional health system losing patients and referrals, we work through how to separate symptoms from causes, targets from constraints, and operational bottlenecks from genuine strategic problems. We examine competing explanations—patient access, clinical capacity, referral friction, economics, the service proposition, and structural market change—and show why each would lead to a different strategy, investment decision, and role for AI. 

The episode also connects Richard Rumelt's concept of the crux with the Playing to Win strategy-choice cascade. Playing to Win helps establish where the organization intends to play and how it intends to win. Crux diagnosis asks what consequential difficulty is preventing the organization from succeeding within those choices.

We explore a simple but demanding test:

If we substantially removed this constraint, would the enterprise outcome materially improve?

That question helps leaders pressure-test whether they are concentrating resources on something that matters—or improving a process that leaves the underlying strategic problem untouched.

The implications for AI are important. Different diagnoses produce different AI strategies. If access and referral friction are the constraint, AI may help classify referrals, extract information, improve routing, match patients with clinicians, forecast demand, and detect leakage. If clinical capacity is the constraint, better AI-enabled scheduling cannot create capacity that does not exist. If contracting or the service proposition is the problem, AI may play a supporting role rather than being the primary intervention.

The diagnosis determines the role of AI. The existence of AI does not determine the diagnosis.

We close by treating the diagnosis as a hypothesis rather than a permanent conclusion: What evidence supports it? What evidence would cause leadership to change it? And if the supposed constraint improves but the enterprise outcome does not respond, when should leadership reopen the diagnosis?

Books referenced in this episode

Richard Rumelt — The Crux: How Leaders Become Strategists
The Crux on Amazon

A.G. Lafley & Roger L. Martin — Playing to Win: How Strategy Really Works
Playing to Win on Amazon

Healthcare AI strategy for hospital executives and health-system leaders — a diagnosis-first approach that turns AI possibilities into measurable operating capability. This episode tracks one real enterprise challenge — losing patients and referrals to competitors and new entrants — through the full strategic sequence, so you can see how each step shapes the next.

In this episode, you'll learn:

• Why AI use-case lists are evidence of possibility, not strategy
• How to read the changing environment before defining the challenge
• How to identify the critical operating flow where concentrated action produces the most progress
• The four roles AI plays in a healthcare response (including when AI is unnecessary)
• How readiness, decision rights, and measurement determine whether AI produces value