Discernment from data
Health systems commit capital and clinician capacity to clinical technology on the strength of vendor claims and pre/post dashboards. I read that evidence as a practicing physician who also writes the queries, and I answer two things: whether clinicians changed what they do, and whether their patients are better for it.
An independent assessment of what each tool delivers against its total cost of ownership, including the clinician time and alert burden that never appear on the invoice. Attribution is causal, so the finding holds up in front of a capital committee.
Informs: contract renewals, enterprise expansion, AI governance, retirement of low-value alerts.
The value thesis is set before go-live: baselines, targets, and an accountable operational owner for each measure. After conversion, delivered value is separated from secular trend and from the other initiatives running alongside it.
Informs: the business case, board reporting, post-implementation optimization priorities.
A plan for getting decisions out of the data you already hold: governance, the roles to hire or develop, and the first studies. For Epic organizations this includes standing up a Cosmos research program.
Informs: analytics investment, academic and quality strategy, data governance charters.
Outcome studies designed with your health system customers to the standard their clinical and financial leaders apply. The result supports renewals and expansion, shortens the enterprise sales cycle, and is publishable.
Informs: evidence strategy, customer value reviews, pricing tied to outcomes.
The buyer's view of your product: how it lands in Epic workflows, what governance and value analysis will ask, and where implementation risk sits. Grounded in executive leadership of clinical informatics on the provider side.
Informs: product roadmap, implementation model, who has to sponsor it.
A physician executive's read of the clinical claim, the evidence behind it, and the integration reality, delivered on a transaction timeline.
Informs: investment decisions, partnership terms, what to fix after closing.
A before-and-after comparison cannot separate a technology's effect from a trend already under way, a seasonal pattern, or a concurrent initiative. Quasi-experimental designs can. I developed an approach that suspends an alert deliberately and measures what real-world outcomes change, which tells a health system which interruptions earn their cost.
I apply the same attribution discipline to ambient documentation, predictive models and care programs, working directly in the data: Epic Clarity, Caboodle and Cosmos, and Cerner Millennium.
A short, fixed-scope assessment that frames the decision, the evidence available, and what a defensible answer requires.
The full study, with findings presented to your governance bodies. Or the build of a new capability, such as a research program: its governance, roles, training and first deliverables.
Ongoing counsel through a program, a build or a renewal cycle.
Tell me the decision in front of you and its timeline. I will tell you what evidence it needs and what it would take to produce.
Engagements run under your agreements and inside your systems. Patient data stays in your environment.