The Statistics Before the Statistics

Realizing It Was Never Just About the Math

Alex Zajichek, Data Scientist

Department of Quantitative Health Sciences
Cleveland Clinic

September 30, 2026

Who Am I?

• Data scientist in the Rotroff Lab
• Statistical, modeling, and software projects
• Hired in QHS, moved, rehired remotely

AI-me*

*I have justification for this

What are these slides about?

My evolution in statistical problem solving and how Dr. Kattan helped shape it.

Starting (not) at the origin

• Hired out of school by Mike. Special thanks to Jim Bena.
• When I started I was at my peak in terms of technical knowledge, statistical theory, etc. I could still do advanced calculus and make derivations
• Little did I know I knew nothing about statistics in the real world.
• Mindset was “if this then do this method”. It was objective/systematic. Run procedure, check diagnostics, calculate pvalue, GET conclusion
• I remember explicitly mikes email on
• roughly my 3rd day. “Here’s something to get your feet wet with.” It was some R code to compute a c-index (using frank harrells rms package). He wanted me to run it for some new predictions he had and send him the results. “Sure here you go!”. It was something like this: (show some R code/output).
  • “That looks backwards”. I was confused. Then he told me the minimum value is .5, you need to flip the labels.
  • I didn’t even know what this meant. He clarified it’s like an AUC. “Oh I’ve done that”. But I never had an intuitive understanding of what it was telling us. Just knew higher was “better”.
  • So then the next part: is this good? (looks at recommended thresholds). Maybe? But what if it’s better than the current standard of care/decision making process? “Hmm I never thought of it like that” I guess there is not an objectively right answer. It depends on the context.
  • This story is the first time I really began thinking that statistics is more than objective math. It’s subjective problem solving.

• Peak technical knowledge
• “Use this formula, GET answer”
• A C-Index of 0.4?

Fresh Graduate → Started at CCF

Falling down the curve

• Mike introduced me to decision curves, tradeoffs, clinical context

• Not just is the model good, but does it help make better decisions?

• We would talk about risk: flipping a coin, is it deterministic? “Given we know EVERYTHING about the coin, it’s a 1 or 0.”. We’re limited by our knowledge about the system. That determines probability.

• Applies to patients. Different patients have different risk up to how much we know about them.

• Building a no-show model forced us to think about resource constraints. The entire implementation around it might influence how you do the MATH.

• Hypoglycemia project. “This could be implemented…”. Why not DO the implementation?

• Experiences led to want to think about tangible output. Starting with problem, decision making process, and building towards that.

• Decision curves and the like
• What is risk and how it applies to patients
• Practical contraints inform the math
• More product focused

Fresh Graduate → Started at CCF

Working biostatistician

An alternative view

• Moved back to Wisconsin. Eventually joined a local health system in the Quality department. Working directly with care teams on clinical operations and executives on continuous improvement/quality initiatives.

• Started to realize how much impact can be made on patient care just from basic reporting, clean/accurate data pipelines, and just having useful information to the right people at the right time

• Began questioning the utility of p-values. CI advisor presenting results, stating p-value, and room is quiet. Nothing intuitive about it. Couldn’t “feel” the result. Needed more to ensure decision.

• Value in basic reporting + clean data
• Insufficiency of p-values (“interocular trauma”)

Fresh Graduate → Started at CCF

Working Biostatistician

Back to WI → Clinical Operations

Renewed focus

• Came back to CCF (remotely) with a renewed focus/perspective.
• Problem first. Thinking of tools, workflows, assumptions, product, END GOAL.

• The stats is only a slice of the pie
• Research diagram (the FULL product)
• Publications as byproduct

Fresh Graduate → Started at CCF

Working Biostatistician

Back to WI → Clinical Operations

Rejoined CCF

Current goals

• Probably annoy people asking “why are we doing this even?”, “what is the end goal here?”
• I still care (a lot) about statistical theory and mathematical soundness. It just doesn’t have as much of a place in getting to the solution. If you’re thinking about modeling, you are 80% of the way there.

• Product-first mentality
• Stats comes last
• Measured outcomes and impact

Fresh Graduate → Started at CCF

Working Biostatistician

Back to WI → Clinical Operations

Rejoined CCF

Current State