The Tool That Shouldn't Exist

Imagine you are a billionaire who needs a kidney. You are worth $4 billion and your estimated remaining lifespan without a transplant is 18 months.

You call your assistant. Your assistant calls a health care consultancy. The consultancy assigns an epidemiologist. The epidemiologist opens a spreadsheet, pulls public data from six federal databases, and begins an analysis that will take roughly two weeks and cost you somewhere between $50,000 and $200,000.

Here is what that analysis looks like.

---

The variables nobody tells patients about

In Part 1, we established that the US organ transplant system is not one list. It is a patchwork of 56 Organ Procurement Organization territories, hundreds of transplant centers with independent acceptance criteria, and a geography-driven allocation framework that creates enormous variation in wait times by location.

The billionaire's epidemiologist knows all of this. What makes their analysis valuable is not access to secret data. Everything they use is published by SRTR, OPTN, the CDC, the Census Bureau, and state health departments. What makes it valuable is knowing which variables matter, how they interact, and how to combine them into a recommendation.

The variables fall into two categories: supply-side (how many organs become available in a given region) and demand-side (how many patients are competing for them).

Supply: where organs come from

Organs for transplant come overwhelmingly from deceased donors. The supply of deceased donors in any region is shaped by factors that most patients never think about.

Motor vehicle fatality rates. Traffic deaths are a major source of transplantable organs, particularly kidneys and livers. Regions with higher per-capita motor vehicle fatality rates tend to have more potential organ donors. The relationship is not perfectly linear (not every traffic death results in organ donation), but it is statistically significant and well-documented.

Here is the detail that matters: the type of trauma affects which organs survive. A high-speed collision that causes massive thoracic injury may destroy the heart and lungs while leaving the kidneys and liver viable. A head injury from a lower-speed crash may preserve thoracic organs. The optimal city for a kidney transplant is not necessarily the optimal city for a heart transplant. The supply chain is organ-specific.

Donor registration rates. Not every eligible death results in organ donation. Families must consent, and in many cases, the deceased must be a registered organ donor. Donor registration rates vary dramatically by state, from under 40% in some states to over 80% in others. A region with high trauma rates but low registration rates may produce fewer actual donors than you would expect.

Demographic composition. Younger donors generally produce higher-quality organs. Regions with younger populations (measured by median age, military base presence, or university concentration) tend to have a higher proportion of donation-eligible deaths. Conversely, regions with aging populations contribute fewer usable organs per capita.

Cause-of-death profiles. The opioid epidemic changed the transplant landscape. Overdose deaths now account for a significant fraction of organ donors in some regions. These donors tend to be younger, which is favorable for organ quality, but they may carry higher infectious disease risk profiles. Some centers accept these donors routinely; others are more cautious. The geographic distribution of overdose deaths is uneven, which means it affects some OPO territories more than others.

Demand: who is competing

Supply alone does not determine your odds. What matters is the ratio of supply to demand in a given territory.

Waitlist size relative to population. Some OPOs serve dense urban populations with large academic medical centers that attract patients from wide catchment areas. These regions may have adequate donor supply in absolute terms but enormous waitlists relative to that supply. Other OPOs serve smaller, less dense populations with proportionally smaller waitlists.

Number of transplant centers competing for the same OPO's organs. When an organ becomes available, it is offered first to patients at centers within the local OPO territory. If that territory contains five transplant centers, those centers are all drawing from the same pool. If it contains one, that center has less competition for local offers.

Center acceptance rates. This is the variable the epidemiologist cares about most, because it is the least visible to patients and the most consequential for outcomes. Each transplant center independently decides which organ offers to accept for its patients. A center with an aggressive acceptance profile (accepting older donors, donors with certain risk factors, organs with longer cold ischemia times) effectively creates a shorter waitlist for its patients, because it says yes to offers that conservative centers decline.

SRTR publishes center-level acceptance data, but interpreting it requires context. A center with a high acceptance rate and good outcomes is genuinely more accessible. A center with a high acceptance rate and poor outcomes is accepting organs it probably shouldn't. The epidemiologist looks at both numbers together.

Why the best city depends on the organ

This is the insight that surprises most patients: the optimal listing location is different for different organs.

A city with high motor vehicle fatality rates and aggressive acceptance criteria might be ideal for a kidney transplant. But that same city could be mediocre for a heart transplant if the trauma patterns tend to damage thoracic organs, or if the local heart transplant program is small and conservative.

Conversely, a city with a large academic medical center that runs an aggressive heart transplant program, located in an OPO territory with favorable donor demographics, might be the best option for a heart but a poor choice for a kidney because of an enormous kidney waitlist.

The billionaire's epidemiologist does not produce a single ranking. They produce an organ-specific recommendation, because the supply chain, the demand profile, and the center-level behavior are different for each organ type.

The Monte Carlo problem

So you have 50 or more variables. You have data on each of them, but the data is noisy. Donor registration rates fluctuate year to year. Motor vehicle fatality rates are volatile in small populations. Center acceptance behavior changes when surgeons join or leave a program.

How do you combine 50 uncertain variables into a single score?

One approach is to weight them deterministically: assign each variable a weight, multiply, sum. But that assumes you know the right weights, and it gives you a point estimate with no uncertainty quantification. If City A scores 72 and City B scores 71, is that difference real or noise?

The alternative is Monte Carlo simulation. You model each variable as a distribution (not a single number), then sample from all distributions simultaneously, thousands of times. Each sample produces one simulated scenario. After 10,000 scenarios, you have a distribution of outcomes for each city, not just a point estimate.

The advantage: you can see not just which city scores highest on average, but how much the rankings fluctuate. If City A beats City B in 9,800 of 10,000 simulations, that is a robust finding. If it beats City B in only 5,200 of 10,000 simulations, the difference is noise and you should weight other factors (cost of living, proximity to family, existing medical relationships).

What the interactive shows you

Play with the sample sizes above. At N=10, the histogram is barely recognizable as a distribution. The mean jumps around. You could not reliably rank two options based on this. At N=1,000, the shape is clear but the mean still has a standard error of a few points. At N=10,000, the estimate has converged: the mean is stable, the distribution shape is well-defined, and you can make rank comparisons with confidence.

This is exactly the logic behind combining uncertain transplant variables. Each "sample" in the real system represents a scenario: one possible configuration of donor supply, waitlist dynamics, and center behavior, drawn from the observed distributions. Run enough scenarios and the ranking of cities becomes reliable even though any individual variable is noisy.

This is what I built

TransPlan is the tool that automates this analysis. It pulls public data across 8 weighted categories spanning donor supply, waitlist dynamics, geographic access, center performance, and demographic factors for 22 US cities. It runs the Monte Carlo simulation. It produces city-level composite scores with confidence intervals.

Everything TransPlan does, the billionaire's epidemiologist could do manually given two weeks and a spreadsheet. The difference is that TransPlan does it automatically, and it is free.

The data sources are entirely public: SRTR, OPTN, CDC WONDER, Census Bureau, state DMV registries, CMS. Nothing is proprietary. Nothing requires special access. The methodology is documented and open source.

This is not a clinical tool. It does not tell anyone whether to get a transplant, where to get evaluated, or whether to list at a second center. It is an analytical layer: it takes publicly available data that is scattered across dozens of sources, combines it with a statistical framework that accounts for uncertainty, and outputs a structured comparison.

The access shift

The analysis TransPlan performs has always been possible. Wealthy patients have been paying for it, in one form or another, since the transplant system was created. The epidemiologist with the spreadsheet is not doing anything exotic. They are reading public data, applying standard statistical methods, and producing a recommendation.

The barrier was never information access. The federal government publishes all of it. The barrier was expertise: knowing which variables to look at, how to combine them, and how to interpret the result. That expertise has historically been available only to people who could hire it.

Making that analysis freely available does not change the transplant system's rules. It does not create organs that did not exist before. It does not move anyone up a waitlist. What it does is remove the information asymmetry that allowed the system to be navigated more effectively by people with resources.

Whether that is a straightforwardly good thing turns out to be a harder question than it sounds.

---

Next in this series: More People Benefit. Is That Enough?. Two worlds, two definitions of fairness, and the question that has no clean answer.