Listen Get

Universal Right to Try with Evidence: Potential Impact of Adoption in All 50 States

Keywords

war-on-disease, 1-percent-treaty, medical-research, public-health, peace-dividend, decentralized-trials, dfda, dih, victory-bonds, health-economics, cost-benefit-analysis, clinical-trials, drug-development, regulatory-reform, military-spending, peace-economics, decentralized-governance, wishocracy, blockchain-governance, impact-investing

Please help us improve this impact analysis. We want to help policymakers and the public understand the enormous potential benefits of universal patient access to clinical trials and a continually learning system for global treatment-effectiveness rankings and outcome labels. Comment below, or highlight any text and log in with Hypothesis to leave a note. Co-authors welcome.

The short version

About 6,650 diseases (95% CI: 5,700 diseases-8,232 diseases) rare diseases still have no effective treatment. The model uses this queue as a proxy for the wider therapeutic frontier. At the current rate, roughly 15 diseases receive their first effective treatment each year. At that pace, exploring the queue takes approximately 443 years (95% CI: 255 years-841 years), and the average therapeutic target waits about 222 years (95% CI: 128 years-420 years).

The problem is not a shortage of possible treatments. Many post-Phase-1 compounds, repurposed drugs, combinations, doses, and disease-specific uses are never tested because nobody can finance conventional Phase 2 and Phase 3 trials.

Universal Right to Try136 with Evidence changes that financing constraint. Licensed centers provide eligible experimental treatments while patients or payers cover treatment-related care, trial-site services, and permitted study costs. Trial participants follow shared protocols, and other eligible uses contribute separately labeled observational outcomes. Treatments that were not economically viable to investigate become testable. State law does not waive federal authorization or evidence requirements.

If all 50 states adopt the framework and it produces the modeled increase in treatment discovery, the conditional central estimates are:

Result Conditional central estimate
Treatment-discovery multiplier 5.48x
Average first treatment arrives 181 years earlier
Premature deaths prevented

9.19 billion
Healthy years restored 483 billion DALYs
Disability-equivalent suffering prevented 1.65 quadrillion hours
Campaign plus shared registry cost

$65 million
Philanthropic cost per healthy year restored

$0.000134

These are conditional lifetime totals across future generations. They do not mean that billions of currently living people are saved, and they do not assume that passing a law automatically produces the modeled scientific effect.

Monte Carlo Distribution: Lives Saved from Universal Right to Try with Evidence (10,000 simulations)

Monte Carlo Distribution: Lives Saved from Universal Right to Try with Evidence (10,000 simulations)

Simulation Results Summary: Lives Saved from Universal Right to Try with Evidence

Statistic Value
Baseline (deterministic) 9.19 billion
Mean (expected value) 9.4 billion
Median (50th percentile) 8.82 billion
Standard Deviation 4.13 billion
90% Range (5th-95th percentile) [3.71 billion, 17.2 billion]

The histogram shows the distribution of Lives Saved from Universal Right to Try with Evidence across 10,000 Monte Carlo simulations. The CDF (right) shows the probability of the outcome exceeding any given value, which is useful for risk assessment.

What the legislation changes

Montana has already created licensed experimental treatment centers. Senate Bill 535 covers treatments that have completed Phase 1 but are not approved for general FDA use. It permits centers to establish payment arrangements, charges a $10,000 license application fee and $5,000 annual renewal fee, and directs 2% of annual net profits toward access for qualifying Montana residents137.

That creates a lawful state-regulated place to provide experimental treatment. It does not by itself create a clinical trial system. Montana’s rules require centers to track treatment, patient outcomes, and serious adverse events, but the rules explicitly state that the centers do not administer clinical trials138. Observational outcomes can reveal promising or dangerous signals. Without a credible comparison, they cannot reliably show that a treatment caused the result.

The model bill adds four requirements:

  1. Enroll. Register a prospective protocol, obtain informed consent, report serious safety events, and place each patient in the applicable federal trial or treatment-access pathway.
  2. Compare. Prespecify outcomes and use simple randomization or another adequate control when feasible. Without an adequate comparison, label the result a signal, not a validated treatment effect.
  3. Pool. Send standardized, de-identified baseline and outcome data to a shared registry.
  4. Publish. Continuously publish results for each treatment-condition pair, including negative and harmful results.

The result is access with evidence. Participating patients receive treatments that sponsors make available and contribute to a common learning system. Every result can improve the next treatment decision.

Why treatments that cannot attract conventional funding matter

A candidate treatment can have completed initial human safety testing and still be commercially impossible to develop. A conventional Phase 3 trial costs about $41,000 (95% CI: $20,000-$120,000) per participant. The existing model estimates that an embedded pragmatic trial can reduce that to approximately $929 (95% CI: $97-$3,000) per participant, a 44.1x (95% CI: 12.8x-210x) reduction.

That does not mean every trial becomes 44.1x (95% CI: 12.8x-210x) cheaper. Complex and high-risk studies will still need specialized sites, procedures, and monitoring. It means routine care can often replace dedicated research visits and duplicate data collection while retaining prespecified outcomes, safety reporting, reliable analysis, and a comparison group.

The candidate pool is much larger than a count of compounds. One compound may be tested for several diseases, doses, combinations, or subtypes. Additional candidates do not multiply disease burden. They increase the chance that the first effective treatment for each condition is found sooner.

One policy-effect input instead of a pretend forecast

Nobody knows how many centers will open, how many patients will enroll, how many shelved treatments will return, or what fraction of well-designed evaluations will find an effective treatment. Modeling each unknown as a separate multiplier would manufacture precision and create opportunities to count the same benefit twice.

The model compresses the entire policy mechanism into one uncertain input: a 5.48-fold treatment-discovery multiplier. It represents:

  • patient or payer funding of treatment and evidence generation;
  • treatments revived because conventional trials were not economically viable;
  • enough candidate-condition pairs and eligible participants to sustain evaluation;
  • protocols capable of producing credible efficacy evidence; and
  • the share of completed evaluations that find a first effective treatment.

The central value is a calibration assumption, not an observed causal estimate. The automatic Monte Carlo samples uncertainty in this multiplier, launch cost, avoidable disease burden, and their upstream inputs. It does not capture political adoption risk or model-form uncertainty in using the rare-disease queue as a proxy for the wider therapeutic frontier. The physical health totals are undiscounted. Applying a time discount rate would reduce their present value, but adding a preferred discount rate would not change the underlying schedule shift.

The calculation, one step at a time

The arithmetic asks how much sooner the average first treatment arrives if the discovery rate increases:

Step Question Conditional central result
1 How long would the remaining treatment queue take at the current discovery rate?

443 years (95% CI: 255 years-841 years)
2 How long does the average therapeutic target wait?

222 years (95% CI: 128 years-420 years)
3 How much earlier does treatment arrive at the modeled discovery rate?

181 years
4 How many global healthy years does that schedule shift restore?

483 billion DALYs
5 What is the campaign and registry-launch cost per restored healthy year?

$0.000134

The full calculation is:

\[ \begin{gathered} Cost_{RTT,DALY} \\ = \frac{C_{RTT}}{DALYs_{RTT}} \\ = \frac{\$65M}{483B} \\ = \$0.000134 \end{gathered} \]
where:
\[ \begin{gathered} DALYs_{RTT} \\ = DALYs_{global,ann} \times Pct_{avoid,DALY} \times T_{accel,RTT} \\ = 2.88B \times 92.6\% \times 181 \\ = 483B \end{gathered} \]
where:
\[ \begin{gathered} T_{accel,RTT} \\ = T_{first,SQ} \times \left(1 - \frac{1}{k_{RTT}}\right) \\ = 222 \times \left(1 - \frac{1}{5.48}\right) \\ = 181 \end{gathered} \]
where:
\[ \begin{gathered} T_{first,SQ} \\ = T_{queue,SQ} \times 0.5 \\ = 443 \times 0.5 \\ = 222 \end{gathered} \]
where:
\[ \begin{gathered} T_{queue,SQ} \\ = \frac{N_{untreated}}{Treatments_{new,ann}} \\ = \frac{6{,}650}{15} \\ = 443 \end{gathered} \]
where:
\[ \begin{gathered} N_{untreated} \\ = N_{rare} \times 0.95 \\ = 7{,}000 \times 0.95 \\ = 6{,}650 \end{gathered} \]

The rare-disease treatment queue is the clearest available count of diseases still awaiting a first effective treatment. The model uses it as a proxy for the pace of exploring the wider therapeutic frontier, then applies the resulting schedule shift to eventually avoidable global disease and aging-related burden. That is the model’s strongest assumption. It is stated directly because it drives the scale of the result.

What the headline numbers mean

The 9.19 billion deaths estimate exceeds the current world population because it sums premature deaths prevented across approximately 181 years of future generations. It is a schedule-shift estimate, not a count of currently identifiable people.

The 1.65 quadrillion hours estimate means disability-weighted equivalent hours. A disability weight of 0.25 sustained for four hours equals one full-disability-equivalent hour. It does not mean that every hour represents maximum conscious pain.

The global estimate assumes discoveries become public knowledge and eventually benefit patients outside the United States. The access right itself remains state law.

The model does not count the uncertain direct health effect on participating patients. Earlier access to an effective treatment could help them immediately. Ineffective or unsafe treatments could harm them. Those effects are real, but they are a separate channel from the treatment-discovery schedule shift and are not included in the headline totals.

What philanthropy and investment pay for

Philanthropy funds the shared public good

Philanthropy pays for the 50-state campaign, common evidence standards, and the shared registry’s first ten years. Participating centers then sustain the registry through license or enrollment assessments. The model does not require a new state appropriation.

The conditional philanthropic cost per DALY divides that public-good investment by the health benefit if all 50 states adopt and the mature system produces the modeled discovery rate. It excludes patient and payer spending on treatment delivery because those payments buy care and evidence generation rather than consuming the philanthropic campaign budget.

At the central estimate, the modeled philanthropic cost per life saved is $0.00707, which is 636.2k times lower than the midpoint of GiveWell’s cited modeled range across top charities7. The comparison is so large because a relatively small campaign and registry are assumed to unlock treatment, sponsor, payer, clinic, and investor spending that philanthropists do not supply.

The cost-per-DALY Monte Carlo shows uncertainty in the conditional estimate itself:

Monte Carlo Distribution: Universal Right to Try with Evidence Philanthropic Cost per DALY (10,000 simulations)

Monte Carlo Distribution: Universal Right to Try with Evidence Philanthropic Cost per DALY (10,000 simulations)

Simulation Results Summary: Universal Right to Try with Evidence Philanthropic Cost per DALY

Statistic Value
Baseline (deterministic) $0.000134
Mean (expected value) $0.000171
Median (50th percentile) $0.000119
Standard Deviation $0.000192
90% Range (5th-95th percentile) [$4.01e-05, $0.000448]

The histogram shows the distribution of Universal Right to Try with Evidence Philanthropic Cost per DALY across 10,000 Monte Carlo simulations. The CDF (right) shows the probability of the outcome exceeding any given value, which is useful for risk assessment.

For a grant decision, multiply the conditional benefit by the donor’s own probability that the grant produces full adoption, mature implementation, and the modeled scientific effect. Equivalently, divide the conditional cost per life saved by that probability. Keeping that judgment outside the model avoids another arbitrary scenario parameter.

Commercial capital funds the parts that can earn revenue

Review, compliance, clinical administration, manufacturing, monitoring, data infrastructure, and successful treatment assets can all produce revenue. Commercial capital can therefore finance:

  • experimental treatment review board services and compliance;
  • licensed treatment and pragmatic-trial sites;
  • manufacturing and distribution;
  • treatment sponsors and holders of shelved post-Phase-1 assets;
  • interoperable registry, analytics, and monitoring infrastructure; and
  • payer contracts that condition coverage on evidence production.

Donors fund the public-good layer that no company can capture. Investors fund services and assets that can earn revenue. Neither group has to finance the entire system.

The flywheel maps possible cash flows, not promised investor returns. Actual returns depend on lawful charging, treatment demand, reimbursement, liability, protocol quality, and whether the evidence changes clinical or regulatory decisions.

The federal boundary

State legislation does not approve drugs, waive federal trial authorization, or lower FDA’s evidence standard. Federal Right to Try is limited to eligible patients who are unable to participate in a clinical trial involving the drug139. A patient enrolled in an interventional pragmatic trial must instead participate under an active investigational new drug application or another applicable FDA-authorized pathway. Patients who cannot join a trial may contribute separately labeled observational outcomes through federal Right to Try or expanded access if eligible.

FDA’s standard remains substantial evidence from adequate and well-controlled investigations. FDA guidance allows one rigorous investigation plus confirmatory evidence in some circumstances, while still requiring sufficient safety evidence and a favorable benefit-risk assessment140. FDA also supports streamlined randomized trials integrated into routine clinical practice141. Project Pragmatica applies this design logic to approved oncology products through simpler eligibility and routine-practice enrollment142. These programs show that pragmatic design can reduce collection burden. They do not authorize states to bypass federal law.

Federal charging rules also remain in force. An IND sponsor needs FDA authorization to charge for an investigational drug and generally may recover only direct drug costs. Trial sites may separately recover pharmacy, nursing, equipment, and study-related procedure costs without authorization under that charging rule143. The model therefore assumes that patients or payers cover treatment delivery, site services, and permitted study costs. Whether lawful revenue is sufficient to sustain the modeled discovery multiplier remains uncertain.

Medicare already uses the analogous payment principle in Coverage with Evidence Development. Specified services may be covered within an approved study while evidence is collected, then CMS can reconsider coverage using the results144. Universal Right to Try with Evidence applies the same basic idea outside Medicare: access now, standardized evidence from every use, and stronger decisions over time. The Continuous Evidence Generation Protocol145 provides one compatible technical design. The legislation does not need to mandate that specific protocol.

Montana can prove it works

Montana has already completed the politically difficult first step. It created the licensed-center and independent-review framework. Infinita describes a coordinated pathway involving experimental treatment review board review, licensed clinical administration, monitoring, evidence generation, and early commercialization for post-Phase-1 therapies146.

The next step is a transparent Montana demonstration that adds interoperable enrollment, credible comparisons, pooled outcomes, documented fees, and safety monitoring. If that system produces useful evidence while remaining commercially sustainable, other states have a model they can copy.

The original Right to Try model spread from its first state in 2014 to 41 states by 2018147. That history establishes a distribution channel, not a guarantee. The evidence amendment must prove its value in operation.

Any implementing bill should require Enroll, Compare, Pool, and Publish. That turns 50 possible access experiments into one continually learning treatment system.