The Repurposing Study Planner
Take a repurposing idea, an existing drug proposed for a new use, and work out what could be answered with real world data, which design would answer it, what would undermine the answer, who would have to run it, and what result would change anything. Built around the way an OHDSI federated network study is planned and run.
How to use this tool
Repurposing questions are observational questions with a particular shape: the exposure is already in use, often off label, the people who receive it differ from the people who do not in ways that predict the outcome, and the sponsor who would normally fund the evidence may not exist. This planner works through that shape stage by stage and assembles a plan you can hand to collaborators.
- Name the candidate. The drug, its approved use, the proposed use, and where the idea came from. Signal provenance changes what kind of evidence comes next.
- Frame the question. Population, exposure, comparator, outcome, and time at risk, written as the trial you would run if you could.
- Choose the design. Answer a short set of questions about the exposure and the outcome, and the planner suggests a design and the matching analysis package, with its caveats.
- Work through the threats. Each threat comes with what it does to a repurposing estimate and the usual safeguards. Record which safeguard you plan to use.
- List the data and the partners. Which data sets could hold the exposure, whether they are in the OMOP common data model, and who to ask.
- Plan the execution. The steps a federated study runs through, with an owner and a target date on each.
- Set the evidence bar. Who has to be convinced, what result would convince them, and where the result goes next.
- Assemble the plan. A readiness check-up, then the plan as a document, as Markdown, or as a save file you can reload.
Everything stays in this browser. The plan autosaves here, and the save file is the way to move it to another machine or send it to someone else.
Name the candidate
Start with the drug and the use being proposed for it. Where the idea came from is part of the record, because a signal from a knowledge graph and a signal from twenty clinicians who have already tried it call for different next steps.
Before you go further
Check whether the question has already been asked. The literature map covers real world evidence studies by disease area and data source, and the catalog lists the data sources, registries, and programs that hold repurposing-relevant data.
Real World Evidence Literature Map The Repurposing Catalog CURE IDFrame the question
Write the trial you would run if you could, then the observational study that emulates it. Being specific here is what makes the design choice in the next stage possible, and it is the part that transfers directly into a protocol.
Turning this into a formal question
The use case generator puts a question into the standard OHDSI form, traces every part of it to standard concepts, writes the site queries that count who is in the population, and gives a feasibility read. Take the population, exposure, comparator, and outcome from here into it, then come back with the counts.
The Analytic Use Case GeneratorChoose the design
Answer these about the exposure and the outcome as they appear in data, not as they would be in a trial. The suggestion below follows from your answers and is a starting point for a conversation with a methodologist, not a substitute for one.
Where these designs are described
- The Book of OHDSI, population-level estimation, for the new-user cohort design and the self-controlled designs.
- The Book of OHDSI, method validity, for negative controls and empirical calibration.
- HADES, the analytics packages each design runs on, and Strategus, which assembles them into one study package.
- Target trial emulation, for writing the protocol as the trial you would have run.
Work through the threats
Each of these can turn a repurposing question into a misleading answer, and most of them have a standard safeguard. Open one, read what it does to an estimate like yours, choose how you plan to handle it, and note anything specific to your candidate.
A threat left unaddressed is a finding, not a failure. Reviewers ask about these directly, so a plan that names the one it cannot solve reads better than a plan that is silent about it.
List the data and the partners
Off-label use is recorded unevenly, so the data sets that hold your exposure are not always the ones you would reach for first. List the candidates, note whether each is already in the OMOP common data model, and record who to ask.
Finding candidate sources
The catalog lists real world data sources, registries, knowledge bases, and regulatory programs by disease area. The ecosystem planner is the one to use when the data you need do not exist yet and someone has to build them.
The Repurposing Catalog The Real World Evidence Ecosystem Planner The OMOP ETL WorksheetPlan the execution
A federated study keeps the data at each site and sends the analysis to them, so most of the work happens before anyone runs anything. Give each step an owner and a target date. Steps without an owner are the ones that slip.
The mechanics, step by step
Network Study Paint By Numbers walks through concept sets, cohort definitions, vocabulary selection, exposure windows, and the feasibility exercises, with a published study worked backwards alongside your own.
Network Study Paint By Numbers The Book of OHDSI, network research Published study packagesSet the evidence bar
Repurposing evidence has to travel somewhere to do any good, and different destinations accept different things. Decide now who has to be convinced, because it determines how the study is designed and how much of it has to be pre-specified.
The money and the value case
The HEOR workbench covers cost-effectiveness, budget impact, and the numbers a payer asks for. The first-time funder guide is for patient organizations that would have to fund this themselves.
The HEOR Workbench The First-Time Funder CURE Drug Repurposing CollaboratoryAssemble the plan
The readiness check-up beside this reads what you have entered and reports where the plan is thin. It is a check of the plan, not of your data or your candidate.
Take it with you
The document is the version to circulate. Markdown is for pasting into a repository or a protocol draft. The save file reloads everything here, including the parts still unfinished.
The study in one sentence
Plan so far
Readiness check-up
This planner is a structured way to think through a repurposing study and to write down what you decide. It is not a protocol, a statistical analysis plan, or regulatory advice, and the design suggestion follows from your own answers rather than from any review of your data. Off-label prescribing, data use agreements, institutional review board (IRB) approval, and any regulatory path all have requirements that differ by country and institution, so take each of those to the people who hold them. Where the planner describes what many programs do, treat it as a starting point to check rather than as a rule.