Stop Moonlighting: Why Expertise In IRT Is Non-Negotiable
Clinical trials are among the most consequential scientific endeavors in medicine. Every role within them — from drug discovery to statistical analysis to regulatory compliance — demands deep, specialized expertise. Yet one critical function is often treated as an afterthought, handed off to whoever happens to be available: Interactive Response Technology, or IRT.
This needs to change.
What Does "Moonlighting" in IRT Actually Mean?
Think for a moment about your own professional expertise. How long did it take you to develop the skills you rely on daily? How specialized are they? Could someone without your background simply step in and perform your role at the same level?
Most professionals would answer confidently: no. And yet, organizations routinely assign IRT responsibilities to people whose primary expertise lies elsewhere — statisticians, clinical operations staff, or project managers who layer IRT oversight onto already full plates. This is moonlighting: doing a second job on the side, without the full focus or foundation the role demands.
The question worth asking is a simple one: why would IRT be any different from the other highly specialized disciplines that make a clinical trial succeed?
IRT Is Serious Business
The stakes of getting IRT wrong are not abstract. They are study-defining.
Participant randomization sits at the heart of a trial's scientific validity. If randomization is implemented incorrectly, the integrity of the entire study can be called into question. This requires not just technical execution, but a deep understanding of randomization methodology and its downstream implications.
Study drug assignment carries real-world consequences for patients. Incorrect assignments can expose participants to the wrong treatment — a patient safety issue, not just a data quality one. Add in the complexity of crossover designs and titration schemes, and the margin for error shrinks further.
Supply management is often where expertise gaps show up most visibly. Getting the right drug to the right site at the right time requires accurate demand forecasting, logistics coordination, and trust in the system's algorithms — none of which are intuitive without experience.
Blind protection is perhaps the most underappreciated area of IRT expertise. Both study design choices and drug assignment logic can create inadvertent opportunities for unblinding. Without experienced eyes reviewing these decisions, the integrity of the blind can be compromised before a single participant is enrolled.
What Only Experience Can Surface
IRT expertise isn't just about avoiding catastrophic errors — it's about making consistently better decisions throughout the life of a study. There is a long list of nuanced, consequential topics that only become visible through deep experience in the field. Here are just a few examples:
Understanding why randomization block size and allocation ratio can directly affect drug supply requirements
Recognizing the distinct implications of rescreening versus re-randomization versus randomization replacement
Navigating the tension between when a randomization list must never be changed and the rare circumstances when it must be
Distinguishing between visit dates and transaction dates, and why conflating them creates data problems
Appreciating the difference between drug assignment, dispensation, and actual dosing — and why each matters independently
Identifying where unblinding risk hides in seemingly routine design and assignment decisions
Applying waste reduction strategies like medication pooling, fractional dosing, and automated supply switching effectively
These are not topics that surface in a project management handbook. They emerge from years of working across complex, varied, real-world study designs.
The Case for Dedicated IRT Expertise
Beyond engaging experienced external partners, organizations running clinical trials should seriously consider investing in dedicated IRT expertise. The reasons are both practical and strategic.
A dedicated IRT expert pays for themselves. The cost is modest compared to the downstream expenses of protocol amendments, supply waste, study delays, or a compromised randomization. They create organizational standards that drive consistency across a portfolio of trials. They serve as an independent check on vendor recommendations, providing validation rather than passive acceptance. And critically, they are more accessible to internal stakeholders than external vendors — able to manage expectations, answer questions, and resolve issues in real time.
For larger organizations with active trial portfolios, a full-time internal resource makes clear sense. But for smaller biotechs and emerging sponsors — where headcount is lean and a single IRT specialist may not be justifiable — a fractional consultant can deliver the same strategic value without the overhead of a permanent hire. A seasoned fractional IRT expert brings deep, cross-industry experience, can be engaged at the level each study demands, and gives smaller organizations access to expertise that would otherwise only be available to larger players.
Either way, the goal is the same: having a trusted, experienced voice in your corner that is genuinely working in your organization's interest — not just executing a vendor's preferred approach.
The Expertise Imperative
Every discipline in clinical research that touches patient safety and scientific validity is treated as a specialist domain. Drug discovery, CMC, statistical analysis, regulatory affairs — none of these are handed to capable generalists and hoped for the best.
IRT deserves the same respect.
The complexity of modern clinical trials — adaptive designs, oncology protocols with visit and cycle distinctions, global supply chains, blinding requirements — demands people who have genuinely seen the full range of challenges this domain presents. Not people who learned IRT while doing something else.
The trials we run today will determine the treatments available to patients tomorrow. That is not a context in which expertise should be optional.