Reasons given include regulatory uncertainty, site resistance, sponsor caution that veers into risk aversion, and ill-adapted feasibility scoring, but none of those account for the full picture.
Here’s the question I’ve sat with since coming to Lightship: if flexible delivery is the superior way to run a trial, why hasn’t the rest of the industry made it standard practice?
New evidence for hybrid and virtual trials arrives every year: better completion rates, broader reach, stronger follow-through to the end of a study. And still, half of clinical trials miss their enrollment targets, and nearly a third of the participants who do enroll drop out before completion.
A review in the Journal of Clinical and Translational Science examined 13 studies of hybrid and virtual (i.e., flexible) methods and found the same pattern: better recruitment in 11 of them, better retention in seven, broader participant representation in six. Additionally, a recent analysis from Tufts CSDD found that a direct-to-participant site model had an 86% completion rate, versus a 36% benchmark for typical sites, and a 20 to 30% share of total enrollment from that single site. A separate randomized trial comparing a digital care program to standard in-person care found the same gap, 81% completion versus 64%. Third-party validation like this matters because it confirms what I know: flexible delivery performs. And it performs better than in-clinic-only delivery.
The Risk-Based Monitoring Precedent
This isn’t the first time an evidence-backed change has stalled at the green light: risk-based monitoring (RBM) offers a clean parallel to the flexible trial adoption lag. The FDA finalized guidance backing RBM in 2013, but years of encouragement barely moved adoption past 53% by 2019. And the pieces requiring real trust, like reduced source verification, stayed stuck in the high teens.
Evidence and regulation moved the easy stuff. It took a pandemic to move the rest, and trust still didn’t budge. Keep that in mind because the four reasons usually given for slow flexible model adoption are versions of the same story.
It’s Change Management. Every Time.
Four reasons get the blame: regulatory uncertainty, site resistance, sponsor caution, and feasibility scoring ill-adapted to new models. Peel back any one of them, and they’re all wearing the same disguise: change management, the work of getting an organization to adopt something new even after the proof already exists.
The first explanation is regulatory uncertainty. Sponsors have pointed to unclear standards for remote consent, drug handling across settings, and data integrity as reasons to hold back. But the FDA closed that gap when it finalized guidance on decentralized trial elements in 2024, giving sponsors a clear path for data quality. If regulatory clarity were the missing piece, RBM would have moved faster, too, because guidance for that existed for years before adoption moved at all. Regulatory clarity already exists: the real gap is change management.
Site resistance? Sounds right, until you look closer. True, when trial activities move to outside vendors like separate nursing agencies and disparate telehealth platforms, investigators have historically been slow to get on board. They understandably don’t trust oversight they can’t see.
Models built as a single team, like Lightship’s, however, change that equation. It’s the same investigators and coordinators, whether the visit happens in a clinic or a living room. But most study teams have only ever seen the fragmented, multi-vendor version of flexible delivery, and they size up every model against that template, bracing for complexity that isn’t there. Call it what it is: change management, driven by fear instead of fact.
What about sponsor caution? Sponsors already treat the failure to achieve representative enrollment as a risk they can’t ignore. A drug that wasn’t adequately tested across the populations who will actually use it becomes a safety and reputational problem the moment it reaches the market. Sponsors plan around that risk long before a product gets anywhere near approval by ensuring adequate representation in recruitment and enrollment. What they haven’t done yet is apply that same urgency to how a trial gets delivered.
Sponsors treat representative enrollment as non-negotiable. They don’t yet treat flexible delivery, one of the most direct levers for reaching the populations traditional sites miss, the same way. Research in Science makes this case directly: site-based requirements exclude people without reliable transportation or a nearby site. The connection between flexible delivery and representative enrollment hasn’t been made explicit, even though the research is there. Draw the line, and the risk calculus shifts the way it already has for representation: from nice-to-have to an absolute necessity. Sponsors are using an old checklist. Why? Change management, unaddressed.
Last but certainly not least, there’s feasibility. And it’s broken for every model, not just new ones. Phase III protocols now carry more than 50 eligibility criteria. Sites predict enrollment against criteria they can’t fully assess, on timelines sponsors rarely make workable. Nobody checks the prediction against what actually happened. 76% of protocols need an amendment after clearing feasibility anyway, a quarter of them avoidable from the start.
There’s an old definition of insanity: doing the same thing over and over and expecting a different result. Study feasibility is a prime example. It runs on historical metrics, participants enrolled per site per month, an investigator’s track record, rewarding a long history and screening out everything else, no matter how a new model actually performs. That’s a misguided idea of what drives study success, propped up by change management nobody’s bothered to fix.
Four reasons. One pattern: change management, dressed up as something new each time.
Where This Leaves Us
Every explanation above is the same problem in a different disguise: the industry is avoiding change.
RBM showed us that regulation and evidence move the easy stuff, but it takes a shock to move the rest, and even the shock doesn’t reach everything. Flexible trials got their own version of that shock. The pandemic forced overnight adoption, and some of that early enthusiasm oversold what the model could actually deliver. The retreat that followed made sense at the time, but the evidence has only gotten stronger since.
It’s strong enough that companies like Lightship have built their entire model around it. Broader adoption hasn’t followed at the same pace, which is a gap worth closing. Sponsors don’t have to wait for the industry to sort this out. The connection between flexible delivery and representative enrollment belongs in the next feasibility scorecard a sponsor writes.
I opened this blog with a question, and here’s the answer: the industry already has the evidence it needs. Acting on it just takes will. If yours is ready, let’s talk.