When a trial's enrollment stalls, the fastest path back to a healthy accrual curve is a 7 to 10 day diagnostic, followed by site triage (stop, rescue, or scale) and short-cycle recruitment actions launched within the first week. These three moves work because they address the three real causes of enrollment decline at once: weak site capacity, excessive participant burden, and slow outreach. Ownership sits with a senior operations lead, who should deliver a diagnostic brief with a site triage list within 48 to 72 hours.
TL;DR:
- Compare weekly site velocity, screen failure reasons, activation lag, geographic fit, and lead response times to separate capacity problems from eligibility barriers.
- Stop sites with structural barriers, but give recoverable sites dedicated staffing or operational remediation that can be deployed within two weeks.
- Move sites below 50% of projected velocity to STOP after two remediation weeks; sites above 120% of target for three weeks can absorb displaced volume.
- Online campaigns can cost less per enrollee than offline recruitment, about $72 versus $199 in one comparison, but only with immediate lead follow up.
- Pilot telehealth or home collection at one or two sites, documenting protocol changes and monitoring and training plans to protect data quality and safety reporting.
Table of Contents
- Run a rapid enrollment diagnostic in 7 to 10 days
- Rank interventions by speed and evidence, not by instinct
- Turn interventions into a 30 to 90 day rescue timeline
- Track the right KPIs and keep the rescue governed
- How an embedded diagnostic partner accelerates a rescue
- What most rescue advice gets backward
- Get a diagnostic on your enrollment timeline
- FAQ
- Sources
Run a rapid enrollment diagnostic in 7 to 10 days
Before changing anything, find out exactly where the funnel breaks. Pull the data that shows the real picture, not the summary dashboard a site coordinator last updated three weeks ago.
- Screening logs and the full screen-to-enroll funnel, broken out by site
- Site activation dates versus planned activation dates
- Staffing rosters and current CRA or coordinator workload per site
- Site-level performance on prior studies, where available
- The competing-trial landscape in each enrollment geography
With that data in hand, run four analyses fast:
- Weekly enrollment velocity per active site, compared against the protocol's planned rate
- Screen-failure reason mapping, to separate eligibility problems from operational ones
- Geographic coverage versus the target population's actual distribution
- Time-to-first-contact for every inbound lead, since delayed callbacks quietly kill conversion
These analyses map directly to AQuIP's approach, which uses frequent, multi-scale accrual monitoring and a recruitment journal to flag site- and study-level problems early enough to act on them. Once the data is in, triage every site into one of three categories: STOP (suspend new enrollment at sites with structural problems, such as eligibility misalignment or loss of a referring clinician), RESCUE (sites that can reach acceptable velocity with dedicated resources, additional training, or an operational audit within two weeks), or SCALE (sites already outperforming target that can absorb more volume or budget).
The minimal rescue team is small on purpose: a senior operations lead running point, a data analyst, a senior CRA, a regulatory liaison, and a PI or medical lead who can approve protocol-adjacent changes quickly. Our own breakdown of where AI can close the gap between protocol design and recruitment covers how this diagnostic can be compressed further when the data pipeline is already instrumented.
Pro Tip: Pull activation-lag data before anything else. A site that opened 60 days late almost never catches up on its own, no matter how good its recruitment messaging is.
Rank interventions by speed and evidence, not by instinct
Once sites are triaged, match the result to interventions ranked by how fast they move the needle and how much evidence supports them. Start with fast wins that can launch within days:
- Telephone reminders to non-responders already in the funnel
- Targeted SMS messages with clear, specific participant information
- Short, intensive online ad bursts timed to available scheduling capacity
- Bulk clinician messaging through EHR portals to surface eligible referrals
A systematic review of recruitment interventions found telephone reminders to non-responders increase recruitment rates significantly, with SMS and financial incentives showing smaller but still positive effects depending on context. Online recruitment adds a second lever: a JMIR meta-analysis found online campaigns can be faster and cheaper per enrollee than offline recruitment, in one comparison roughly $72 versus $199, but only when leads get immediate follow-up. Run online bursts as short, intense windows tied to real scheduling slots, not as an always-on spend.
For RESCUE sites, pair fast wins with structural fixes: a dedicated rescue CRA or senior monitor, scheduling and transportation support for participants, satellite or mobile visit options, and local HCP partnerships that widen the referral base. Where the protocol allows it, add decentralized elements like telehealth visits or home sample collection to cut participant burden. FDA guidance on decentralized clinical trials requires that any such element be described in protocol documentation with a monitoring and training plan that protects data quality and safety reporting, so pilot these changes with one or two sites before rolling them out broadly.
Centralize the operational layer too: one point of contact for participants, a single scheduling queue instead of per-site booking, and a fast referral pathway from community clinicians. Any eligibility adjustment or financial incentive needs IRB and regulatory sign-off before launch, with the rationale documented and its effect on data quality tracked from day one.
Pro Tip: Loosening eligibility criteria can widen the funnel, but only pursue it alongside your IRB and only where it leaves primary endpoints intact.
Turn interventions into a 30 to 90 day rescue timeline
A rescue plan without dates is just a wish list. Break execution into three phases with clear owners and decision points.
- Days 0 to 7: Close out the diagnostic, finalize site triage, launch fast-win campaigns (telephone, SMS, online bursts), and route any protocol-adjacent change through regulatory and IRB review. Owner: rescue lead, with the data analyst producing the triage brief.
- Weeks 2 to 4: Begin remediation at RESCUE sites or start onboarding replacement sites for STOP sites. Pilot decentralized elements in a small subset if the protocol supports it. Scale whichever outreach channel is converting fastest.
- Months 2 to 3: Expand the channels that proved their worth, finalize any staffing changes, and compare cost per enrollee and enrollment velocity against the original baseline before committing to longer-term investment.
Set decision gates in advance, not in the moment:
- A site hitting below 50% of projected velocity after two weeks of remediation moves from RESCUE to STOP
- A channel with cost per enrollee materially above the campaign average gets cut at the 30-day mark
- A site exceeding 120% of target for three consecutive weeks becomes a candidate to absorb volume from a STOP site
Our site readiness checklist is a useful reference for compressing activation timelines during the week 2 to 4 push.
Track the right KPIs and keep the rescue governed
A rescue effort without daily visibility drifts back into the same problems that caused the decline. Track a small set of metrics at the right cadence instead of drowning the team in dashboards.
| Metric | Cadence | Use |
|---|---|---|
| Daily enrollment count | Daily | Operational check, catch stalls immediately |
| Weekly enrollment velocity per active site | Weekly | Tactical, compare against triage targets |
| Screen-to-enroll ratio | Weekly | Flags funnel leakage versus eligibility issues |
| Time-to-first-contact | Weekly | Surfaces slow lead handling before it compounds |
| Site activation lag | Monthly | Strategic input for future site selection |
| Cost per enrolled participant | Monthly | ROI check on channel spend |
Our site performance analytics piece lays out a fuller KPI set for teams building this out as a permanent capability. Once the rescue stabilizes, fold what worked into standing SOPs and feasibility checklists so the next protocol starts with better site selection assumptions.
How an embedded diagnostic partner accelerates a rescue
We built our approach around backtracking from a client's strategic goal, usually first-patient-in or database lock by a specific date, to find exactly where the operational bottleneck sits. Rather than handing over a generic playbook, we embed with the team and work the diagnostic alongside clinical operations and regulatory staff.
- Our AI Velocity Diagnostic is built to compress the data-pulling and funnel analysis described above into a faster first week
- Typical deliverables include a diagnostic brief, a prioritized site triage list, a pilot plan for the fastest-moving interventions, and a governance handoff so the client team owns the dashboards going forward
- A short pilot engagement typically starts with a scoped diagnostic window, followed by a defined handoff period where client staff take direct ownership of the new workflows
Our overview of where AI can improve trial planning covers the broader thinking behind this model.
What most rescue advice gets backward
Most enrollment rescue advice treats recruitment tactics, better ads, more reminder calls, a flashier referral campaign, as the fix. They rarely are. The practitioner literature on rescue efforts points at a more uncomfortable cause: junior staff get handed the rescue, and the organization keeps leaning on one or two strong-enrolling sites instead of fixing the structural gaps underneath everyone else.
My view, based on the evidence here, is that the diagnostic is the whole game. Teams that skip it and jump straight to tactics usually see a short-lived bump and then a return to decline, because they never addressed whether the problem was site capacity, participant burden, or genuine eligibility scarcity. The fix that actually holds is senior-led, data-driven, and treats the rescue as a one-time investment in permanent operational capability, not a one-off scramble. If you only have the bandwidth to do one thing well, do the diagnostic properly and let it tell you which lever to pull first. Everything downstream gets easier once you know that.
— John
Get a diagnostic on your enrollment timeline
We work alongside clinical operations staff to build the diagnostic, the triage list, and the pilot plan rather than handing over another dashboard to maintain.
A typical engagement includes:
- A rapid diagnostic mapped to your current accrual data and site performance history
- A prioritized remediation plan with realistic timelines and cost trade-offs
- A governance handoff so your team owns the workflow once the engagement ends
Reach out through our services page to scope a diagnostic against your current enrollment timeline.
FAQ
What is an enrollment rescue plan in clinical trials?
An enrollment rescue plan is a structured, time-boxed effort to diagnose why participant recruitment has stalled and to fix it through site triage and targeted interventions. It typically combines a rapid data diagnostic with prioritized actions like outreach campaigns, site remediation, or protocol-adjacent changes within a 30 to 90 day window.
How quickly can an enrollment rescue show results?
Fast-win interventions like telephone reminders and targeted online campaigns can show measurable effects within one to two weeks of launch, based on recruitment intervention research. Site-level fixes and decentralized elements typically take two to four weeks longer before their full effect on velocity shows up.
Which sites should be shut down versus rescued?
A site is a reasonable STOP candidate when the problem is structural, such as a lost referral source or a misaligned eligible population, since these rarely resolve with added resources. A site is a better RESCUE candidate when a remediation package, like a dedicated CRA plus scheduling support, can be deployed within about two weeks and is projected to meaningfully improve velocity.
Are decentralized trial elements safe to add mid-study?
Decentralized elements like telehealth visits can reduce participant burden, but FDA guidance requires that any such element be documented in the protocol with a monitoring and training plan that protects data quality and safety reporting. Pilot any decentralized change in a small cohort before applying it trial-wide.
What KPIs matter most during an enrollment rescue?
Daily enrollment count, weekly enrollment velocity per active site, and time-to-first-contact for new leads are the three metrics that surface problems fastest. Cost per enrolled participant and screen-to-enroll ratio matter more on a monthly cadence for judging whether an intervention is worth scaling.
Sources
- Accrual Quality Improvement Program for clinical trials
- Guidance | Conducting Clinical Trials With Decentralized Elements (FDA)
- Online recruitment efficiency meta-analysis (JMIR, 2020)