Recruitment
Advanced Recruitment Forecast
Model site activation, the delay to first screening and the screening period before randomisation; forecast whole randomised patient counts.
Your assumptions
Build a recruitment forecast
Enter your own study assumptions. The example is illustrative, not a clinical benchmark. Yellow fields are editable.
Recruitment rates by site performance
Enter each rate per month or per year—the other field updates automatically. Rates are randomised patients per site, not screened patients. Percentages below are of recruiting sites only and must total 100%.
High-recruiting sites
Medium-recruiting sites
Low-recruiting sites
Yearly rate = monthly rate × 12. Converted values may be rounded for display; calculations retain the full rate you entered.
Enter your study assumptions to see the forecast.
Cumulative recruitment forecast
The blue line shows whole randomised patient counts obtained by rounding the cumulative expected-value forecast to the nearest patient (0.5 rounds up). The dashed line is the target.
Monthly site activation
Whole sites per month: an early activation ramp-up, a steadier middle and a late ramp-down. The schedule always totals the planned sites.
View monthly recruitment calculations
Patient counts are whole numbers. Monthly patients are calculated from the change in rounded cumulative patients, so they always reconcile. “Sites contributing” is the average expected recruiting-site equivalent during that month and may be fractional.
| Month | Month begins | New sites | Sites activated | Average sites contributing (expected equivalents) | Randomised this month | Cumulative randomised |
|---|
How it works
Model assumptions
Site activation: Enter the months needed to activate all sites. The model allocates whole sites across an early activation ramp-up, a steady middle and a late ramp-down. These are simplified planning shapes; actual country and site readiness can vary. A constant profile is possible with both ramp-up and ramp-down set to 0. Activation means a site is ready to recruit; it is not the same as initiation or first patient.
Calendar dates: Month 1 starts at the entered first site activation date. FSI on the Study Duration page is a separate milestone. An optional actual or planned FPI can be entered for comparison; it does not change the forecast. FPI means the first patient randomised, not the first patient screened.
Screening and randomisation timing: Enter the days from activation to first screening, plus the screening period in days. These are separate inputs: screening occurs before randomisation. Earliest possible screening and randomisation dates assume a pre-identified eligible patient at the first site and successful screening. To estimate the recruitment curve, sites activated in each month are represented at that month's midpoint; their randomised recruitment starts after both day-based intervals. The model prorates their first recruiting month by actual eligible days, rather than assuming a fixed half-month contribution. Non-recruiting sites are excluded first; other sites follow the high/medium/low mix at fixed rates.
Whole-patient figures: Calculations retain fractional expected values internally. Cumulative patients are rounded to the nearest whole patient (0.5 rounds up); patients in each month are the difference between consecutive rounded totals. Estimated LPI is the approximate date the expected cumulative value reaches the rounding threshold for the whole-patient target (target minus 0.5), rather than the date the unrounded value reaches the exact target. This aligns the LPI month with the displayed whole-patient total.
Recruitment rates are for randomised patients. Do not deduct screen failure again.
Limitations: The model assumes a chosen activation profile, fixed recruitment performance, a fixed delay to first screening and a fixed screening period for all sites. It does not separately simulate country differences, start-up delays, seasonality, recruitment pauses, screening volumes, screen failures, eligibility, site caps or uncertainty ranges. Screening dates and forecast FPI are estimates, not verified patient events.
Estimated FPI and LPI: These are mathematical expected-value planning estimates, not guaranteed operational dates. The forecast does not model individual sites or patients and is limited to 120 months from first site activation.
Planning estimate only. Validate assumptions against the protocol, individual site feasibility and operational plan. The forecast is limited to 120 months.