How many sites, in which countries, finishing when?
Plan recruitment the way the evidence says: every site recruits at its own pace, and the answer is a range.
The principles
Every site recruits at its own pace
The rate at each site is drawn around its country average, so a few sites carry a study and some recruit nobody. This is the Poisson-gamma model of Anisimov and Fedorov (2007), the standard in the recruitment literature.
Sites open over months, not on day one
Each country has its own start-up window, from contracts and ethics approval to the first patient, based on published start-up times.
The answer is a range
Thousands of simulated studies give a fast, a central and a slow finish. The chance of meeting the deadline is written in words, such as a three-in-five chance, because people read frequencies more accurately than decimals.
Plan backwards from the deadline
Set the patients, the deadline and the confidence you need, and the planner finds the fewest sites per country that get there, within the number of sites each country can realistically supply.
Protocol design moves the numbers
A study design questionnaire covers the risk factors found in the research, such as eligibility, visit burden and competing trials, and shows how each answer changes recruitment in each country.
Every assumption shows its evidence
Each number is marked as evidence-anchored, direction from evidence or illustrative, with its source. A CRO with trial intelligence services would recalibrate them against its own history; here they are honest estimates.
People decide
Recommendations drafted by AI never change a plan on their own. A person accepts or dismisses each one, and in a real system every decision would be audit-logged.
See the dashboard
The dashboard is open to view: six fictional studies and how their sites are doing, the planner and the KPIs and QTLs. If you want to talk to me about patient recruitment, KPIs or QTLs, please contact me.