epiClinicalA fictional CRO

epiClinical is fictional. It is a portfolio demonstration by Adam Daniels. Its studies, sponsors, sites and enrolment figures are synthetic. The evidence behind its assumptions is real, and every number says how far it can be trusted.

EPI-512, Phase III. Recruitment complete

Moderate to severe ulcerative colitis

WGB-512 induction against placebo in moderate to severe ulcerative colitis. Sponsor: Woodgrove Biologics (fictional), Immunology and GI.

Which sites are ahead, on target or behind their expectation, where each loses its patients, and what might help.

Sites at a glance

Data to September 30th 2026. Sponsor: Woodgrove Biologics (fictional). Study-level KPIs and QTLs have their own tabs: KPIs and QTLs.

Patients randomised

471 of 471

100% of the target

Sites open

70 of 70

Every planned site is open

Patients screened

930

Up: +49 in the last three months

Sites behind their expectation

40 of 70 judged

Behind 18 ahead, 12 on target

See sites behind

Agreed, of patients approached

56%

55% in published cancer trials

Randomised, of patients screened

51%

The plan assumed 52%

Are enough patients being screened?

Over the last three months the study screened about 16 patients a month; since it began, 51% of patients screened have been randomised. It has reached its randomisation target, so no more screening is needed.

Screening is keeping pace with the plan

Patients screenedCumulative, against the screens the plan implies: its expected randomisations divided by the 52% it expects to pass screening

Show data
Patients screened: planned and actual, by month
MonthPlannedActual
Sep 202100
Dec 202100
Mar 202221
Jun 20221717
Sep 20225051
Dec 202299107
Mar 2023159168
Jun 2023221226
Sep 2023283275
Dec 2023345335
Mar 2024407384
Jun 2024469447
Sep 2024531523
Dec 2024593580
Mar 2025655635
Jun 2025717693
Sep 2025779755
Dec 2025841825
Mar 2026903881
Jun 2026965930

Sources: epiClinical synthetic screening logs; screen failure by therapeutic area from the planner parameters (D23)

How each kind of site is doing

Each site is compared with what it was expected to recruit in its months open, given its type, track record, country and the protocol. Below 75% of that is behind; above 125% is ahead.

40 out of 70 sites are behind; sites of the same kind differ widely

Patients randomised as a share of those expectedEach dot a site, by profile; log scale; the shaded band is on target

Show data
Each judged site: patients randomised against expected, by profile
SiteProfileRandomisedExpectedObserved ÷ expectedStanding
512-US-01, PhoenixSite new to research103.3×2.99Ahead
512-US-02, BaltimoreCommunity specialist practice04.3×0.00Behind
512-US-03, MiamiRegional teaching hospital105.7×1.74Ahead
512-US-04, NashvilleDedicated research centre1310.3×1.27Ahead
512-US-05, PhiladelphiaAcademic medical centre55.9×0.84On target
512-US-06, BostonCommunity specialist practice34.2×0.72Behind
512-US-07, HoustonDedicated research centre69.8×0.61Behind
512-US-08, Salt Lake CityRegional teaching hospital118.6×1.28Ahead
512-US-09, New YorkAcademic medical centre78.5×0.82On target
512-US-10, AtlantaCommunity hospital52.4×2.08Ahead
512-US-11, CharlotteCommunity specialist practice24.4×0.45Behind
512-US-12, ClevelandCommunity hospital04.7×0.00Behind
512-US-13, DallasCommunity hospital26.6×0.30Behind
512-US-14, SeattleRegional teaching hospital95.6×1.61Ahead
512-US-15, DenverDedicated research centre13.6×0.28Behind
512-US-16, ChicagoAcademic medical centre128.0×1.50Ahead
512-US-17, Los AngelesAcademic medical centre33.1×0.98On target
512-US-18, PittsburghCommunity hospital44.7×0.86On target
512-US-19, San DiegoSite new to research25.5×0.36Behind
512-US-20, MinneapolisCommunity hospital35.1×0.58Behind
512-UK-01, BristolRegional teaching hospital43.5×1.15On target
512-UK-02, SheffieldRegional teaching hospital26.8×0.30Behind
512-UK-03, NewcastleRegional teaching hospital59.8×0.51Behind
512-UK-04, LondonDedicated research centre28.4×0.24Behind
512-UK-05, CardiffRegional teaching hospital36.8×0.44Behind
512-UK-06, BirminghamAcademic medical centre43.5×1.16On target
512-UK-07, LiverpoolCommunity hospital25.9×0.34Behind
512-UK-08, SouthamptonRegional teaching hospital16.6×0.15Behind
512-ES-01, BarcelonaAcademic medical centre17.3×0.14Behind
512-ES-02, Santiago de CompostelaRegional teaching hospital399.8×3.98Ahead
512-ES-03, SalamancaRegional teaching hospital77.2×0.97On target
512-ES-04, SevilleAcademic medical centre59.9×0.50Behind
512-ES-05, MurciaRegional teaching hospital47.0×0.57Behind
512-ES-06, PamplonaRegional teaching hospital107.6×1.31Ahead
512-ES-07, BilbaoSite new to research05.8×0.00Behind
512-ES-08, AlicanteSite new to research65.3×1.13On target
512-ES-09, ValenciaSite new to research75.4×1.29Ahead
512-ES-10, GranadaSite new to research35.4×0.56Behind
512-PL-01, LublinDedicated research centre195.5×3.43Ahead
512-PL-02, KatowiceSite new to research28.6×0.23Behind
512-PL-03, RzeszówSite new to research25.5×0.37Behind
512-PL-04, ŁódźCommunity specialist practice47.3×0.55Behind
512-PL-05, WrocławAcademic medical centre1113.3×0.83On target
512-PL-06, KrakówAcademic medical centre19.8×0.10Behind
512-PL-07, BydgoszczSite new to research38.4×0.36Behind
512-PL-08, PoznańDedicated research centre312.4×0.24Behind
512-PL-09, OlsztynDedicated research centre816.9×0.47Behind
512-PL-10, WarsawAcademic medical centre19.9×0.10Behind
512-PL-11, BiałystokCommunity specialist practice06.8×0.00Behind
512-PL-12, ZamośćCommunity specialist practice46.8×0.58Behind
512-PL-13, GdańskAcademic medical centre89.1×0.88On target
512-PL-14, SzczecinSite new to research59.0×0.56Behind
512-PL-15, LublinCommunity specialist practice137.5×1.73Ahead
512-PL-16, KatowiceRegional teaching hospital39.6×0.31Behind
512-TR-01, DiyarbakırCommunity hospital16.4×0.16Behind
512-TR-02, AdanaRegional teaching hospital2710.0×2.69Ahead
512-TR-03, AntakyaSite new to research15.4×0.18Behind
512-TR-04, MalatyaRegional teaching hospital3610.9×3.31Ahead
512-TR-05, BursaRegional teaching hospital23.8×0.53Behind
512-TR-06, KayseriRegional teaching hospital17.1×0.14Behind
512-TR-07, IstanbulCommunity hospital36.1×0.49Behind
512-TR-08, AnkaraCommunity specialist practice187.9×2.27Ahead
512-IN-01, HyderabadAcademic medical centre128.6×1.40Ahead
512-IN-02, ChandigarhRegional teaching hospital38.6×0.35Behind
512-IN-03, KochiSite new to research58.1×0.62Behind
512-IN-04, VelloreDedicated research centre910.5×0.86On target
512-IN-05, ChennaiCommunity hospital3210.3×3.09Ahead
512-IN-06, BengaluruAcademic medical centre108.7×1.15On target
512-IN-07, JaipurSite new to research106.1×1.63Ahead
512-IN-08, BhubaneswarRegional teaching hospital14.2×0.24Behind

Source: epiClinical synthetic sites

Sites behind, and what might help

Largest shortfall first. Each site's screening log is split into three stages against the study as a whole; the stage marked is where it loses most. A likely cause needs both a leak at its stage and the fact that goes with it, so it was found for 26 of 40 sites. Suggestions wait for a person to accept or dismiss them. Nothing changes the plan on its own, and in a real system each decision is audit-logged.

Behind

512-PL-08, Poznań

Dedicated research centre, Poland · 3 randomised against 12.4 expected in 45 months open
  1. Approached×0.167 patients, 44 expected
  2. Agreed–7 patients approached, too few to judge
  3. Passed screening–3 patients screened, too few to judge
Likely cause
Competing trials take the same patients. The site approaches fewer patients than a site like it should, and runs other trials for the same patients. Investigators offer each patient one trial, so this study gets only a share of them. Known about this site: two or more competing trials open at the site.
Likely cause
Too few patients referred. The site approaches fewer patients than a site like it should, and has no clinic of its own for these patients. It depends on other doctors remembering to refer. Known about this site: relies on referrals, with no formal referral network.
Also known
Does not check records against the criteria before inviting patients; Two in five patients or more travel over an hour.
  1. Agree with the investigator how patients are offered the competing trials.

    An operational step with no outcome evidence; it makes the site's choice visible.

  2. Central campaigns, recruitment vendors and referral networks. Referral routes from centres without the competing trials widen the pool.

    D Evidence grade D. No quantitative evidence found in peer-reviewed comparative studies within this review. Context: in cancer care, no trial was available at the patient's centre for 55.6% of patients, which referral networks aim to address.

  3. Add sites or countries. If several sites share the problem, add sites where competition is lower.

    D Evidence grade D. No quantitative evidence on adding sites as a rescue. Diminishing returns follow from activation lag and site variability rather than from a measured effect. Context: median 0.95 participants per centre per month (IQR 0.42 to 2.60) in NIHR trials.

  4. Central campaigns, recruitment vendors and referral networks. A formal referral network or a central campaign brings patients the site cannot reach itself.

    D Evidence grade D. No quantitative evidence found in peer-reviewed comparative studies within this review. Context: in cancer care, no trial was available at the patient's centre for 55.6% of patients, which referral networks aim to address.

  5. Opt-out contact for potential participants. Where the site holds a patient database, opt-out contact reaches more of it; needs ethics and data-protection approval.

    C Evidence grade C. RR 1.39 (95% CI 1.06 to 1.84), but from a single study of 152 people; not among the high-certainty findings of the 2018 update.

Behind

512-PL-10, Warsaw

Academic medical centre, Poland · 1 randomised against 9.9 expected in 46 months open
  1. Approached×0.7727 patients, 35 expected
  2. Agreed×1.0056%, study 56%
  3. Passed screening×0.137%, study 51%
Likely cause
Patients fail screening. More of this site's screened patients fail the criteria than at other sites in the study. Sites that invite patients before checking their records find out at screening, after the patient has consented and had tests. Known about this site: does not check records against the criteria before inviting patients.
Also known
Two or more competing trials open at the site; Two in five patients or more travel over an hour.
  1. Screening-log review to find the leaking funnel stage. Find which criteria the site's patients fail before choosing a fix.

    C Evidence grade C. No quantitative evidence that logs raise recruitment. In one oncology trial, logs showed 91% of eligible patients approached but 40% agreeing, pointing at the consent conversation; logs were incomplete (47% returned; 28% of enrolled patients missing).

  2. Pre-screen records against the main criteria before inviting patients.

    A process step with no outcome evidence; it moves the failure before consent, sparing patients needless tests.

  3. Site re-training and extra site support visits. If the log shows avoidable errors, a refresher on the criteria targets them.

    D Evidence grade D. No quantitative evidence found; the 2021 Cochrane retention review found only two studies targeting site staff.

Behind

512-PL-09, Olsztyn

Dedicated research centre, Poland · 8 randomised against 16.9 expected in 44 months open
  1. Approached×0.5533 patients, 60 expected
  2. Agreed×1.4179%, study 56%
  3. Passed screening×0.6131%, study 51%
Likely cause
None of the known causes fits the stage that is leaking.
Also known
Does not check records against the criteria before inviting patients.
  1. Screening-log review to find the leaking funnel stage. No known cause fits, so start by finding which stage loses patients and why.

    C Evidence grade C. No quantitative evidence that logs raise recruitment. In one oncology trial, logs showed 91% of eligible patients approached but 40% agreeing, pointing at the consent conversation; logs were incomplete (47% returned; 28% of enrolled patients missing).

Behind

512-PL-06, Kraków

Academic medical centre, Poland · 1 randomised against 9.8 expected in 45 months open
  1. Approached×0.062 patients, 35 expected
  2. Agreed–2 patients approached, too few to judge
  3. Passed screening–1 patient screened, too few to judge
Likely cause
Too little coordinator time. The site approaches fewer patients than a site like it should, and has little coordinator time for the study. Finding and approaching patients is usually the first task dropped when a coordinator is stretched or leaves. Known about this site: half a coordinator or less on this study.
Also known
Does not check records against the criteria before inviting patients.
  1. Increase site or investigator payments. Fund protected coordinator time, if the budget allows; there is no evidence it raises recruitment on its own.

    D Evidence grade D. No quantitative evidence for patient recruitment; a 2005 systematic review found almost no experimental evidence and raised conflict-of-interest concerns. One RCT of paying doctors concerned recruiting doctors as study subjects, not patient recruitment.

  2. Site re-training and extra site support visits. Extra CRA support visits while the site rebuilds its team.

    D Evidence grade D. No quantitative evidence found; the 2021 Cochrane retention review found only two studies targeting site staff.

  3. Close or replace non-enrolling sites. If the site still cannot staff the study after an agreed period, redirect the budget to a site that can.

    D Evidence grade D. No quantitative evidence on the effect of closing sites. Context only: 11% of activated sites enrol no one and 37% under-enrol (Tufts CSDD 2012 data).

Behind

512-PL-11, Białystok

Community specialist practice, Poland · 0 randomised against 6.8 expected in 42 months open
  1. Approached×0.174 patients, 24 expected
  2. Agreed–4 patients approached, too few to judge
  3. Passed screening–0 patients screened, too few to judge
Likely cause
Too few patients referred. The site approaches fewer patients than a site like it should, and has no clinic of its own for these patients. It depends on other doctors remembering to refer. Known about this site: relies on referrals, with no formal referral network.
Also known
Does not check records against the criteria before inviting patients; Screening scans are done at another hospital.
  1. Central campaigns, recruitment vendors and referral networks. A formal referral network or a central campaign brings patients the site cannot reach itself.

    D Evidence grade D. No quantitative evidence found in peer-reviewed comparative studies within this review. Context: in cancer care, no trial was available at the patient's centre for 55.6% of patients, which referral networks aim to address.

  2. Opt-out contact for potential participants. Where the site holds a patient database, opt-out contact reaches more of it; needs ethics and data-protection approval.

    C Evidence grade C. RR 1.39 (95% CI 1.06 to 1.84), but from a single study of 152 people; not among the high-certainty findings of the 2018 update.

Behind

512-PL-16, Katowice

Regional teaching hospital, Poland · 3 randomised against 9.6 expected in 44 months open
  1. Approached×0.6221 patients, 34 expected
  2. Agreed×1.2871%, study 56%
  3. Passed screening×0.3920%, study 51%
Likely cause
Patients fail screening. More of this site's screened patients fail the criteria than at other sites in the study. Sites that invite patients before checking their records find out at screening, after the patient has consented and had tests. Known about this site: does not check records against the criteria before inviting patients.
  1. Screening-log review to find the leaking funnel stage. Find which criteria the site's patients fail before choosing a fix.

    C Evidence grade C. No quantitative evidence that logs raise recruitment. In one oncology trial, logs showed 91% of eligible patients approached but 40% agreeing, pointing at the consent conversation; logs were incomplete (47% returned; 28% of enrolled patients missing).

  2. Pre-screen records against the main criteria before inviting patients.

    A process step with no outcome evidence; it moves the failure before consent, sparing patients needless tests.

  3. Site re-training and extra site support visits. If the log shows avoidable errors, a refresher on the criteria targets them.

    D Evidence grade D. No quantitative evidence found; the 2021 Cochrane retention review found only two studies targeting site staff.

Zero out of 70 judged sites are outside the range the planner's model allows by chance. Site rates vary so widely (coefficient of variation 0.9) that one site's shortfall is rarely proof of a problem. Behind is a prompt to look, not a verdict.

What sites behind have in common

Share of sites with each fact, by how they are doing. A fact far more common among sites behind is worth checking at feasibility for the next study. In this synthetic data the links were built in; with real data, this table is where they would first show.

FactAheadOn targetBehind
Does not check records against the criteria before inviting patients22%25%60%
Screening scans are done at another hospital11%0%18%
Half a coordinator or less on this study0%0%35%
Two or more competing trials open at the site0%0%28%
Relies on referrals, with no formal referral network0%0%13%
Two in five patients or more travel over an hour0%0%18%

Every site

Show all 70 sites
SiteProfileMonths openExpectedRandomisedApproachedScreenedPassed screeningStanding
512-US-01, PhoenixSite new to research50.73.310312148%Ahead
512-US-02, BaltimoreCommunity specialist practice48.34.3020–Behind
512-US-03, MiamiRegional teaching hospital49.15.710301567%Ahead
512-US-04, NashvilleDedicated research centre50.110.313473043%Ahead
512-US-05, PhiladelphiaAcademic medical centre50.65.95181145%On target
512-US-06, BostonCommunity specialist practice47.34.238650%Behind
512-US-07, HoustonDedicated research centre48.09.86251443%Behind
512-US-08, Salt Lake CityRegional teaching hospital52.68.611372055%Ahead
512-US-09, New YorkAcademic medical centre52.18.57391450%On target
512-US-10, AtlantaCommunity hospital48.12.458771%Ahead
512-US-11, CharlotteCommunity specialist practice50.14.4212729%Behind
512-US-12, ClevelandCommunity hospital47.54.70110%Behind
512-US-13, DallasCommunity hospital47.66.6212825%Behind
512-US-14, SeattleRegional teaching hospital47.75.69281656%Ahead
512-US-15, DenverDedicated research centre49.63.615333%Behind
512-US-16, ChicagoAcademic medical centre48.78.012422060%Ahead
512-US-17, Los AngelesAcademic medical centre51.83.138560%On target
512-US-18, PittsburghCommunity hospital47.04.7420944%On target
512-US-19, San DiegoSite new to research49.95.5232100%Behind
512-US-20, MinneapolisCommunity hospital52.05.138560%Behind
512-UK-01, BristolRegional teaching hospital52.63.5411757%On target
512-UK-02, SheffieldRegional teaching hospital50.96.8221825%Behind
512-UK-03, NewcastleRegional teaching hospital52.89.85301338%Behind
512-UK-04, LondonDedicated research centre50.78.4215922%Behind
512-UK-05, CardiffRegional teaching hospital50.86.8317475%Behind
512-UK-06, BirminghamAcademic medical centre52.43.5412757%On target
512-UK-07, LiverpoolCommunity hospital52.45.927633%Behind
512-UK-08, SouthamptonRegional teaching hospital49.96.616333%Behind
512-ES-01, BarcelonaAcademic medical centre46.57.3111617%Behind
512-ES-02, Santiago de CompostelaRegional teaching hospital45.09.8391056857%Ahead
512-ES-03, SalamancaRegional teaching hospital46.37.27341644%On target
512-ES-04, SevilleAcademic medical centre45.69.9515956%Behind
512-ES-05, MurciaRegional teaching hospital44.87.04171136%Behind
512-ES-06, PamplonaRegional teaching hospital48.97.610362148%Ahead
512-ES-07, BilbaoSite new to research49.35.8020–Behind
512-ES-08, AlicanteSite new to research45.35.36251060%On target
512-ES-09, ValenciaSite new to research46.55.47261450%Ahead
512-ES-10, GranadaSite new to research46.15.4319838%Behind
512-PL-01, LublinDedicated research centre40.45.519532966%Ahead
512-PL-02, KatowiceSite new to research41.88.6252100%Behind
512-PL-03, RzeszówSite new to research44.55.529633%Behind
512-PL-04, ŁódźCommunity specialist practice44.57.34241040%Behind
512-PL-05, WrocławAcademic medical centre43.413.311432152%On target
512-PL-06, KrakówAcademic medical centre44.89.8121100%Behind
512-PL-07, BydgoszczSite new to research41.08.43191030%Behind
512-PL-08, PoznańDedicated research centre45.412.4373100%Behind
512-PL-09, OlsztynDedicated research centre44.316.98332631%Behind
512-PL-10, WarsawAcademic medical centre45.69.9127157%Behind
512-PL-11, BiałystokCommunity specialist practice41.76.8040–Behind
512-PL-12, ZamośćCommunity specialist practice41.76.849944%Behind
512-PL-13, GdańskAcademic medical centre41.79.18372433%On target
512-PL-14, SzczecinSite new to research43.99.05291242%Behind
512-PL-15, LublinCommunity specialist practice45.77.513311776%Ahead
512-PL-16, KatowiceRegional teaching hospital44.29.63211520%Behind
512-TR-01, DiyarbakırCommunity hospital47.86.415333%Behind
512-TR-02, AdanaRegional teaching hospital46.010.027704461%Ahead
512-TR-03, AntakyaSite new to research46.35.415333%Behind
512-TR-04, MalatyaRegional teaching hospital49.910.9361046060%Ahead
512-TR-05, BursaRegional teaching hospital48.33.8262100%Behind
512-TR-06, KayseriRegional teaching hospital45.27.119617%Behind
512-TR-07, IstanbulCommunity hospital45.66.1312560%Behind
512-TR-08, AnkaraCommunity specialist practice48.37.918422669%Ahead
512-IN-01, HyderabadAcademic medical centre50.08.612382157%Ahead
512-IN-02, ChandigarhRegional teaching hospital50.18.6312650%Behind
512-IN-03, KochiSite new to research50.48.15231242%Behind
512-IN-04, VelloreDedicated research centre48.710.59251656%On target
512-IN-05, ChennaiCommunity hospital50.710.3321065657%Ahead
512-IN-06, BengaluruAcademic medical centre50.58.710371856%On target
512-IN-07, JaipurSite new to research47.66.110171663%Ahead
512-IN-08, BhubaneswarRegional teaching hospital49.04.2110250%Behind

Rules behind this page

  • A site that has recruited below this share of its expected count is behind: 75%. epiClinical assumption, between the enrolment flag (0.8) and the point NCI trials rarely recovered from (0.6); about half of sites fall short of plan in practice (11% enrol nobody, 37% under-enrol; Tufts, register S-01).
  • A site that has recruited above this share of its expected count is ahead: 125%. epiClinical assumption, set symmetrically to the behind line on a ratio scale (1 ÷ 0.8).
  • Patients a site must have been expected to recruit before it is judged: 2. epiClinical assumption: below this, a site's count says more about chance than about the site.
  • Patients offered a trial who agree to take part: 55%. 55.0% of patients offered a cancer treatment trial agreed, across 35 studies (Unger 2021, JNCI, https://doi.org/10.1093/jnci/djaa155).
  • Patients needed at a funnel stage before a site's rate there is compared with the study's: 8. epiClinical assumption: with fewer, a single patient moves the rate by more than ten points.
  • A site's rate at a funnel stage is out of line when so poor a rate would happen by chance less often than this: 0.1. epiClinical assumption: a loose screen, because it prompts a look at the screening log rather than a judgement.
  • Patients fail screening: Screen failure is high and varies widely by therapeutic area (register PM-05); restrictive criteria are the commonest design barrier (PB-02, PD-01).
  • Patients wait too long for scans: epiClinical assumption from CRA practice; procedures per protocol rose 40.6% since 2010, which weighs on sites without services on site (register S-24).
  • Too little coordinator time: Coordinator turnover of 10 to 37% a year, rising to 35 to 61% in 2022; 45% of sites say capacity limits new studies (register S-09, S-11).
  • Competing trials take the same patients: Competing trials for the same patients reduce enrolment where the eligible pool is scarce (register EB-02, S-21).
  • Too few patients referred: Research-focused sites recruit about 41% better, but depend on referral routes (register S-07); referral networks lack outcome evidence (action A14).
  • Patients decline because of the journey: Procedural and visit burden lowers enrolment and retention (register PD-03); decentralised elements help mainly where travel is the barrier (PD-20).