All apps

CTRMS · Study & registry operators

CuRE Control

RWD-native study operations and registry lifecycle, with a persona-aware tenant model.

What it does

RWD-native study operations and registry lifecycle, with a persona-aware tenant model. Patient → Data → Studies: registries and trials layer on longitudinal patient data instead of trapping each patient record inside one protocol.

Key capabilities

  • Study setup + enrollment
  • Registry lifecycle (enrollment → submission)
  • Data-flow consent gates + re-consent tracking (governance)
  • Tenant + identity management (CRO orgs and sponsors as first-class types)
  • Canonical USDM-shaped study-design model (the MDR design plane)
  • Orchestration of Conduct + downstream products
  • Design-time Trial-Design Explorer (operating characteristics, GSD sizing, allocation-rule evaluation — consumes Calculate)
  • Risk-based monitoring visit planning (Caliber KRI-driven)
  • CTA / site-contract lifecycle
  • Research billing / coverage analysis
  • Cost accrual & budget-vs-actual
  • RWD-conditioned feasibility & site selection
CuRE Control — study-operations dashboard with active studies, sites, enrollment, and registries
Sandbox preview — synthetic demo data

What sets it apart

  • Patient → Data → Studies: registries and trials layer on longitudinal patient data instead of trapping each patient record inside one protocol.
  • Anchors the canonical, USDM-shaped study-design model — the Metadata Repository spine that the CRFs, SDTM mapping, and define.xml downstream are generated from, not hand-rebuilt against.
  • Clinical-ops parity is named directly: startup, site activation, monitoring trips, milestones, issues, and enrollment funnels belong in Control.
  • The RBQM loop closes inside the platform: a Caliber KRI/QTL breach re-triggers and re-scopes monitoring visits — the ICH E6(R3) RFP bar.
  • CRO CSV → OMOP → live Control views: outsourced study data can become governed operational signal without rebuilding the study stack.
  • Owns the canonical user/org graph — every other app resolves identity through Control.
  • Eligibility surfaced at the point of care in Cue warm-hand-offs into Control's enrollment funnel — the referral-to-enrollment gap most CTMSs leave open.
  • As the Translational bundle extends upstream toward first-in-human IND, Control orchestrates the IND journey — preclinical-to-FIH milestones on the same governed record.
CuRE Control · study startup + enrollment funnel
Live demo — synthetic data, runs in your browser

Walk a study through startup, then watch its enrollment funnel

A CTMS stands a study up — protocol, contract, IRB, per-site activation — and then watches subjects flow through the enrollment funnel. Switch the study phase and the stage matrix reshapes; pick a site and drag its screening and enrollment counts, and watch the cumulative funnel, the screen-fail rate, and the recruitment-health band recompute — graded against each site's slice of the planned-enrollment curve. The all-sites panel scores every synthetic site the same way.

A synthetic study — ravucizumab (VELT-042), 240-subject target across 4 sites.

Funnel stages: Screened → Enrolled → Randomized → On Treatment → Completed

SelectedContractingIRB approvedGreen-lightActivated

Activated — open to screening, so it contributes to the funnel.

These are the mutually-exclusive current-status counts. The funnel shows the cumulative "ever reached" count — every later stage rolls up into the earlier bars.

Study-startup milestones

Protocol & SAP finalized
Master contract & budget
Central IRB approval
Site activation (3 / 4)
First subject in
Northwood Medical Center · Site 104
Leadingtarget
Screened50
12screen fails
Enrolled4284%
3withdrawals
1lost to follow-up
Randomized3672%
On Treatment3162%
Completed918%
Conversion
84%
screen → enroll
Screen-fail
19%
of all screened
Enrolled share
51%
of study enrolment
Target Δ
+28%
plan ≈ 33 now

Recruitment health grades this site's enrolment against its slice of the planned-enrollment trapezoid — below 75% of plan is lagging, above 110% is leading. With no site target it falls back to the site's conversion vs. the study-wide rate.

All sites · recruitment health

1 leading1 on-track1 lagging
Screened115Enrolled83Study conversion72%Plan now≈ 119(behind plan)
SiteScreenedEnrolledConv.Health
Northwood Medical Center · Site 104504284%Leadingtarget
Cedar Valley Research · Site 211392974%On tracktarget
Harborview Clinical · Site 278261246%Laggingtarget
Summit Trials Institute · Site 340IRB approved00No activitytarget

Every count and band is computed in your browser from the synthetic status snapshots — no backend. Notice the funnel is cumulative: a site with a low current Screened count can still show a tall Screened bar because everyone downstream passed through it. And a busy site with a high screen-fail rate isn't the same as a healthy one — recruitment health grades enrolment against the plan curve, not raw volume.

Why this is more than a toy

The funnel model and every recruitment metric are a faithful, dependency-free port of Control's real methodology, from three source files under apps/control/src/server/dashboard/: the funnel.ts enrollment-funnel stage model — the ADR-CTL-010 StudyType stage matrix (Phase III shows Randomized; long-term follow-up and observational studies relabel On Treatment → In Follow-up), the cumulative "ever reached stage X" derivation over mutually-exclusive snapshot statuses, and the terminal-status attrition annotations — the site-funnel.ts per-site comparison layer (screen→enroll conversion, screen-fail rate, enrolled share, and the recruitment-health band with its 0.75 lagging / 1.1 leading thresholds), and the plan.ts planned-enrollment curve — a piecewise-linear ramp/steady/taper trapezoid whose area equals the target — ported line-for-line. The startup-milestone chain makes the feasibility → screening activation handoff (ADR-CTL-066) legible: a site that hasn't been activated contributes nothing to the funnel, no matter how promising its feasibility count. Per ADR-PLT-044 all of this is Control's own operational read-math — pure, bound to no statistical-methodology artifact — so it lives in-app rather than in Calculate. Here it all runs entirely client-side on a synthetic study — no backend, no real data.

See CuRE Control in action

Every research ecosystem is unique. Let's discuss how CuRE can be configured for your needs.