KSCOUT Intelligent Site Selection

    Clinical trial site selection, backed by clinical intelligence.

    KScout helps sponsors and CROs identify the right clinical research sites for each protocol using Kitsa's Clinical Research Network - the world's largest network of recruiting and research-capable sites. It brings together site and investigator experience, past trial history, ICD-10 level disease prevalence data, demographic data, geographic coverage, site capabilities, and disease-specific markers so teams can build a defensible, protocol-fit shortlist.

    KScout · Atlas · KT-204
    KScout KT-204 · Phase II Live
    412 sites scored 38 shortlisted Top score 94
    SITE 042· Oncology · Academic
    Dana-Farber Cancer Institute
    Boston, MA · United States
    Score
    94
    Past trials
    84
    Capacity
    High
    SITE 117· Oncology · NHS Foundation
    The Royal Marsden
    London · United Kingdom
    Score
    91
    Past trials
    67
    Diversity
    Strong
    SITE 203· Oncology · University
    Charité Universitätsmedizin
    Berlin · Germany
    Score
    88
    Past trials
    92
    Quality
    A+
    SITE 318· Oncology · Academic
    National University Hospital
    Singapore
    Score
    86
    Past trials
    58
    Capacity
    High
    11 days
    Median feasibility-to-shortlist
    1,874
    Investigators scored per Phase II
    2.3×
    Protocol-fit coverage vs. baseline
    How KScout works

    A clinical trial site selection engine, not a static site list.

    KScout ingests protocol requirements, scores clinical research sites against site experience, investigator experience, past trial history, disease prevalence, and demographic data, explains every ranking dimension, and produces a defensible, protocol-fit shortlist.

    01 Inputs
    Inputs
    • Protocol & I/E criteria
    • Country strategy
    • Enrollment assumptions
    • Sponsor preferences
    Protocol ingested
    02 Signals
    Signals
    • Site experience
    • Diversity & data quality
    • Investigator & past trials
    • Disease prevalence & demographics
    Protocol signals live
    03 Score
    Score
    • Weighted ranking
    • Protocol-specific fit
    • Composite score
    • Recomputed live
    94
    88
    04 Explain
    Explain
    • Score decomposition
    • Dimension rationale
    • Evidence traceability
    • Named drivers
    05 Shortlist
    Shortlist
    • Ranked site list
    • Top candidates surfaced
    • High-confidence picks
    • Portfolio comparison
    06 Defend
    Defend
    • Audit-ready rationale
    • Shareable review output
    • Decision trail
    • Sponsor / ops alignment
    Audit trail complete
    One source of truth

    Every shortlisted site is backed by the same scoring logic.

    Hover any step or shortlist output to see how KScout connects protocol needs, site evidence, and ranked recommendations into one defensible decision path.

    01The problem

    Site selection is the most expensive guess in clinical trials.

    Sponsors send 400-question feasibility surveys to a list inherited from the last trial. Sites with capacity get missed. Sites without capacity say yes anyway. Three months and a non-trivial seven-figure spend later, 30% of activated sites enroll zero patients.

    Traditional site selection
    KScout
    Site evaluation basis
    Traditional site selectionStatic feasibility surveys returned weeks late
    KScoutEvery investigator scored against this protocol, not last quarter's
    Enrollment visibility
    Traditional site selection70% of activated sites under-enroll or miss FPI
    KScoutPer-site explainability across experience, capability, and disease-prevalence signals
    Selection approach
    Traditional site selectionCRO recommendations bias toward familiar relationships
    KScoutA new site enters the rank the moment new data arrives
    Diversity and operational fit
    Traditional site selectionDiversity gets a column, not a strategy
    KScoutSite experience, data quality, diversity, disease prevalence, demographics, and site capability signals - unified
    02The scoring model

    One score.
    Six honest dimensions.

    Every site KScout ranks gets a composite score on a 0–1 scale. The score is fully decomposable: every basis point traces to a named dimension, the underlying data, and the protocol-specific weighting your team can tune.

    01
    Site experience

    Past trial participation, investigator experience, therapeutic-area exposure, and site capability signals help evaluate whether a site is a strong fit for the protocol.

    WEIGHT28%
    02
    Therapeutic depth

    Investigator publication record, prior trial role, and named-investigator presence at the site for this indication.

    WEIGHT22%
    03
    Data quality

    EDC query rate, deviation rate, missing-data rate, and audit findings across the last 24 months of trials.

    WEIGHT18%
    04
    Operational capacity

    Real-time read on competing trials, coordinator headcount, and the site's declared start-up windows.

    WEIGHT14%
    05
    Diversity index

    Catchment demographics matched to your protocol's representation goals - not as a checkbox, as a constraint.

    WEIGHT10%
    06
    Geographic distribution

    Travel-time analysis, regional regulatory exposure, and currency/CPI-adjusted cost per patient.

    WEIGHT8%
    03The output

    A shortlist that defends itself.

    Every site on the list comes with the rationale and the underlying signals. Reviewers see the math; auditors see the trail.

    KT-204 · Phase II NSCLC · EGFR+
    412 sites scored · 38 shortlisted · 5 displayed
    Geo: GlobalMin capacity: 8 pts
    #SiteScoreTrialsTop drivers
    1
    Moffitt Cancer Center
    #14021 · Tampa, FL
    0.9434
    Experience 0.96Depth 0.91Diversity 0.71
    2
    UPMC Hillman
    #9318 · Pittsburgh, PA
    0.9237
    Site fit 0.92Depth 0.95Capacity 0.82
    3
    Karolinska Institutet
    #22107 · Stockholm, SE
    0.9044
    Quality 0.97Depth 0.88Geo 0.76
    4
    MD Anderson
    #31190 · Houston, TX
    0.8939
    Prevalence 0.86Diversity 0.74Depth 0.93
    5
    Dana-Farber
    #5504 · Boston, MA
    0.8742
    Quality 0.91Capacity 0.79Depth 0.92
    33 sites below the fold · model card v2.4 · last refresh 14 min ago
    04The data underneath

    The honest answer requires honest data.

    KScout connects the data sources teams need for protocol-specific clinical trial site selection - including ICD-10 level disease prevalence data, demographic data, site and investigator experience, past trial history, geographic coverage, site capabilities, and custom disease-specific markers. Every shortlist is built from transparent, explainable signals rather than static assumptions.

    REGISTRY
    ClinicalTrials.gov
    480k+ trials
    REGISTRY
    EU CTR
    120k+ records
    PUBLICATIONS
    PubMed
    36M+ abstracts
    OPERATIONAL
    CMS / Open Payments
    11M+ records
    OPERATIONAL
    Sponsor EDC archives
    Per engagement
    EPIDEMIOLOGY
    ICD-10 Disease Prevalence
    County-level
    CLINICAL
    Disease-specific markers
    Per indication
    DEMOGRAPHIC
    Census ACS
    Block-group
    Pilot · Rare disease Phase III

    We replaced a 6-month feasibility cycle with an 11-day shortlist - and out-enrolled the original plan by 38%.

    A specialty sponsor running a competitive Phase III in adult-onset hereditary disease had a 32-site plan with three-year enrollment. KScout re-scored the candidate pool against the protocol, surfaced 14 sites the original list missed (mostly in the Southeast US and the Nordics), and projected a 27-month complete-enrollment scenario. Twenty months later, the trial is running 8% ahead of the new plan.

    Phase IIIRare diseaseGlobal · 7 countries
    14 sites
    Surfaced beyond the inherited site list
    Half outside the sponsor's prior network
    -9 months
    Projected complete-enrollment delta
    Versus the originally proposed plan
    $11.4M
    Estimated cost avoidance to date
    From compressed start-up + reduced rescue activations
    05Guardrails

    Decisions you can defend in a regulatory inspection.

    Trial-start-up decisions get audited. KScout is built so the audit goes well - every score traces to evidence, every model version is stamped, every shortlist is reproducible.

    No unexplainable matches

    Every score decomposes into named dimensions and the underlying records. If the model can't justify it, it doesn't ship it.

    Bias monitoring on by default

    Demographic parity and access metrics are tracked across every shortlist, with drift alerts when a model update materially shifts representation.

    Sponsor data stays sponsor data

    Your EDC archives, sponsor library, and protocol are processed in your tenant. Cross-customer learning is opt-in and aggregated.

    Versioned model cards

    Every shortlist is stamped with the model version, the data freshness, and the protocol-specific weights - reproducible six months later.

    Questions, answered.

    Bring smarter site selection forward.

    See how KScout helps teams identify, compare, and prioritize clinical trial sites with connected operational and geographic intelligence.

    Request a demo →