Genius Cohort

Clinical trial feasibility and recruitment

How many patients in Brazil fit your protocol, and where they are.

Genius Cohort applies the study criteria to real SUS records and shows the eligible cohort, the arrival of new cases, the sites that treat these patients and the recruitment time, in minutes.

Download the Genius Cohort presentation (PDF)
Map of Brazil with the cohort patients and the sites that treat them

Real example: breast cancer treated with trastuzumab (HER2+), adults aged 18 to 75, SUS data.

SUS since 2008

Diagnoses, procedures, chemotherapy, specialized-component medications and hospitalizations from DataSUS, linked by patient.

CNES

Address, beds, accreditations and equipment of every health facility in the country.

ClinicalTrials.gov

More than 11,000 studies with sites in Brazil, with investigators and recruitment status, updated every week.

No patient identification

Aggregate counts only; any number below 10 shows as "< 10".

From protocol to recruitment plan

The images below come from the product itself, with a real search. Every number on screen comes from the search over SUS records; the AI helps build the protocol and interpret, but never invents numbers.

Build the protocol

From the protocol PDF to the criteria, no spreadsheet

Upload the protocol in Word or PDF, or paste the eligibility criteria: the AI reads the text and proposes each criterion with the SUS codes. You review, switch on and off, and decide what goes in.

  • Diagnoses (ICD-10), procedures and medications (SIGTAP), tests from a selected list (record that they were done), chemotherapy regimens, hospitalization, age, sex, recent SUS contact and death.
  • Criteria combined with AND and OR, with time windows and a minimum number of treatment months.
  • What DataSUS cannot assess (test results, scales, biomarkers) goes to a medical-record pre-screening checklist.
Protocol panel with inclusion and exclusion criteria

Protocol feasibility

See which criterion blocks recruitment

The funnel applies one criterion at a time and shows, for each, how many patients would come back without it. When the cohort falls below target, the system suggests what to remove; the simulator asks the AI for alternative criteria and measures each one with a real search.

  • Patients with the disease, final cohort and eligible new cases per month.
  • The most costly criterion is highlighted.
  • Saved scenarios side by side to compare protocol versions.
Eligibility funnel

Care distribution

Aggregate sociodemographic profile and distribution of care sites

Geographic distribution of SUS facilities providing care for the study condition, combined with aggregate sociodemographic composition (race/skin color, sex and macro-region) to plan cohort representativeness without individual identification.

  • Ranking of facilities with the most patients and new cases in the last 12 months.
  • Map with zoom by region.
  • Aggregate cohort diversity, as recorded in SUS.
Cohort diversity

Site selection

Which sites to open, and in what order

Hospitals and specialty centers with the most eligible patients in SUS and research sites with a track record in the disease or with industry, even without SUS patients, with each one's clinical research experience and competition from studies recruiting the same disease. Emergency and primary care units come in as a referral network. The AI suggests each site's role (anchor, complementary, to develop, avoid) and an activation strategy; the system calculates how many sites meet the target on time.

  • Excel export with each site's address, contacts and investigators.
  • Site profile with the investigators and the diseases of their studies.
  • The AI writes the reasoning; the numbers come from the search.
Site selection with AI-suggested roles

My site

The research site sees its own patients and who can refer

How many cohort patients your site already treats, where they come from, the sites treating the same cohort within up to 300 km, the capacity registered in CNES and competing studies in the region.

  • Patients treated, distance from where they live and the whole cohort within the radius.
  • Referrers: other sites treating the cohort near you.
  • The site's accreditations, beds and equipment.
My site panel

Recruitment

A recruitment curve to bring to the sponsor

With the target, timeline and conversion rate, Genius Cohort estimates the recruitment time and draws the curve of patients enrolled month by month, with the conversion uncertainty band. The image can be downloaded and goes into the PDF report.

  • New cases per month over the last 36 months, compared with the pace the target requires.
  • Recruitment time at 5%, 10% and 20% conversion.
  • For the whole of Brazil or just your site.
Estimated recruitment curve
New patients per month

Real example: breast cancer treated with trastuzumab (HER2+), adults aged 18 to 75, SUS data.

Download

Genius Cohort corporate presentation

27 pages: the challenge, the data sources and one real search end to end — protocol reading, funnel, simulator, map, recruitment time, site selection, referral network, budget and AI analysis — with screens from the product itself. To bring to a meeting or share with your team.

Download PDF (4 MB)

One question for each role in research

Pharmaceutical industry

Is this protocol feasible? Which criterion blocks recruitment?

Criterion-by-criterion eligibility funnel, how many patients each criterion costs, and a simulator that tests alternative protocol versions.

CRO

Where to open sites, and how many are needed?

Sites that treat the cohort, monthly arrival of new cases, research experience and competition on ClinicalTrials.gov, and how many sites meet the target on time.

Research site

Do I have patients for this study? Who can refer them?

How many protocol patients your site already treats, where they come from, who treats the same cohort nearby, the estimated recruitment curve and a pre-screening checklist for the medical record.

Frequently asked questions

How much does it cost?

There is a free trial: for research sites, 30 days or 3 studies, whichever comes first per site; for CROs and sponsors, 2 months or 2 studies, whichever comes first per company. After the trial, each new study is purchased separately in Genius Cohort itself, or under a contract with TechTrials.

Are patients identified?

No. The tool only shows aggregate counts (how many patients, where they live, where they are treated), and any number below 10 shows as "< 10". No patient data is displayed or exported.

Where does the data come from, and how recent is it?

From public SUS records (DataSUS: outpatient care, chemotherapy, medications and hospitalizations) since 2008, the facility registry (CNES) and ClinicalTrials.gov, updated every week. The most recent 3 months of outpatient data are still incomplete, because SUS takes up to 3 months to receive the records, and monthly-pace calculations leave them out.

What doesn't DataSUS show?

Test results, clinical scales, biomarkers and people treated only by private insurance. Of the tests, those on a selected list (genetic testing, viral load, PSA, echocardiography and others) are included, only as a record that they were done. That is why Genius Cohort separates what the search confirms from what must be checked in the medical record, in a pre-screening checklist. A comorbidity only appears if it was coded: excluding by it gives a ceiling, including gives a floor.

Do I need to install anything?

No. It runs in the browser, in Portuguese and English, with the same TechTrials login as the other products.

Research sites

Free access for research sites

For people working at a clinical research site in Brazil. Enter your site's CNES: Genius Cohort uses it to show the patients you already treat.

Free trial for your site: 30 days or 3 studies, whichever comes first. The limit applies to the whole site, adding up the studies of all its researchers. After the trial, each new study is purchased separately.

All patient data is aggregated and anonymized: no individual is identified, and any number below 10 shows as "< 10". Data handling follows the LGPD, Brazil's data protection law (Law No. 13,709/2018).

By requesting access, you agree to the Terms of Use and the Privacy Policy.

CROs and pharma

Free access for CROs and sponsors

Protocol feasibility, site selection and recruitment planning at scale, with AI analyses and export of site lists with contacts and investigators.

Free trial for your company: 2 months or 2 studies, whichever comes first. The limit applies to the whole company, adding up the studies of all its people. TechTrials reviews the request and grants access; after the trial, each new study is purchased separately or under a contract.

All patient data is aggregated and anonymized: no individual is identified, and any number below 10 shows as "< 10". Data handling follows the LGPD, Brazil's data protection law (Law No. 13,709/2018).

By requesting access, you agree to the Terms of Use and the Privacy Policy.