Recruitment page and pre-screening

A public page of the study with the approved description and a short questionnaire; the consent, the site's leads, a patient from a lead, deletion and the counts.

A study can have a public recruitment page: the description of the study in plain language approved by the ethics committee, the sites taking part, and a short pre-screening questionnaire. A person who may fit leaves a first name or initials and an e-mail or a phone, and the team of the chosen site calls back. The page has its own random address (/r/…) for posters, leaflets and letters; it needs no account.

The page is off by default and is switched on per study, only after the ethics committee’s approval of its text is recorded. Every change of the setting, every lead, status and deletion is in the audit trail, without the person’s name, contact or answers.

The study, not the product

In Russia, advertising of prescription drugs to the public is prohibited (Federal Law No. 38-FZ “On Advertising”, article 24). So the page carries the patient information the ethics committee approved: what the study is, who may take part, what taking part means. No trade name of the drug, no promise of benefit, no call to be treated. Ask the sponsor’s lawyer to check the text before it goes to the committee. In the EU and the US the same holds in practice: the ethics committee or IRB approves recruitment material before it is used.

The text, the questions and the approval

Study menu → Recruitment → Page and approval (the project administrator). In Draft of the text and the questions write the title, the text (a blank line starts a paragraph, a line starting with # is a heading), its language and the questions. A question is one of three kinds, and you mark which answer fits:

• Yes or no: for the diagnosis (fits: yes) or a key exclusion (fits: no);
• A number in a range: the age, from 18 to 75 years, for example;
• One of several answers: one answer per line, a + at the start marks an answer that fits.

Up to 15 questions. The arrows move a question, the bin deletes it when the draft is saved. Saving the draft does not change the page.

When the committee has approved the text, press Record an approval of the draft: the version of the text, the date, the decision number and, if you like, the committee. The draft as saved becomes the text of the page, with its SHA-256 fingerprint; it never changes (the database refuses it). A new text is a new draft and a new approval record with a new version.

The consent, the data and the sites

In Consent, data and sites choose who processes the answers (the controller): the sponsor named in the study settings, or the site the person chooses. Give the contact for questions and withdrawal, how long the data of a lead are kept (6 months by default; then they are deleted automatically), the legal basis the consent names and the sites taking part (named by name and city, or by city only).

Before anything is sent, the person reads and ticks the consent: it names the controller, the purpose, who sees the data, the retention period, how to withdraw, and the legal basis: a written consent to the processing of health data under 152-FZ articles 9 and 10, an explicit consent under GDPR article 9(2)(a), or, in the US, a note on HIPAA (the answers are protected health information where the controller is a HIPAA covered entity such as a hospital site). The text of the consent is kept with each lead.

Only what is needed is asked: a first name or initials (no surname, no full name), an e-mail or a phone, the answers. They are encrypted with the study’s key when the study keeps personal data encrypted (study settings, Security).

Switching the page on

The block at the top shows whether the page is on, its address with Copy, and what still stops it: an approved text with at least one question, the controller’s contact, the sponsor’s name when the sponsor is the controller, a site taking part, enrolment open. Switch off stops new answers at once: the address then says the study is not recruiting there. New address makes the old one stop working, for example when it spread where it should not.

What the person sees

The approved text with the date and number of the approval, the sites, the questions, the name and contact fields and the consent. After sending, only one of two answers: “You may fit this study, a site will contact you” with a reference to name on the phone, or “Thank you, this study does not fit you”. The page never says which answer decided and never gives a diagnosis. A person who may fit also gets a link that withdraws the consent and deletes the data at once.

An answer that does not fit keeps nothing personal: only the site and the day, for the counts.

The site’s leads

Study menu → Recruitment → Leads. The investigators see the leads of their own sites, the project administrators of every site; data managers, monitors and the sponsor do not see them. The menu counts the new leads. A lead’s page has the name, the contact, the answers marked “fits”, the consent given, the status (new, contacted, screening visit booked, became a patient, not eligible, did not answer, withdrew consent) and notes.

Making a patient

Make a patient on the lead’s page opens the enrolment with the lead’s site only. Nothing of the pre-screening is copied into the patient: the investigator enters the patient as usual, after the informed consent. The lead becomes “Became a patient” and keeps the link to the patient.

Deletion

When the person asks, or withdraws the consent, Delete the personal data on the lead’s page deletes the name, the contact, the answers and the notes at once. The person can do it alone with their link. The retention job deletes the data of every lead past its period by itself. The lead stays in the counts without its data.

The counts on the study overview

The overview shows the funnel (visits to the page, once a day per browser; answers; may fit; contacted; became patients) and the leads by site and status: numbers only, for the sites you see.

Protection from bots

No outside service: a hidden field bots fill, the time it takes to fill the form, limits of answers per address and per hour on the study’s page. The page loads no outside script, font, counter or analytics; search engines are asked not to index it.