Screening questionnaires

Any stage in Trialflare can be switched into questionnaire mode. That turns the form into a public web page with its own link, which anyone can complete without an account, an app, or a participant record.

It is designed for two jobs: ad-hoc data collection like a poll or survey, and — the focus of this article — participant screening, where you collect contact details from people interested in your study, score their answers automatically, and act on the result.

How screening works, end to end

  1. You build a stage with your screening questions and switch on questionnaire mode.
  2. You add scoring rules to the questions that determine eligibility.
  3. You add rules that watch the total score and take actions — marking the submission, messaging the person, or turning them into a participant.
  4. You share the public link.
  5. Submissions arrive in the Questionnaires area for review.

Step 1 — Build the stage

Create a stage as normal (Stages > New stage) with the questions you want to screen on. Each question is backed by a data type, exactly as in any other stage.

Two things to keep in mind while writing screening questions:

  • Conditional fields still work. A question hidden by a condition is not shown and does not contribute to the score.
  • A few question types are not supported on public questionnaires. If you use one, the public page shows a note in its place rather than the field, so it is worth opening your own link once before sharing it.

On the Content tab you can also set post-submission text — rich text shown to everyone who completes the form, whatever their answers.

Step 2 — Turn on questionnaire mode

Open the stage and go to the Questionnaire tab, then tick Enable questionnaire mode for this stage.

Your public link appears once the stage has been saved:

https://app.trialflare.com/q/<stage id>

There is a QR code button next to it, which is handy for posters, clinic leaflets and slides.

Anyone with this link can submit. There is no login, so treat it as public: do not put anything in a screening questionnaire that you would not put on a public web page.

Custom links point at the same form but carry hidden attributes with them — name/value pairs that are attached to every response submitted through that link.

Create one per recruitment channel and you get attribution for free:

Link name Attribute Value
GP surgery poster source gp-poster
Facebook ad source facebook
Partner site source partner-clinic-a

Each custom link gets its own URL and its own QR code. The attributes are stored with the response and come through in your export, so you can see which channel produced which submissions — and, later, which produced participants who actually completed the study.

Contact details

Under Contact details, choose whether to require an email address, a phone number, or both. The public form then asks for them and will not submit without them.

For screening you almost always want at least one. They are what the automated email and SMS actions send to, and what lets you get back in touch with someone who screened well.

Verifying them

Each contact detail you require has a Verify it option beneath it. With it on, the form does not just collect the detail — it checks the person can receive at it. They enter their email address or phone number, press Send verification code, and a six-digit code arrives by email or text; the form will not submit until they have entered it.

A code lasts 15 minutes, and can be re-sent if it does not arrive. After five wrong attempts it locks and a new one has to be requested. If someone changes their email address or phone number after verifying it, the verification is dropped and they verify the new one — so the detail that reaches you is always the one that was confirmed.

Verification is off unless you turn it on, and verifying by text message uses your trial wallet like any other SMS. It is worth the cost when contact details matter: on a public link that anyone can submit through, it is the difference between an address someone typed and one you know reaches them.

Limiting the form to people who have already consented

Beneath Verify it, each contact detail also has It must already have consented to this trial. With it on, a submission is only accepted if the email address or phone number the person enters appears on a consent already completed for this trial — so the questionnaire is closed to everyone else. This is what you want for screening that happens after consent rather than before it.

Turning it on turns Verify it on too, and holds it there. Both checks are needed together: without verification anyone who knew a participant's email address could screen in as them, and matching the consent would prove nothing. The contact detail is checked against your consent records only once the code has been entered, so the form never tells a passer-by whether a given address or number is on your trial.

A phone number matches whether the person consented by SMS or by WhatsApp. If eConsent is not enabled for the trial there are no consent records to match against, and nobody would be able to submit — the stage editor warns you when that is the case.

If you are screening after consent, also look at Continue to a questionnaire in your consent form's settings (Setting up eConsent): it puts a button at the end of consent that brings people straight here with their contact details already filled in, and links their answers back to the consent they gave.

Step 3 — Score the answers

Scoring lives on the individual questions, not on the questionnaire as a whole.

In the stage Content editor, open a question and go to its Scoring tab. Add one or more rules of the form if the answer … then add N points:

Condition Meaning
exists / does not exist Whether the question was answered at all
equal / not equal Matches a specific value
more than / less than Numeric and date comparisons

Points can be negative, which is often the cleaner way to express an exclusion criterion.

The total score for a submission is the sum of the points from every question. That single number is what your rules act on, so it is worth sketching your thresholds before you start: for example, +1 for each inclusion criterion met and a large negative for any exclusion, so that anyone excluded lands well below the pass mark.

A few specifics:

  • Questions hidden by a condition never score.
  • Calculation fields score on their output, which must be a number.
  • PROMIS forms score on their T-score, and PROMIS batteries on their PROPr score. A form or battery that could not be scored counts as not existing.

Step 4 — Add rules

Back on the Questionnaire tab, the Rules section defines what happens when a submission arrives.

Each rule has conditions and actions.

Conditions

Conditions compare the total score: equal to, greater than, or less than a threshold you set. Multiple conditions on one rule are combined with AND — so greater than 5 AND less than 10 gives you a band.

A rule with no conditions applies to every submission, which is the simplest way to send everyone the same thank-you message.

Score-based conditions need a score. If a submission produces no score at all — for example, every scored question was hidden by a condition — score-based rules do not fire. A rule with no conditions still does.

Actions

Action What it does
Mark as successful Flags the response green in the Questionnaires list. Your "passed screening" marker.
Mark as unsuccessful Flags it red.
Allocate to next unallocated participant Turns the submission into a real participant — see below.
Email participant Sends an email to the address given in the submission.
SMS participant Sends a text to the number given in the submission.
Show on-screen message Displays a message immediately, on the confirmation screen.

Email and SMS messages are plain text. The on-screen message is rich text, so it can include formatting, images and links — which makes it the natural place to send someone onwards to a booking page or the next form.

You can add several actions to one rule, and several rules to one stage. Every rule whose conditions are met fires, so a submission can be marked, emailed and shown a message. If more than one rule matches, all of their on-screen messages are shown in order.

A typical screening setup is three rules:

  • Score greater than 7 → mark as successful, email participant, show on-screen message.
  • Score less than 8 → mark as unsuccessful, show a polite "you're not eligible this time" on-screen message.
  • No conditions → email participant a copy of what they submitted, if you want everyone to get an acknowledgement.

Turning a submission into a participant

Allocate to next unallocated participant is how screening feeds recruitment. It takes the oldest unallocated participant record in your study, attaches the submission to it, and marks it allocated. Any email, phone or WhatsApp number from the submission is copied onto the participant record where that field is empty.

This means you need to have created a pool of unallocated participants in advance — that is exactly what unallocated records are for. If the pool is empty when a submission qualifies, the allocation cannot happen, so keep an eye on how many you have left during active recruitment.

Because the person is now a real participant, everything else in Trialflare becomes available to them. In particular, under Messages > Automatic you can set a message to send when the participant completes stage , which gives you a richer follow-up than the plain email and SMS actions — attachments, your choice of channel, and the study's own messaging thread.

Step 5 — Review the responses

A Questionnaires entry appears in the study sidebar as soon as any stage has questionnaire mode enabled.

Pick your stage from the dropdown and you get every submission with its score, status, contact details and — once allocated — a link to the participant record. You can filter by status to see just the successful or unsuccessful ones.

Submissions are listed newest first. Click the Date column to flip between newest and oldest first; whichever you choose is remembered for the next stage you look at. Long lists are shown 50 at a time, with page controls underneath.

From here you can:

  • View the full response, with any validations and queries against it — clicking the row does the same.
  • Auto-allocate a submission to the next unallocated participant, or Allocate… to a specific one you choose. This is the manual equivalent of the rule action, for submissions you have reviewed by hand.
  • De-allocate a response, if it was attached to the wrong person.
  • Change the status manually, overriding what the rules decided.
  • Delete a response.
  • Export every submission for the stage as a CSV, including the attributes from any custom links.

Queries work here as they do anywhere else. You can raise one by hand against any answer in a submission, and the rules under Queries > Automatic are applied to submissions as they come in, so a questionnaire can raise its own queries before anyone is enrolled. Because a submission belongs to nobody until it is allocated, such a query is recorded against the response rather than a participant: it shows in the trial's queries list with Questionnaire submission in place of a participant ID, and the link in the assignment email opens this page with the submission in question, rather than a participant record that doesn't exist yet. Rules about a participant's group or site can never be met by a submission, for the same reason — nobody has been assigned to either yet.

Things to be aware of

The link is public and unauthenticated. Anyone who has it can submit, as many times as they like. Trialflare has no way to tell two submissions from the same person apart, so build in a contact detail requirement and expect to do some de-duplication during recruitment. Turning on Verify it for a contact detail raises the bar — every submission then comes from someone who could receive a code at the address or number they gave — but it does not stop the same person submitting twice.

Submissions are not participants until you allocate them. Until then they sit against the study's questionnaire placeholder, which is why the participant column shows Unallocated.

Contact details are personal data. Everything the questionnaire collects is stored in your study and is subject to the same privacy statement participants agree to on the form itself. Collect only what you need to screen and to make contact.

Permissions

To do this You need
Create and edit stages, including questionnaire mode, scoring and rules trial.write (or trial administrator)
View stage configuration without editing it trial.readStages — implied by trial.read
See the Questionnaires area and view responses trial.readParticipantStudyData
See the email and phone columns on responses trial.readParticipantPersonalInformation
Allocate, de-allocate, change status, or delete responses Trial administrator — trial.admin, or trials.manage at team level
Export responses to CSV trial.exportData

See Roles and permissions for the full list.