For clinicians, treatment centers and researchers

Real data between sessions

A therapy session is short. A patient says “I was low”, but cannot show when, how much and after what. Bivela makes exactly that easy.

What a patient brings you

Trend instead of memory

A chart of mood, sleep and energy across weeks and months, with the most frequent triggers.

Session prep

The patient writes their topics and questions beforehand.

Raw questionnaire scores

With dates, to see the trend — not for diagnosis.

An encrypted export

The patient gives you whatever they choose, with a password they set.

The clinician workspace

A separate section in the app and on the clinicians' site (doctor.bivela.me, separate from the admin panel), only for accounts a center's administrator has given the clinician role in the panel. A clinician keeps a file for each patient under a name or a code and records assessments before and after treatment — at admission, weekly, at discharge and at follow-up.

Go to the clinicians' site

  • Established instruments, complete: DASS-21, PHQ-9 and PHQ-2, GAD-7 and GAD-2, CGI-S and CGI-I, and ASRM. Each is shown with its source and the source of its Persian text.
  • Two instruments shortened by Bivela's choice: the Bivela psychosis rating, adapted from the BPRS, with eight items (every kind of delusion is one item, social isolation and psychomotor (catatonic) signs by Bivela's choice, and a ready question for the patient on hallucinations and delusions) and without the 18-item BPRS's severity bands; and the Bivela suicide-risk screen, adapted from the C-SSRS screener, without the method and plan questions and without any example. Their scores are not comparable with the originals, and earlier results stay with their own version.
  • The Persian text, honestly: the suicide-risk screen's questions are Columbia's official Farsi text and the DASS-21 the published Persian version. For PHQ, GAD, ASRM, BPRS and CGI no official Persian version was available, so Bivela's own translation is used and its sources say so. Each instrument's form, library entry and report state where its Persian text comes from, and any clinician can check a text against its source, part by part, and approve it in the panel.
  • Score entry for ten more instruments — HAM-D-17, YMRS, PANSS, FAST, MDQ, PSS-10, ISI, WHO-5, AUDIT and WHOQOL-BREF: their text is not in Bivela; its copyright belongs to its owners, or a valid Persian text has not reached us yet. The clinician gives or rates a licensed copy on paper and enters only each item's score on the clinicians' site, by number and a short name; Bivela adds up the scores and applies the published criteria (for example HAM-D response and remission, YMRS remission and response, PANSS linked to the CGI, a positive MDQ screen, the AUDIT risk zone and the WHOQOL-BREF and FAST domains). These forms are never sent to a patient's phone and never appear in patient mode.
  • At admission and discharge: on the clinicians' site every patient shows “admission assessment” or “discharge assessment”; a discharge opens with the admission's instruments so the two results can be compared, and the report puts admission and discharge side by side.
  • Assessments on the clinician's phone or browser: the patient's part runs in a locked mode that needs the phone's lock or the clinician's account password to leave. Answers that call for attention (such as thoughts of self-harm) are shown to the clinician at once.
  • The clinician's own questions: kept apart from the validated instruments, never scored, and stored with the exact wording asked in each session.
  • Change over treatment: any two sessions compared, and each instrument's course from its first to its latest use; reliable change (Jacobson–Truax), response and remission only where a published criterion exists.
  • A printable report: in the app and on the clinicians' site (print or PDF).
  • Answering on the patient's own phone: the clinician creates an invitation code; the patient sees the clinician's name in their own app and links only if they agree. The clinician sends questionnaires and the patient answers them there. Questions and answers are encrypted with a separate key for every link and the server cannot read them. The patient is told that answers are not seen right away and, if an answer is worrying, is shown where to get urgent help. An eight-digit check code is shown on both sides, so when you are together you can confirm no key was swapped. The patient can end the link at any time, and nothing of their data other than the answers to those questionnaires is sent — except a numbers-only summary of recent mood and sleep (never any notes), and only if the clinician asked for it and the patient, having seen the numbers, switches it on, or the sharing the patient turns on themselves (below).
  • Share with my therapist: a linked patient may, in “My doctor”, turn on sharing a daily summary with you, after seeing exactly what is sent. The daily numbers of their own check-ins over the last 90 days — mood, energy, anxiety, irritability, hours and quality of sleep — are updated every few hours when they have logged something new; their warning signs only if they allow that separately. Notes, the journal and triggers are never sent. The numbers are encrypted for you and appear on the patient's page on the clinicians' site as charts and a table, and a patient whose warning signs are showing comes first in “Follow-ups”. This is not an urgent alert, and the patient is told so. The patient can turn it off at any time; what you have seen until then stays with you.
  • Doctor mode in the app (from version 0.5.34): at the start the clinician taps “I am a doctor” and gets their own home instead of the patient's screens: follow-ups, answers that arrived and shared summaries, warning signs first; the patient list; and one way to add a patient — an invitation code for a patient with their own phone, or an account for an inpatient (below).
  • A fuller record: diagnosis (F31.x or other), the episode at admission and psychotic features, voluntary or involuntary admission, medicines and doses at admission and discharge, the discharge outcome and destination, readmission; and a completeness figure that says what is still missing.
  • “Everything” and “Bivela's analysis”: a linked patient may, with a third switch in “My doctor”, also send the questionnaires they filled in, their medicines and when they marked them taken, the sleep diary, the daily rhythm and their programs — never the text of notes or thought records, only how many. On the patient's page, “Bivela's analysis” writes its findings in your browser, each with its basis — the instruments' own alerts and reliable change, the patient's own PHQ-9 and GAD-7, warning signs, a medicine not marked for days, several nights in a week of long sleep latency or waking, days without a check-in — then a written summary you can copy. It is not a diagnosis and does not replace clinical judgement.
  • Excel of all patients: a file made in your own browser, with names or without (a code per patient): the records, every score and what patients shared. Entries made on a clinician's phone are marked apart, so they are not taken for the patient's own report.
  • Your own letterhead and signature, without Bivela's name: the full report, the discharge letter, the paper forms, the quality report and the Excel files print with your centre's name and logo, and under letters and reports come your own name, title, medical council number and an image of your signature or stamp; Bivela's name and logo are on none of them. In a centre's shared space the centre's letterhead is used and the signature is still the doctor's own. The two instruments Bivela shortened print without Bivela's name but still say “adapted from …”, so they are not taken for the originals.

Privacy: the files are encrypted with a key made from the clinician's own account password; the server keeps only ciphertext and the clinicians' site encrypts and decrypts the files in the browser. So the clinician's password matters: anyone who has it can see the files, and if it is forgotten, only the recovery code the clinician makes in the clinicians' site's settings opens them.

Not there yet: the full text of the score-entry instruments is not in Bivela until their rights holders' permission is granted. This section itself has not been evaluated in a clinical study. The full list of questionnaires, with the source of each Persian text.

For treatment centers (trial)

If a patient has no access to a phone — for example, while admitted — the center can create an account for them and enter the data together. Accounts are independent: each patient has their own account and password, and when they get their phone back they sign in with the same email and password and everything comes back.

A few important points:

  • Consent: before creating an account for a patient, their own informed consent or their legal representative's is needed, following the center's procedure. The app cannot replace it.
  • The center holds the account password and for that reason can see the data, like anyone who has an account's password. Our server cannot.
  • Phone security: a phone that has several patients' passwords on it must be locked and not shared. Turn on the app lock (“Patients on this phone” requires it). The current early-access build does not yet block screenshots or the “recent apps” preview; close the app from recent apps after each patient.
  • Patients on this phone (from version 0.5.25, Android only for now): several patients, each with their own account, on one phone; each patient's data is kept apart from the others', and tapping a patient's name opens the app for them. When you move to another patient, the current patient's changes are sent if sync is on, and the app opens again and asks for the lock. Reminders come only for the patient who is open. The name you give each patient stays on that phone. Removing a patient deletes their data from the phone; their account and its encrypted copy on the server stay.
  • An inpatient's account from the doctor's phone (from version 0.5.34): after confirming that the hospital's paper consent was obtained, the doctor makes the patient's account on their own phone — no e-mail and no first-run steps — and sharing with that doctor is on from the start (the patient sees it in “My doctor” and can turn it off). The password is kept in the doctor's own encrypted backup, so the patients come back on the doctor's other phones. At discharge a sign-in card is printed; the patient signs in on their own phone, sets their own password and takes the account over, and from then on the doctor's phones no longer have access. “Sign out of all phones” on the clinicians' site ends the doctor's sessions and those of the accounts they still hold.

Admin panel

A web panel (Persian, right-to-left, works on mobile too) for a center's administrators:

  • A dashboard and usage statistics.
  • A user list with consents, last activity and the number of items — without content.
  • Anonymous research data with a date filter and CSV export, and pseudonymous research as an Excel file with a data dictionary.
  • Managing and publishing educational content.
  • An audit log of every administrator action.
  • Sign-in with mandatory two-step verification (an authenticator app or an emailed code) and a picture code.
  • A content viewer for accounts whose password you hold; it decrypts inside the browser and every opening is logged.
  • Clinical team: the records of patients who turned on “Share with Bivela's clinical team” in the app, opened with the team's key in the administrator's browser, with “Bivela's analysis”; every opening is logged under the administrator's name, and the patient turning it off closes it at once.
  • Usage counts and errors per account and in total — how often each part is opened and the kinds of errors, no content; the user can turn it off.

Research

Research mode is optional and turns on with explicit consent. Only anonymous scores are sent and the account ID does not travel with them. This is data collection, not a clinical study; no such study has been done yet.

Pseudonymous research, separate and optional: for each finished day, numbers and choices only — mean mood, energy, anxiety and sleep, the sleep diary, the daily rhythm, medicines marked taken, the warning-signs level, session counts and questionnaire scores — under a random code made on the phone; no name, account or text. A researcher can follow each person's course over time without knowing who they are. The panel's researchers download an Excel file with a data dictionary, and every download is logged. Turning it off or «Delete all data» deletes the code and all its days from the server.

Being built (not there yet)

  • A public install of the iPhone version — built and in testing.

These are plans, not dated promises.

Talk to us

To discuss a trial at your center, write to:

Get the app