Product preview · Fictional patient data
Status language
Words like “sent”, “done” and “handed off” collapse preparation, portal posting, patient reading and clinical review into a single checkmark. Vague status labels are where false completion begins.
The instruction, portal update or clinical note has been assembled from the current signed source and is waiting for an authorized person to review and continue in the clinic's own EMR or portal.
An authorized staff member reviews the prepared text against the signed source and continues inside the clinic's own EMR or portal.
Product rule Prepared content has not been published.
Names and times are fictional examplesPick a status, or use the arrow keys. The same words appear on every screen, from the desktop rail to the watch.
For your team
Eleven statuses, each with one meaning, a stated limit and one next step.
Today“Sent”, “done” and “confirmed” mean different things to different people.
With OvaFlowEleven statuses, each with one meaning on every screen, device and role.
TodayA status hints at more than it proves.
With OvaFlowEvery status says what it does not establish.
TodaySimilar states get merged into one badge.
With OvaFlowFour pairs are never merged, so each gets the right check.
TodayNobody is sure who picks it up.
With OvaFlowEach status names the next step and who takes it.
Never merged
Each pair needs a different check or a different person. OvaFlow keeps them as separate states and never rounds one up to the other.
One is about a send that may have happened; the other is about losing sight of the screen. They need different checks.
An offer, even one that has been read, leaves the current owner responsible.
A published record proves the clinic posted it, not that the patient saw or understood it.
Acknowledging an instruction is not the same as reporting when a dose was actually taken.
Plain words for every status
In ordinary clinical software, a single green “Done” or “Sent” badge can mean that a draft was staged, that a portal send was attempted before a timeout, that a message posted to the portal (though the patient has not opened it), or that a shift handoff message was read (though nobody accepted the callback). Health services research shows that conflating transmission with receipt—or reading with action—is a primary driver of dropped outpatient follow-up [1, 2].
OvaFlow uses a closed vocabulary of eleven canonical statuses across every surface. Each status states three things in plain language: what facts are known, what facts are still missing (including explicit “It does not mean…” negations), and the exact next action and role required to move it forward. And four forbidden equivalences ensure that critical pairs—like “Publication outcome unknown” vs “Observation interrupted”, or “Coverage offered” vs “Coverage accepted”—are never merged into a single ambiguous state.
Did the physician sign the change?
Before 17:10 PDT, a staged proposal reads “Awaiting clinician approval” and never alters version 2. Once Dr. Mira Lee signs version 3 at 17:10 PDT, the portal draft becomes “Ready for review in the EMR”.
Did the clinic portal record the post?
A dropped confirmation at 17:14 PDT shows “Publication outcome unknown” (holding resends). Reading back the portal record at 17:16 PDT shows “One matching publication verified”.
When did Elena take the injection?
Elena’s report (received at 20:22 PDT) that she injected at 20:20 PDT is kept as “Patient-reported use · 20:20 PDT” and opens “Clinical review pending” for Dr. Mira Lee without rewriting version 3.
Who owns the evening callback?
Offering coverage at 20:26 PDT (and Alex reading at 20:27 PDT) keeps “Coverage offered · Nora Ruiz, RN remains responsible” until “Alex Kim, RN accepted evening nursing follow-up” at 20:28 PDT.
At 17:04 PDT, the proposed 20:00 PDT trigger shift sits under “Awaiting clinician approval” (which explicitly does not alter the signed version 2 order). Once Dr. Mira Lee signs version 3 at 17:10 PDT, the portal draft becomes “Ready for review in the EMR”.
When confirmation drops at 17:14 PDT, the status reads “Publication outcome unknown” (holding resends). At 17:16 PDT, reading back the EMR record transitions it to “One matching publication verified” (closing only the send task).
Elena’s report (received at 20:22 PDT) that she injected at 8:20 PM, Oct 30 (20:20 PDT) is recorded as “Patient-reported use · 20:20 PDT” (preserving her words without rewriting version 3) and opens “Clinical review pending” for Dr. Mira Lee.
At 20:26 PDT, the status reads “Coverage offered · Nora Ruiz, RN remains responsible” (even when Alex reads it at 20:27 PDT). At 20:28 PDT, pressing Accept updates it to “Alex Kim, RN accepted evening nursing follow-up” while Dr. Lee’s “Clinical review pending” remains open.
System boundaries
Clear operational boundaries keep clinical authority, prescribing, and laboratory witnessing inside your clinic’s own systems. See AI & clinical boundaries for full governance rules.
Questions & primary sources
Because they require completely different actions. “Publication outcome unknown” means a send was just attempted and may have posted, so you must check the EMR before resending. “Lost sight of the EMR” (“Observation interrupted”) means the screen bridge dropped while you were viewing, so you simply re-open the EMR tab to see the current state.
None of the eleven statuses use cryptic codes. Every chip states the rule in plain English right on the screen—for example, “Coverage offered · Nora Ruiz, RN remains responsible”—so a nurse on her first shift knows who owns the callback without consulting a manual.
Posting an instruction to your portal doesn’t prove you saw it, and tapping “Got it” at 5:30 PM doesn’t tell your doctor the exact minute you took the injection at 8:20 PM. Keeping those steps distinct ensures your care team acts on real facts.
See it
Recorded from an interactive design prototype. Every patient, clinician and clinic is fictional, and EMR screens are modeled reconstructions.
Beside a modeled Artisan view, the companion compares Dr. Priya Shah's approval with the unchanged Artisan fields, stops on Maya's open chart, and won't assume an update it didn't see.
A late estradiol result, a newly signed trigger time, the wrong chart caught before publishing, and a lost confirmation checked rather than blindly resent.
Next step
We'll compare the phrases your team uses today with the exact status phrases the prototype uses. Fictional records only; please don't send patient data.