01Right patient
Wrong-chart stop
If the open chart isn't the patient the work belongs to, placement stops and the draft is kept.
Stops on the wrong chart · draft kept
Trust
Clinical decisions stay with your team.
OvaFlow is designed to coordinate around decisions your clinicians make, never to make them. This page explains who decides what, how separate facts are kept separate, what AI may and may not do, how data would be handled, and exactly how mature each part is today.
Three safeguards
Right patient. Right owner. Right next step.
Flip a card to watch the safeguard happen in a prototype walkthrough.Keep scrolling: each card turns over to show the safeguard in a prototype walkthrough. You can flip any card yourself.
01Right patient
In the walkthrough · 0:18
Maya Chen's Cycle Sheet is open in the browser, so continuation stops.
Prototype walkthrough · Fictional patient dataArtisan · Modeled workflow example · Not yet checked against a real clinic build
02Right owner
Offered is not accepted
Seeing or reading a handoff never transfers responsibility. Only an explicit accept does.
20:26 PDTCoverage offered · Nora Ruiz, RN remains responsible
02Right owner
On the on-call iPhone · 0:07
20:26 PDTCoverage offered · Nora Ruiz, RN remains responsible
20:28 PDTAlex Kim, RN accepted evening nursing follow-up
Alex accepts the evening nursing callback. Dr. Mira Lee's clinical review stays open and can't be edited here.
Design prototype · Fictional patient data
03Right next step
Unknown is not done
When a send may or may not have happened, the task stays open and the EMR is checked before anything is repeated.
Publication outcome unknown
03Right next step
In the walkthrough · 0:23
Publication outcome unknown
Check the published record in the EMR
One matching publication verified
The portal publish is submitted, but no confirmation comes back. The status reads “Publication outcome unknown”, and resending is held.
Prototype walkthrough · Fictional patient data
Five commitments
What you can hold us to
Each one has its own page with the detail a medical director, a nursing lead or a privacy lead would ask for.
- Who decidesYour clinicians decidePhysicians author and sign. Embryologists hold specimen decisions. OvaFlow prepares, tracks and checks, and never prescribes or signs.Clinical authority
- Separate factsSeparate facts stay separateReceived, signed, published, delivered, read, acknowledged, reported and accepted are different facts with different evidence.Safety model
- AI boundariesAI with firm boundariesExtraction, comparison, grounded summaries and admin drafts. No diagnosis, dosing, triage, embryo ranking or success prediction.AI boundaries
- DataData handled by design, decided per deploymentThis public site uses fictional data only. A clinic deployment requires a formal security and privacy review and signed agreements first.Privacy & data
- MaturityHonest about maturityPrototype walkthroughs, an interactive sandbox, and a pilot path. Nothing here is presented as running clinic software.Readiness
Worked example & safety research
One trigger evening, end to end
Follow one fictional evening (Elena Rivera, Oct 30, 16:45–20:28 PDT) across physician signing, wrong-chart prevention, patient confirmation and on-call nursing handoff—alongside the published clinical safety literature that shaped each rule.
Evening trigger timeline
Fictional patient & staff walkthroughHow a signed trigger change reaches the patient and evening nurse
Nora Ruiz, RN · IVF nurse coordinator
Tracks Elena Rivera's morning estradiol draw from an outside lab as an expected result and offers follow-up coverage to Lee Park, RN.
Until Lee Park explicitly accepts, Nora Ruiz still owns the follow-up.
Dr. Mira Lee · REI physician
Reviews the arrived estradiol result and signs version 3 of Elena's trigger order: Ovidrel 250 mcg at 20:00 PDT on Oct 30 (retrieval Nov 1 at 07:00 PST, 36h across the clock change).
OvaFlow retires version 2 (20:30 PDT) as a superseded instruction; the clinical decision is Dr. Lee's.
Nora Ruiz, RN · IVF nurse coordinator
Prepares to place the updated trigger instruction into the EMR while Jordan Patel's chart is open on her desktop.
Placement stops before anything is written, and the draft waits intact for Elena's chart.
Nora Ruiz, RN → Elena Rivera · Portal publish & stale-link block
After a connection drop on publish, Nora checks the portal's activity record to reconcile; on Elena's phone, an older 8:30 PM link blocks and opens version 3 (8:00 PM, Oct 30).
Elena's 17:24 PDT syringe question is logged as an open question, never a taken dose.
Elena Rivera · Patient
Reports she injected at 20:20 PDT (8:20 PM, Oct 30); the clinic receives the report at 20:22 PDT (35h 40m before the Nov 1 07:00 PST retrieval, against 36h 00m planned).
The signed 20:00 PDT order stays as signed, and Dr. Mira Lee reviews the 20-minute difference.
Nora Ruiz, RN → Alex Kim, RN · Evening on-call handoff
Nora offers the evening nursing callback to on-call nurse Alex Kim, RN at 20:26 PDT; viewing the card does not transfer ownership until Alex accepts at 20:28 PDT.
Nursing callback moves to Alex Kim at 20:28 PDT; Dr. Mira Lee's clinical review of the timing interval stays with her.
Published safety research
Full 33-gap research audit →Four clinical studies that shaped these safeguards
Starmer et al., NEJM 2014 (PMID 25372088) ↗
Structured handoffs and preventable adverse events
Published finding: Across nine hospitals, implementing a standardized handoff bundle (I-PASS) reduced medical errors by 23% and preventable adverse events by 30% without increasing handoff time.
OvaFlow design rule: OvaFlow separates 'Coverage offered' from 'Coverage accepted' and keeps clinical review ownership separate from nursing follow-up.
Adelman et al., JAMIA 2013 (PMID 22753810) ↗
Wrong-patient electronic orders via retract-and-reorder
Published finding: Surveillance of computerized provider order entry identified thousands of near-miss wrong-patient orders driven by juxtaposed patient lists and multi-tab context switches.
OvaFlow design rule: Every draft is bound to a single patient and episode; if the open EMR chart does not match, a wrong-chart stop blocks placement.
Ancker et al., BMC Med Inform Decis Mak 2017 (PMID 28395667) ↗
Alert clutter, repeated warnings, and clinician override rates
Published finding: Clinician likelihood of accepting a decision-support reminder fell by 30% for each additional reminder in the same encounter and by 10% for every 5% rise in repeated alerts.
OvaFlow design rule: OvaFlow avoids modal alert spam, deduplicates repeated source updates, and uses calm inline status badges.
ONC HealthIT.gov SAFER Guides (2024/2025) ↗
Test results reporting, CPOE, and downtime contingency resilience
Published finding: Closed-loop tracking of expected test results and structured paper-to-electronic reconciliation after system outages prevent missed abnormal findings and duplicate orders.
OvaFlow design rule: Expected labs are tracked so silence is never read as normal, and post-timeout states default to 'Publication outcome unknown.'
Today and later
Public preview versus a clinic deployment
What this site shows is a design prototype with fictional people. Real clinic use comes after testing in the clinic, review and signed agreements, in that order.
Operated today on this public site
What this public site does today
- Every patient, clinician and clinic shown is fictional.
- EMR screens are modeled reconstructions, and the walkthroughs are recordings of a design prototype.
- The /demo sandbox is a simulation with no EMR connection.
- Please don't submit patient information through any form on this site.
- Walkthrough requests by email or phone: The walkthrough form on this preview site prepares an email for you to send; it does not send or store your entries itself. Walkthrough inquiries are coordinated directly via info@ovaflow.com or 760-898-3271 (please do not include patient data).
Required before any clinical use
Clinic testing, security review, and signed agreements
- Each operation is tested in the clinic's own EMR, permissions and devices before it is enabled.
- A formal security and privacy review and signed agreements come first.
- Operations that can't meet their contract stay as explicit human steps.
For how visitor inquiries on this website are handled, see the privacy policy or write to info@ovaflow.com.
Common questions
Clinical authority and evaluation questions
Straight answers on clinical decisions, public demo data and what a deployment needs first.
Can OvaFlow prescribe medication or set a trigger time on its own?
No. OvaFlow is designed strictly to support workflow coordination and documentation around fertility care. A proposed note or draft update is never treated as a clinician-signed plan. Only licensed clinicians author, approve, and sign clinical orders inside the clinic's system of record.
Does this public website or interactive demo process real patient health information?
No. Every patient walkthrough on this public site uses fictional demonstration personas, alongside modeled EMR reference screens and simulated transcripts. Visitors should never enter patient names, identifiers, or clinical information into our public walkthrough inquiry form.
What is required before OvaFlow could be used in a live clinical pilot or deployment?
This public site is a product preview, and no compliance status, regulatory clearance, or live vendor connection is claimed today. Any future clinical pilot or clinic deployment would require clinic-specific environment testing, a formal security and privacy review, and signed commercial and data-handling agreements scoped with the clinic's legal and IT teams.
Why does OvaFlow require an explicit accept before a handoff moves ownership?
Multi-center clinical safety research (including the NEJM I-PASS handoff study by Starmer et al.) shows that ambiguous sign-outs and assumed receipt are a primary driver of preventable medical errors. In fertility care, evening trigger changes, abnormal lab callbacks and post-retrieval symptom checks cross nursing, on-call and laboratory shifts. OvaFlow keeps the current owner responsible until a named colleague explicitly accepts the specific scope.
How does OvaFlow help prevent wrong-patient chart mistakes when staff work across multiple browser tabs?
Health-IT safety studies (such as Adelman et al. on retract-and-reorder wrong-patient electronic orders) show that clinicians and coordinators frequently switch between open patient charts under time pressure. Before OvaFlow places a draft note or stages an update beside the EMR, it checks the patient identifier on the active EMR screen against the patient tied to the draft. If they differ, placement stops immediately and the draft is preserved.
How does OvaFlow avoid adding pop-up clutter and alert fatigue to a busy clinic?
Empirical research on clinical decision support (Ancker et al., 2017) demonstrates that acceptance of clinical reminders drops by 30% for each additional alert per encounter and drops further when identical warnings repeat. Rather than firing repetitive modal pop-ups, OvaFlow uses quiet state badges, single-line readiness gates, and deduplicated queue items that stay calm until an expected result is actually overdue or a chart mismatch occurs.
What happens if the clinic EMR or portal times out right after a nurse clicks send?
Blindly resending after a network timeout can deliver duplicate or conflicting medication instructions to a patient. Following ONC SAFER guidance for test-result and order resilience, OvaFlow records the state as 'Publication outcome unknown,' holds automatic retries, and prompts the staff member to perform a readback check in the clinic's own EMR first.
Where does artificial intelligence stop in OvaFlow, and can a clinic disable or govern model updates?
AI in OvaFlow is restricted to source-linked field extraction, version comparison, grounded visit summaries and administrative drafts. It never diagnoses, chooses doses, triages symptoms, ranks embryos or predicts cycle outcomes. When a source field is absent, the model abstains and shows the field as missing. Any model change requires clinic medical-director approval.
Policies
The fine print, in plain words
Evaluation discussion
Walk through clinical boundaries with your team.
Bring your medical director, nursing leadership and IT stakeholders. We will show how each boundary behaves in a modeled workflow example, with fictional records only.