The Review Loop: Two Formularies and a Long Email Thread
Being in Florida for UCF didn't stop the review loop. Catherine walks Dental and Triage form by form over email, and two independently built formularies start converging into one.
Building at a Distance
Since the sandbox went live, the actual mechanism for finishing the forms has been email. UCF started in August, and the build now runs from Florida while the people whose forms these are, station leads and physicians in Houston, review from there. The sandbox is what makes that possible: a link and a password instead of a laptop passed hand to hand.
Two threads carried most of the traffic this stretch: a line-by-line review of Dental and Triage with Catherine Short, and the convergence of two formularies built independently on opposite sides of the same problem.
Dental and Triage, Line by Line
Catherine was Costa Rica's dental lead and one of the clinical leads who helped shape the Dental form in the first place. In mid-July I sent her the rebuilt Dental and Triage forms directly in the sandbox, pointed at a completed encounter she could reopen, edit, and abandon without anything saving.
The questions I asked her were specific on purpose: does the tooth-charting screen actually match the paper chart, are the tap targets big enough on a tablet, can a cavity be charted without picking a procedure first, is a flag enough for a patient returning to finish an extraction. Vague questions get vague answers.
Hers weren't vague either. The tooth diagram was landing well, but she flagged that Costa Rica's dentists mark an extracted tooth with an X on the chart, a convention the form didn't account for and that still needs confirming with Linda. She also asked how to edit allergies from her tablet.
The honest answer was that she couldn't. Allergies can only be added at first registration or through an admin-only edit screen, neither reachable from Triage. Answering that question surfaced a real gap: a returning patient checking in for a new visit is never asked about allergies again at all. Testing the form in isolation would not have caught that. It took someone who actually uses it asking a plain question about her own workflow.
Two more items came out of the same exchange, both new scope: a place for nursing to chart routine care, dressing changes, insulin, blood-sugar education, and a section for the durable medical equipment handed out at a visit. Neither exists in the form yet. Both are open, pending Catherine's input on the shape: free text, a checklist, or something in between, and whether it needs to record who did it and when.
One question in the original list was a mistake, written for the registration lead and sent to Catherine by accident. She caught it immediately and moved past it without losing any time.
As of this writing, her answers to the follow-up questions are still outstanding. Not a blocker. Just where the thread stands.
Two Formularies, One Version
Dr. Byrd has been reviewing the 96-drug sig worksheet built directly into the app since the last sprint. Independently, Dr. Suzanne Condron, already credited for the Pediatrics form, built something different: a standalone 72-medication reference PDF with her husband and AI assistance, alphabetized by English generic name, with Spanish translations, a color-and-shape dispensing key (self-administer, physician-revisable for chronic care and antibiotic stewardship, or health-department-only), and patient-facing Spanish instructions for topical medications. A companion document gives the pediatric referral criteria in English, translated from the Spanish original.
Two formularies built independently for the same mission is not a problem to route around. It is a decision to make on purpose, the same way this system got rebuilt once already rather than run two half-working versions side by side. The plan now is to compare both directly and land on one version everyone trusts, with Dr. Byrd folding the result into therapeutic classes, anti-infectives, anti-hypertensives, diabetes medications, and so on, rather than a flat alphabetical list.
The same thread brought in Dr. Carminia Davidsohn, Chief Medical Officer for the Philippines mission and a former member of the Epic implementation team at Kelsey-Seybold. She walked me through a full patient encounter from the clinical side back in April, and I promised to follow up once there was something real to show her. There is now a live sandbox instead of a description of one.
Two smaller items came out of the same email, neither about the formulary itself: whether the diagnosis field should become a real trackable list instead of free text, and physician review of a ten-category review-of-systems checklist that, as built, was drafted without a doctor's input at all.
What's Still Open
Catherine's remaining answers on Dental and Triage. The two formularies, reconciled into one. The nursing-charting and DME sections, designed once there is input on their shape. The extraction-marking convention, pending Linda. None of this is new to the project. It is the same paper-as-spec discipline that shaped every form before it, running now over email instead of across a table.