The Fragmentation Problem Nobody Talks About
Most people with a chronic condition or a complex medical history have the same mess: a shoebox of referral letters, a GP who knows about the cardiologist but not the osteopath, a medication list that hasn't been updated since the last hospital admission, and zero record of when that referral to Dr. Kempster was supposed to expire. The problem isn't that people don't care about their health records. The problem is that no single source of truth exists outside the hospital system — and the hospital system doesn't talk to itself, let alone to you.
A scattered personal health record fails at exactly the moments it matters most. Sitting in an ED waiting room at 11 PM, being asked which anticoagulant you're on and at what dosage — that's not the time to be scrolling through six months of SMS reminders from three different practices. The information needs to be structured, linked, and immediately retrievable. That's the only standard worth building to.
This Memento template is built around a relational model that mirrors how clinical encounters actually work: a visit has a date, a service type, a provider, one or more conditions, and one or more medications. Those entities are linked libraries, not repeated text fields. Change a provider's phone number once and it updates everywhere. Discontinue a medication and that flag persists in the record without deleting the dosage history.
How a Single Encounter Gets Captured
The core entry is a clinical event. The Date field captures the exact datetime of the visit — not just the day, which matters when you've had a cardioversion and a follow-up consultation on the same afternoon. Service Type is a checkbox field that can hold multiple values simultaneously: a GP appointment that also results in a prescription and a referral gets all three ticked. This isn't cosmetic — it drives filtering. When you need to pull every prescription event from the last 18 months for a medication review, that checkbox is your query anchor.
Service/Procedure goes deeper. The option list covers the full clinical range: ECG, TOE Cardiogram, Colonoscopy, Lung Function Tests, Ablation, Cardioversion, Care Plan Update. These aren't generic categories — they're the specific procedures that accumulate in a patient's history over years. Selecting "Cardioversion" on a record dated three years ago, then cross-referencing it against the Medication library entry for the flecainide that was prescribed the same week, gives you a longitudinal picture that no single clinical letter ever provides.
The Prescriptions and Prescription 2 fields are linked entries into the Medication library. Each medication record carries its own structure: drug name, strength, dosage frequency, the condition it treats, a boolean for whether it's currently being taken, and a discontinuation date if it's been stopped. Prescribed by links back to the Practitioner/Provider library. This means a single cardiovascular medication entry connects to the cardiologist who prescribed it, the condition it was prescribed for, and the exact date it was discontinued when the ablation made it redundant. That's a clinical narrative compressed into a database record.
What a Three-Year Record Actually Looks Like
After three years of consistent logging, this database stops being a journal and starts being a clinical decision support tool. Filter by Service Type: Major Surgery and you get every surgical event in chronological order, with associated conditions, medications prescribed at the time, and the referring provider. Filter by Continuing condition: true in the Condition library and you get the active problem list — not a static document that goes stale, but a live query against current flags.
The New Referral Required boolean on each visit entry is a small field that does a lot of work. At the end of any appointment, toggling it on creates a filter-visible flag. A weekly review of entries where New Referral Required = true AND Next Appointment = empty surfaces the gaps before they become a missed specialist visit. The Referral expires field in the Practitioner/Provider library closes the loop — when a referral to a specialist lapses, the date field tells you exactly when, and a new entry for the renewal consultation gets logged against the same provider record.
The OldCondition field, marked explicitly "Obsolete - Do not use," is left in because removing it would break historical entries. That kind of honest documentation of a schema evolution is worth preserving. It tells you the system was actually used, not just set up and abandoned.