Personalized medicine for a solo practice — where one physician runs a program to a research team’s standard
iksa is the harness for n‑of‑1 medicine. While you’re with another patient, your brain health program keeps following its markers and asking a new symptom its safety questions, then waits for you.
A gene sets the stakes, and a battery, a night and the labs say what you can act on
You counsel the patient before the genome goes off, read it yourself and sit with them when it says ε4, so each patient is a trial of one with a doctor who knows all of it. preDIVA’s multidomain program found no fewer dementia cases (The Lancet), and US POINTER moved cognitive scores with no dementia endpoint (JAMA). Two incretin trials didn’t slow early Alzheimer’s disease (The Lancet). The trial that did count less dementia treated uncontrolled hypertension (Nature Medicine), so your program ties each reading to something you’d do this year.
- APOE, from the genomeThe ACMG and NSGC guideline doesn’t recommend APOE testing to predict risk and allows it at a clinician’s discretion with counseling. One copy of ε4 raises the stakes of what’s treatable without changing the list, since US POINTER’s structured program helped ε4 carriers no differently from non-carriers. Once, after counseling before the test and the patient’s yes to knowing.
- Cognitive baselineBrief computerized batteries vary widely in validation, so you choose on that validation and stay on one. Practice can hide a fall: in the Mayo Clinic Study of Aging, people heading for MCI or dementia still gained on memory at their second visit. Every year, on alternate forms where the battery has them, rested and at the same hour.
- Home sleep studyIn the Wisconsin Sleep Cohort, an AHI of 15 or more went with poorer memory and word-fluency scores in ε4 carriers alone, though only cross-sectionally. Apnea with sleepiness earns treatment anyway, by CPAP or, with obesity, a GLP‑1 and GIP dual agonist labeled for moderate to severe apnea. The American Academy of Sleep Medicine sends a negative home test on to a lab study. Once, when snoring, sleepiness or witnessed pauses point to it, with CPAP use reviewed at four weeks.
- Labs, and p‑tau217 if you order itB12 and TSH for what’s reversible, ApoB and HbA1c for what’s vascular. A p‑tau217 with no symptoms is your clinical call: the Alzheimer’s Association’s guideline covers objective impairment in specialty care, and in the Mayo Clinic Study of Aging, kidney disease moved it about as much as brain amyloid did. Yearly, with blood pressure at every visit.
The gene, the low score and the night all reached one doctor, who treated the night first
In an ε4 carrier, processing speed at the 21st percentile invites one story, though it’s low average and not impairment. The night offered another, since apnea goes with slower speed and attention, so CPAP went in for the apnea itself.
Start with what your program does in the visit after a patient’s ε4 result, from what they hear first to what goes in your note.
Request a walkthroughRequest a walkthrough for a solo practiceA prevention study names someone for counseling, cover and safety, and one physician holds those standards with iksa carrying the record
Your panel is small on purpose, because your attention is what each patient pays for. No payer stands behind a patient who feels well, so the clinical, commercial and regulatory record comes with your program, and iksa holds their labs and your notes in one place for you to decide from.
Genetic counseling
The Alzheimer’s Prevention Initiative built eight components around telling a participant an APOE result. Your program carries that into one visit, with the patient’s long-term care insurance and their children’s risk both on the table.
Sub-investigator
A study names a sub-investigator who can step in having read the file. When you take a week off, the colleague you name needs that same read: your patient’s genome, and that ε4 came up.
Safety reporting
A mask the patient has stopped wearing, morning headaches, the word their wife noticed slipping: a study logs each on a form, while here each comes by message to the doctor who knows what it means for this patient.
Control arm
If processing speed climbs at the next battery, that could be CPAP, practice on the test or a low score drifting back toward the patient’s average. A study sorts the three with a control arm, and a trial of one reads them against what your note said to expect.
The list a nurse would work and the questions they’d take are yours today, and theirs the day you hire one
The harness goes on following the markers your plan names while you’re with someone else, and whatever it raises waits for your decision.
A year on, the patient asks what you told them about their children’s risk, and the note you wrote that day answers
Clinical Intelligence answers from the record and quotes the note behind each answer in your own words, so if you never wrote down what you said, no answer can quote it.
Clinical Intelligence: clinical AI and lab trackingThe flagged labs are yours to clear for now, and whoever writes in holding a consumer APOE result is yours to answer
You review and resolve what’s flagged on the panel’s list. Agentic CRM drafts that reply from your approved answers and stops where they stop, so the medical part waits for you to write it.
Agentic CRM: scheduling and follow-upYour patient hears about ε4 from you, since Patient AI turns on person by person, and then it takes their mask questions
Patient AI works from what you wrote about the fitting and the four-week review, and if they write about morning headaches, the safety questions come first and, when that needs a person, a card waits on your list.
Patient AI: the treatment companion
Whether treating apnea changes cognitive decline in ε4 carriers is unsettled, and small panels pooled together could help a study ask it
Each patient on your panel carries a genotype, a battery and a sleep study, all dated, and your note on what you did about them. De-identified, with each patient’s consent and an ethics review, your panel is a site a research group could pool with others. How small panels pool into one study →
Questions from the physician who runs the whole program
I have no nurse. Does this still work?
You’d work their list.The panel’s list, the flagged cards and the drafts come to you, and nothing Agentic CRM drafts leaves until you’ve approved or changed it, so the first read and the first draft come off your plate. Whoever you hire gets a login from us and that same list.
If a patient reports new confusion while I’m with someone, who sees it?
The patient’s told first, then you.Stroke signs or new confusion don’t wait for the rest of the safety questions: the patient is told to get emergency care now. A red card then sits on your list with what they reported, when it began and a suggested action, until you’ve resolved it.
When I take a week away, who watches my patients?
Whoever you name.We give the colleague you name a login for the week, which opens the same records and the same list, while drafts from Agentic CRM wait for you. Whatever never reached a note stays with you, the ε4 conversation included, so that’s the note to write before you leave.
Could Patient AI bring up a patient’s APOE before I have?
Not while it’s off for them.It reaches a patient only on their individual grant, which we turn on with you one patient at a time, and you can pause it for any one of them, so the ε4 conversation can come first. Once it’s on, it answers from their own record, their genetics included, and explains what your note says.
Do I need a genetic counselor before giving an APOE result?
Not if you’ve trained for it.The ACMG and NSGC guideline wants counseling from someone with the expertise, and a three-generation family history, which tells you whether PSEN1, PSEN2 or APP is the real question. iksa only reads the genotype from a supported vendor’s report, and what you covered, or where you referred the patient, goes in your note.
A patient brings a p‑tau217 from a wellness lab. Where does it go?
On file, with your reading.It goes in their chart and comes through with the rest of their record. Its meaning is contested: the Alzheimer’s Association’s revised criteria would call an abnormal result Alzheimer’s disease with no symptoms while recommending against testing such people outside research, and the International Working Group, in JAMA Neurology, would call them at risk. The Association’s blood-test guideline warns that many commercial tests miss its accuracy thresholds, so that reading is yours.
I have no office manager to run a migration from my EHR. How does this start?
With one patient’s folder.You send in one patient’s documents once your BAA is on record, and Clinical Intelligence reads them. You keep charting where you chart today.
What is n‑of‑1 medicine?All practicesiksa beside your EHRProtocol management softwareWhat iksa adds
Let us be the second reader of the brain health program you built, from the counseling visit to the yearly battery
Bring a protocol you run. We’ll walk one patient through it.