What North State Prevention is built on

Every reading is classified in src/lib/htn-rules.ts against the ACC/AHA 2017 bands and drives what the patient sees, whether an alert is created, and whether a 5-minute recheck is required.

Blood pressure categories follow the American Heart Association, Understanding Blood Pressure Readings.

Chronic titration decisions are never made on a single reading: the engine uses the mean home BP over a rolling 14-day window and advances every 2 weeks per the UC Way protocol. More readings and more distinct days improve confidence, but the 2-week interval is the primary gate; patients with sparse data are flagged low-confidence and routed to adherence outreach. Only acute severe readings (≥ 180/120) act on a single measurement.

Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults. Hypertension. 2018;71(6):e13-e115.

Shimbo D, Artinian NT, Basile JN, et al. Self-Measured Blood Pressure Monitoring at Home: A Joint Policy Statement From the American Heart Association and American Medical Association. Circulation. 2020;142(4):e42-e63.

The med engine in src/lib/titration.ts follows the UC Way HTN medication algorithm reviewed by UC Davis. It starts with asingle-pill combo (losartan/HCTZ) rather than titrating two separate scripts, then advances every 2 weeks based on home BP monitoring. We default to ARB + thiazide instead of ACEi + thiazide to avoid the ~10% cough rate that quietly drives non-adherence; lisinopril remains an equivalent substitute. Prioritizing ARB/ACEi + thiazide also means a single BMP covers safety monitoring for both drug classes, keeping lab overhead low for uninsured patients.

1. Step 1 (Start) — Losartan/HCTZ 50/12.5 single-pill combo
2. Step 2 — Losartan/HCTZ 100/25 single-pill combo
3. Step 3a → 3b → 3c — Add Amlodipine 2.5 mg, then 5 mg, then 10 mg every 14 d if still above goal
4. Step 4a → 4b — Add Spironolactone 12.5 mg, then 25 mg — only when the patient is on a thiazide, eGFR ≥60, and K+ <4.5
5. Step 5a → 5b → 5c — Metoprolol succinate XL 25 mg, then 50 mg, then 100 mg, each rung requiring a resting pulse ≥55 bpm (also the direct next rung when spironolactone criteria are not met)
6. Step 6 — Refer to specialist (resistant HTN workup)

A 14-day gate prevents same-visit re-titration, and the engine blocks advancement when K+ > 5.0, Cr rises > 30%, eGFR drops below step-specific thresholds, or pulse is <55 bpm on a rate-lowering agent.

UC Way Hypertension Medication Algorithm. University of California Office of the President. BMJ Open 2026; in press / internal protocol document.

Titration decisions are cross-referenced against KDIGO's eGFR × albuminuria matrix. Patients in the "orange" or "red" cells (e.g., eGFR 30–59 with UACR > 300) are held from thiazide advancement in favor of a loop diuretic pathway and get earlier lab follow-up.

Kidney Disease: Improving Global Outcomes (KDIGO). 2024 Clinical Practice Guideline for the Evaluation and Management of CKD. Kidney Int. 2024;105(4S):S117-S314.

The program runs on a single consolidated BMP + UACR panel so uninsured patients get one lab visit instead of several. Cadence, thresholds, and blockers below are the same rules the titration engine in src/lib/titration.ts and the demo data generator enforce at runtime.

When labs are drawn

1. Baseline — 7 days after enrollment, for every patient.
2. Post-titration — 1–2 weeks after any Start, Step 2, or Step 4a/4b change on a RAAS or diuretic.
3. Closer lab monitoring (every 90 d) — Step 4a/4b regimens, CKD orange/red, or K⁺ borderline on last panel.
4. Ongoing standard (every 180 d) — stable Step 2–3c on RAAS + thiazide.
5. Not required — lifestyle-only or amlodipine monotherapy.

Engagement modulates cadence: sporadic patients drift ~30 days late; lapsed patients freeze at their last real draw until they re-engage.

What we measure and the ranges we use

What abnormal labs do to the engine

1. K⁺ > 5.5 — hard hold on any RAAS advance; clinician review required.
2. K⁺ 5.0–5.5 on ACEi/ARB — block spironolactone add; recheck BMP.
3. Spironolactone eligibility — on a thiazide, eGFR ≥60, and K+ <4.5 all required; failing any one criterion routes to the metoprolol XL rungs instead.
4. Cr > 1.5 on ACEi/ARB — hard hold; recheck renal function before advancing.
5. eGFR < 30 on HCTZ — swap thiazide → loop diuretic (furosemide).
6. KDIGO orange/red — CKD carve-out: mono-RAAS at Start; chlorthalidone or furosemide instead of HCTZ downstream.

Which advances require a fresh BMP first

1. Start (from no meds) — no prior BMP required; follow-up BMP at 1–2 wk.
2. Advance to Step 2 or Step 4a/4b — BMP within the prior 14 days required; the titration button blocks until one is on file.
3. Advance to Step 3a–3c or 5a–5c — no fresh BMP gate (amlodipine and metoprolol have no electrolyte effect).

When a reading is Stage 1 or Stage 2, the patient's log flow surfaces the full MI guide — importance/confidence rulers, change-talk elicitation, barriers, and a concrete commitment. Follow-up readings in the same cycle switch to a lighter MI check-in so patients aren't asked the same questions repeatedly.

See every MI screen live → · Full patient decision tree → · Patient-facing protocols →

Miller WR, Rollnick S. Motivational Interviewing: Helping People Change. 3rd ed. Guilford Press; 2013.

Because the program is not staffed 24/7 and targets uninsured patients, the triage logic is intentionally conservative on severe readings:

1. ≥ 180/120 → ask the patient to rest 5 min and re-check.
2. If the recheck is still ≥ 180/120, ask the nine red-flag questions.
3. ≥ 160/100 also triggers the same symptom screen — patients with prior stroke or vascular disease can develop stroke-like symptoms at this level.
4. Any positive symptom → 911 / ED.
5. Confirmed severe but asymptomatic → urgent, time-bound in-person evaluation (not "call your clinic when you can"), so a very high pressure is not left to drift.

The program manages chronic hypertension. Hypertensive urgencies and emergencies are explicitly routed out of the platform to the correct level of care.

North State Prevention targets uninsured and under-insured patients on a low monthly subscription. Sustainability requires that infrastructure + device cost per patient stay below the subscription price. Device tiers meet patients where their tech comfort is:

See the public device tier one-pager for the full cost comparison.

North State Prevention is UC Way-aligned with a KDIGO overlay. That means we depart from strict ACC/AHA 2017 guidance in a handful of specific, deliberate ways. Each deviation is listed here with the reasoning so reviewers, partner clinicians, and future auditors can see the tradeoff we chose.

1. First-line in Black adults without CKD or heart failure. ACC/AHA prefers thiazide or CCB monotherapy in this subgroup. We start losartan/HCTZ, so the thiazide half is aligned but the ARB half is off-guideline as monotherapy for this subgroup. Rationale: single-pill adherence and program simplicity. The clinician can swap to amlodipine/HCTZ at their discretion.
2. Combo pill from Step 1. ACC/AHA reserves two-drug initiation for Stage 2 or BP ≥ 20/10 above goal. Rationale: the UC Way algorithm and real-world combo-first programs show that population-level adherence gains outweigh the mild over-treatment risk for our Stage 1 cohort.
3. Stage 1 treatment threshold. ACC/AHA gates medication start on ASCVD 10-yr risk ≥ 10% (or diabetes, CKD, or clinical CVD). We treat all Stage 1 cycles that fail lifestyle change. Rationale: our target population often lacks recent lipids, so an ASCVD gate would delay enrollment. Planned for v2 once the lab partnership is live.
4. ARB preferred over ACEi. ACC/AHA treats them equivalently. Rationale: cough-driven non-adherence is a bigger clinical problem than the small cost delta in our formulary. Lisinopril remains a recognized substitute in the engine.
5. Severe asymptomatic → clinic callback. ACC/AHA recommends outpatient follow-up within 24–48 hours. The app does not instruct the patient; it creates an urgent item in the clinic queue carrying the coded alias, timestamp and reading, with a 60-minute deadline for staff to acknowledge and call. Unacknowledged items escalate to the clinic's backup staff contacts. Rationale: the clinic holds the alias-to-patient crosswalk and is the right party to make clinical contact. Every step is logged as a callback audit trail.

Aligned with ACC/AHA: <130/80 target, cycle-based diagnosis confirmation, KDIGO overlay for CKD, and 911/ED routing for symptomatic severe hypertension.