Data

Tools

Indexes

Applications

Experiments

Awesome

An open API service indexing awesome lists of open source software.

https://github.com/clay-good/sophiewell.com

A privacy-first, zero-dependency suite of offline-capable healthcare utilities for medical billing, clinical scoring, and field medicine, delivering fast, deterministic answers with no servers, no AI, and no data tracking.
https://github.com/clay-good/sophiewell.com

clinical-decision-support clinical-math deterministic ems-tools field-medicine hcpcs healthcare icd-10 medical-billing medical-calculators medical-coding medication-safety nursing-reference offline-first open-source-healthcare patient-tools privacy-focused pwa static-site zero-dependencies

Last synced: 13 days ago
JSON representation

A privacy-first, zero-dependency suite of offline-capable healthcare utilities for medical billing, clinical scoring, and field medicine, delivering fast, deterministic answers with no servers, no AI, and no data tracking.

Awesome Lists containing this project

README 

          
# sophiewell.com



Deterministic healthcare utilities, free forever, no servers, no accounts.



sophiewell.com is a single-page static website built for the **nurse on

shift** — primarily ICU and acute-care RN (CCRN / PCCN), then ED RN

(CEN), then floor / med-surg RN, then OR / PACU RN, then L&D / NICU RN —

with every relevant tile still served to doctors, pharmacists, RTs,

billers, coders, and EMS providers. It is, in posture, MDCalc with no

ads, no login, no upsell, no cookie banner, no email capture, and no

network call after first paint. Everything runs in the browser. There

is no AI of any kind. The data sources are public, bundled with the

page, and refreshed on a schedule. The site costs nothing to operate

beyond the domain renewal, so it can be free forever. The product

thesis, in one sentence: **MDCalc, but free, login-less, ad-less,

offline-capable, with a catalog tuned to the nurse on shift**

(see [docs/spec-v29.md](docs/spec-v29.md)). The one-line scope test:

a tile must consume at least one user input and produce a computed

output; "searchable lookup of static facts" does not qualify. See

[docs/spec-v10.md](docs/spec-v10.md) for the audience and

dependency-budget commitments and

[docs/spec-v29.md](docs/spec-v29.md) for the nurse-first pivot

and the v29 catalog ledger. At v52 close the catalog is 255

deterministic tiles — every one of them computes from at least

one user input (the new `pa-lint` tile in spec-v52 consumes

dropped files instead of form fields and produces a

deterministic findings report, the first instance of the

`shape: 'document-linter'` tile shape). Catalog-truth invariants

([docs/spec-v46.md](docs/spec-v46.md)) fail CI on any drift

between `UTILITIES.length` and the public marketing copy.

Sophie's eight posture commitments

([docs/spec-v50.md](docs/spec-v50.md)) — no ads, no login, no

telemetry, no third-party fetch, no AI, no cookies, no paid

tier, MIT-licensed forever — are listed at

[/commitments/](https://sophiewell.com/commitments/) and

enforced by automated checks on every commit. Select tiles

expose a collapsed "where does this come from?" derivation

block ([docs/spec-v48.md](docs/spec-v48.md)) with the verbatim

source formula, study cohort, and limits of validity.



## The problem



Bedside math — drug dose, drip rate, anion gap, eGFR adjustment,

sepsis-bundle clock, restraint re-check, Braden re-score —

lives in published clinical literature and institutional protocols.

The nurse on shift, the resident at 3 a.m., the medic in the

ambulance, and the pharmacist verifying a renal dose all reach

for the same calculators and arrive at the same number from the

same inputs. Existing options are paywalled, ad-laden, login-

gated, account-tied, or quietly telemetered — every one of which

adds friction to a 30-second decision at the bedside.



The meta-problem is that the workers who would benefit most from

fast, free, deterministic math are the ones least likely to have

a paid app handy at the moment they need it.



## The solution



Take the public datasets and the published clinical formulas, ship them to

the browser, do the lookups and math locally, and never phone home. A

single static page, a Content Security Policy that forbids outbound network

connections, and a service worker that caches the page for offline use.



## Quick start



Visit https://sophiewell.com. Save the page for offline use if desired. To

run locally, clone the repository and run `npm run dev`, which starts a

zero-dependency static server on http://localhost:4173 that applies the

production security headers. Any static file server will also work.



## How it works and how to use it



After the spec-v29 nurse-first prune (plus the spec-v30

thermal-emergency decision tiles, the spec-v31 Beers

deprescribing checker, the spec-v32 non-verbal pain scales, the

spec-v33 opioid-sedation + neonatal-pain extensions, the

spec-v34 pediatric-ICU bedside extensions, the spec-v35

SOS withdrawal companion, the spec-v36 MEOWS maternal

track-and-trigger, the spec-v37 CPSS + LAMS prehospital stroke

triage scales, the spec-v38 RACE prehospital LVO predictor, and

the spec-v39 ROSIER ED stroke-recognition scale, and the

spec-v40 GUSS post-stroke bedside dysphagia screen, and the

spec-v41 FOUR Score ICU coma scale, and the spec-v42 Katz ADL

discharge-planning functional-status index, and the spec-v43

Lawton IADL instrumental-ADL companion, and the spec-v44 Barthel

Index rehab-nursing weighted ADL, and the spec-v45 C-SSRS

bedside suicide-risk screener added on top), the

site organizes 255 deterministic calculators

across the bedside-shift surfaces a

nurse, doctor, pharmacist, RT, EMS provider, biller-coder, or

educator actually reaches for. Every tile takes at least one

user input and produces a computed output; searchable indexes

of static facts are explicitly out of scope (see

[docs/spec-v29.md §3](docs/spec-v29.md) and

[docs/spec-v10.md §2.3](docs/spec-v10.md)).



**Clinical math & conversions** covers BMI, BSA suite, MAP /

pulse pressure / shock index, anion gap with delta-delta,

corrected Ca / Na, osmolal gap, A-a gradient and P/F suite,

Winter's formula, eGFR suite (CKD-EPI 2021 / MDRD / Cockcroft-

Gault), FENa / FEUrea, maintenance fluids (4-2-1), QTc suite,

pregnancy dating, pack-years, the universal unit converter,

sodium-correction planner (Adrogue-Madias), free water deficit,

predicted body weight + ARDSnet tidal volume, and RSBI.

**Medication & infusion** covers drip rate, weight-based dose,

concentration-to-rate, pediatric dose bounds, insulin drip,

anticoagulant reversal, opioid MME (CDC 2022), steroid and

benzodiazepine equivalence, antibiotic renal-dose adjustment,

vasopressor dose-to-rate (with VIS / Wernovsky IS), TPN

macronutrient, iron deficit (Ganzoni), and the v29 nursing-

shift additions for insulin correction, electrolyte

replacement ladders, CRRT dose, ECMO titration, and the MTP

ratio tracker.

**Clinical scoring & risk** covers GCS, APGAR, NIHSS / mNIHSS,

Wells PE / DVT, CHA2DS2-VASc, HAS-BLED, TIMI, GRACE, HEART,

PERC, Geneva, CURB-65, PSI, qSOFA / SOFA, MELD-3.0 / Child-

Pugh, Ranson / BISAP, Centor / McIsaac, Caprini, Bishop,

Alvarado / PAS, PHQ-9, GAD-7, AUDIT-C, CAGE, EPDS, Mini-Cog,

CIWA-Ar, COWS, ASCVD PCE, PREVENT 2023, Light's criteria,

Mentzer index, SAAG, R-factor liver injury, KDIGO AKI, ICH

Score, Hunt-Hess / WFNS, plus modified Sgarbossa, revised

cardiac risk index, PEWS / NEWS2 / NEWS2-escalation, every

v17-v28 risk-score tile, and the v29 bedside nursing screens

(Braden, Morse, Hendrich II, RASS, BPS, CPOT, ICDSC, CAM,

CAM-ICU, Aldrete / PADSS).

**Clinical criteria & diagnostic bundles** packages NPIAP

pressure-injury staging, Norton / PUSH wound assessment, the

VIP / INS extravasation criteria, ABO / Rh blood-product

compatibility, and the v17-v28 diagnostic bundles.

**Workflow & templates** carries the patient-visit generators

that survive v29 (appeal letter, HIPAA Right of Access, HIPAA

authorization, ROI request, discharge instructions, specialty-

visit questions, the wallet-card generator, and the SBAR

handoff template), plus the v29 nursing-shift workflow timers:

restraint timer, sepsis-bundle clock, code-blue clock, device-

day counter, Bristol / abdominal-girth tracker, and the vent

SBT readiness / ARDSnet PEEP-FiO2 ladder.

**Field medicine** covers NEXUS / Canadian C-Spine, CDC Field

Triage, START / JumpSTART, peds-weight-dose, burn surface area

and fluid resuscitation (Parkland / modified Brooke),

pediatric ETT sizing, naloxone dosing, the EMS documentation

helper, and the AVPU / GCS quick reference.

**Public health & infectious disease** covers tetanus

prophylaxis, rabies PEP, bloodborne pathogen exposure, TB

testing interpretation, and STI screening intervals. **Billing

& coding** is now sparse — the time-based E/M code selector,

the NDC 10 ↔ 11 digit converter, and the HIPAA 60-day breach

clock. Every static index (ICD-10-CM, HCPCS, CPT, NDC, POS,

modifier, revenue, CARC / RARC, NUBC, DRG, APC, ICD-10-PCS,

RxNorm, NDC↔RxNorm) was retired in spec-v29 wave 29-2 §2.1;

use your EHR, CMS, FDA, or NUBC source instead.



The user flow is simple: type what you need into the hero search

("wells PE", "CHA2DS2-VASc", "ICD-10", "magnesium replacement") or

pick a tile from the disclosure-collapsible home grid, enter input,

read output. The hero consults a hand-curated synonym table

(`data/synonyms.json`, spec-v7 section 3.2) before falling back to

fuzzy matching, and shows a one-line breadcrumb explaining why a

tile is recommended. (The spec-v7 §3.1 patient-artifact dropzone was

retired in the 2026-05-18 clinical-staff-first pivot; the

deterministic classifier under `lib/artifact-*.js` is retained for

possible future reuse on a clinical-input surface, but the home view

no longer wears it.) Every utility opens with its

inputs pre-filled from a worked example so the empty state is never

empty (a "Reset to example" link restores them after editing), and

the References region at the bottom shows the formula citation or

dataset stamp the result came from (spec-v9). Calculator state is

encoded in the URL hash so any view can be bookmarked or shared as

a permalink.



All computation happens in the browser. For the full picture, see

[docs/architecture.md](docs/architecture.md).



## System design and architecture overview



The application is one HTML file, one CSS file, one JavaScript module set,

a service worker, and a data folder of sharded JSON. There is no backend.

The browser receives static files from the same origin and runs everything

locally. Data shards are loaded only when a utility that needs them is

opened. The service worker pre-caches the application shell on first load and

caches data shards on first access, keyed to the build hash so new

deployments invalidate old caches cleanly. The

application has zero runtime dependencies. A weekly CI job runs the data

refresh pipeline and opens a pull request with any updated data. For the

long version, see [docs/architecture.md](docs/architecture.md).



## Deterministic logic versus LLM usage



The product uses zero LLM inference and zero AI of any kind. All operations

are deterministic functions over public datasets and published formulas.

There is no model in the loop, no embedding, no inference call, no API

key. If a future sibling project explores AI-driven workflows, it will be

a separate, clearly labeled product. sophiewell.com itself never calls a

language model.



## Stability commitments



The site is stable and predictable by design. These commitments are hard

rules, not soft preferences.



- **No A/B testing, ever.** Every user sees the same version of every

  utility. A clinician who used a calculator on Monday and got one result,

  then got a slightly different result on Tuesday because they were in a

  treatment group, would correctly stop trusting the site.

- **No feature flags visible to users.** No "experimental" toggle, no

  "beta" feature. If something is on the site, it is for everyone.

- **No tracking.** No analytics script, no tracking pixel, no heatmap, no

  session replay, no error reporter, no third-party telemetry. The CSP

  enforces this with `connect-src 'self'`. The site never logs user-agent

  strings or IP addresses.

- **No notifications, no email capture.** The site never asks for an

  email address, never displays a notification permission request, never

  shows a "Sign up for updates" form. There is no newsletter and no

  account system, because there are no accounts.

- **Versioned releases.** Every release is tagged with a semantic version

  and described in [CHANGELOG.md](CHANGELOG.md), linked from the footer.



## CLI reference



| Command                  | Description                                                       |

|--------------------------|-------------------------------------------------------------------|

| `npm run dev`            | Serve the directory locally on http://localhost:4173              |

| `npm run build`          | Copy static files into `dist/` for deployment                     |

| `npm test`               | Run the full test suite (unit, a11y, grep, data integrity)        |

| `npm run test:unit`      | Run Node's built-in unit tests (1,782 tests)                      |

| `npm run test:e2e`       | Run Playwright integration tests against a real browser           |

| `npm run test:a11y`      | Run accessibility checks on every utility view                    |

| `npm run lint`           | Run ESLint with the project rules (bans innerHTML, eval, others)  |

| `npm run data:refresh`   | Re-fetch and re-shard every public dataset                        |

| `npm run data:verify`    | Verify shard SHA-256 hashes against the manifests                 |

| `npm run sbom`           | Regenerate the CycloneDX SBOM (`sbom.json`, `sbom.md`)            |

| `npm run release:check`  | One-shot pre-release gate: lint + test + sbom + build             |

| `npm run clean`          | Remove `dist/` and other build artifacts                          |



## Safety guarantees



- The application makes no outbound network requests at runtime.

- The application does not store user input anywhere.

- The application is read-only with respect to all bundled data.

- The application does not write to any storage location other than the

  service worker's own cache of its own static files.

- Clinical input is processed in memory and discarded when the page

  is closed.

- There is no `localStorage`, no `sessionStorage`, no cookies, and no

  IndexedDB. All four are verified empty by the integration test suite.

- `innerHTML`, `outerHTML`, `insertAdjacentHTML`, `eval`, and the

  `Function` constructor are banned by the ESLint config and a grep

  check; the `el()` DOM helper throws on any attempt to set raw HTML.

- A CycloneDX SBOM (`sbom.json`, `sbom.md`) is regenerated on every

  build and ships with the site, hashing every runtime asset and source

  module with SHA-256 plus a per-build buildId.



## Limitations



- The Field Medicine layer reproduces only the *numeric facts* (drug

  doses, intervals, energy levels, weight ranges) from AHA ACLS/PALS/BLS

  guidelines, with attribution to the AHA guideline edition. AHA

  algorithm flowcharts are *not* reproduced. The Broselow-Luten

  color-band system is *not* bundled (licensed by Vital Signs); the

  pediatric dose calculator works in straight kilograms instead. See

  [docs/field-medicine-citations.md](docs/field-medicine-citations.md).

- Static code indexes (ICD-10-CM, HCPCS, CPT, NDC, POS, modifier,

  revenue, CARC / RARC, NUBC, DRG, APC, ICD-10-PCS, RxNorm,

  NDC↔RxNorm) and reference tables (adult / pediatric lab ranges,

  TDM, tox levels, ISMP high-alert wallet, AHA CPR wallet card,

  NIOSH Pocket Guide, DOT ERG, AGS Beers Criteria, ASA Physical

  Status, Mallampati, Modified Rankin) were retired in spec-v29

  wave 29-2 — Sophie's edge is computation, not indexing. Use your

  EHR, the upstream source, or your institutional protocol.

- The Appointment Prep Question Generator uses deterministic keyword

  matching against a hand-curated bank, not language understanding.

- The site is not medical, legal, or financial advice. It does not

  replace clinician judgment, institutional protocols, professional

  billing review, or legal counsel.

- Clinical calculators are math aids only. Institutional protocols

  govern any clinical decision. Field-medicine utilities additionally

  defer to local protocols and online medical direction.

## Security



Vulnerability reports: see [SECURITY.md](SECURITY.md) for the private

disclosure channel and the threat model summary. The CSP, security

headers, and supply-chain posture (pinned dev deps, SBOM on every

build, integrity-verified data shards) are documented in

[docs/threat-model.md](docs/threat-model.md).



## Documentation



- [docs/spec-v10.md](docs/spec-v10.md) — current positioning spec:

  clinical-first audience, runtime-dependency budget, permanent

  out-of-scope list

- [docs/spec-v11.md](docs/spec-v11.md) — correctness-floor spec:

  per-tile audit protocol, specialty-named groups, optional

  source-quoted `interpretation` field

- [docs/scope-mdcalc-parity.md](docs/scope-mdcalc-parity.md) —

  long-horizon scope: every actionable clinical calculator a

  healthcare worker would otherwise reach for MDCalc to find,

  shipped slowly at the v11 quality bar

- [docs/architecture.md](docs/architecture.md) — runtime architecture,

  data flow, no-backend rationale

- [docs/data-sources.md](docs/data-sources.md) — every bundled dataset

  with canonical URL and refresh cadence

- [docs/clinical-citations.md](docs/clinical-citations.md) — every

  formula and scoring system with citations

- [docs/field-medicine-citations.md](docs/field-medicine-citations.md) —

  Group I citations, including AHA non-derivation posture

- [docs/legal.md](docs/legal.md) — data sourcing posture, AMA CPT

  handling, attributions

- [docs/accessibility.md](docs/accessibility.md) — WCAG 2.2 AA checklist

- [docs/threat-model.md](docs/threat-model.md) — threats considered and

  the controls that mitigate each

- [docs/stability.md](docs/stability.md) — full stability policy

- [docs/performance.md](docs/performance.md) — performance budget and

  measurement methodology

- [docs/operations.md](docs/operations.md) — data refresh workflow,

  manifest format, integrity verification

- [docs/release.md](docs/release.md) — Cloudflare Pages release runbook

- [docs/deployment.md](docs/deployment.md) — deployment configuration

- [CHANGELOG.md](CHANGELOG.md) — every release with date, version, and

  user-visible changes

- [SECURITY.md](SECURITY.md) — vulnerability disclosure policy