V
Vancouver Ultrasound Finder

Methodology

How we build, maintain, and audit Vancouver Ultrasound Finder. Version 1.0 · published 2026-06-16 · next scheduled review 2026-09-15.

This page explains every decision we make about which clinics appear, how they are classified, how they are ranked, and how often the data is refreshed. We publish this in full because medical decision-support is a YMYL ("Your Money or Your Life") category — readers and search engines should be able to verify our process before trusting our output.

1. Clinic inclusion criteria

A Metro Vancouver ultrasound provider appears in our directory if it meets at least one of the following objective criteria:

As of the 2026-06-02 DAP snapshot, our directory holds 69 DAP-accredited clinics plus 9 Class C keepsake operators (78 total). Inclusion is automatic when a clinic meets the criteria — it is not contingent on payment, partnership, or any commercial relationship.

Exclusion is rare and always documented. A clinic may be excluded only for: (a) confirmed closure, (b) confirmed accreditation revocation, (c) duplicate listing where two records resolve to the same physical address and operating entity. Exclusion decisions are logged in our public changelog (see section 9).

2. Reimbursement classification

Each clinic receives one — and only one — reimbursement class. The class drives how we describe the clinic, who we recommend it to, and what fee expectations we set. Classification is determined by DAP scope + MSP enrollment + service marketing, in that order.

Edge case: clinics offering both classes. A small number of DAP-accredited Class B clinics (BreastCare Imaging, Lifespan Medical Services, Fast Track Ultrasound, etc.) also sell 3D/4D packages as a private-pay add-on. We classify these by their primary service profile — Class B — and tag the keepsake offering on the clinic detail page rather than reclassifying the entire facility as Class C.

3. Data sources and provenance

Every field on every clinic detail page is derived from a named public source. We do not republish broker-only data, scraped competitor databases, or unattributed claims.

We do not use: carrier-internal training material, broker-only quote engines, paid lead-aggregator databases, scraped competitor sites, or AI-generated clinic descriptions. Any inference we make from primary data is labeled inline as inference, not as fact.

4. Data pipeline and reconciliation

The end-to-end build pipeline for each refresh:

  1. DAP PDF ingest: parse the CPSBC PDF into a structured row per accredited facility (name, address, scope, expiry, license #)
  2. Google Places resolution: for each DAP row, query Places Text Search with name + address; resolve to a canonical place_id with name + address fuzzy-match threshold 0.92
  3. False-negative audit: any DAP row that fails to resolve goes into a manual review queue. Common causes: DAP-listed facility doing business under a different consumer-facing name, recently relocated, recently merged. Resolved manually with a citation in our internal data audit log (data/dap-false-negatives-and-exclusions.json).
  4. Class C enumeration: keepsake operators are added by direct search of consumer-facing marketing channels (Google Maps "3D ultrasound Vancouver" + Yelp + local parent-community references). Each is verified as a real operating business via phone or recent online activity before inclusion.
  5. Service taxonomy mapping: each clinic's stated services are mapped to our 9-service taxonomy. Mappings are explicit (e.g. "Vascular Doppler" → vascular_doppler) and edge cases are conservative (when in doubt, we exclude rather than over-claim).
  6. Schema validation gate: every record passes a Zod schema validation step at build time. Any record missing required fields fails the build and is fixed at source — we never silently drop or impute data.
  7. Conflict resolution: when DAP and Google Places disagree (e.g. different addresses), DAP wins for accreditation-related fields; Google Places wins for consumer-facing contact data. The conflict and resolution are logged.

5. Coverage Checker ranking algorithm

When a user completes the 6-step Coverage Checker, matching clinics are ranked by a deterministic, transparent sort. There is no opaque ML model.

Sort keys, in priority order:

  1. Service type match (the user's selected scan — echocardiography, obstetric, vascular, etc.)
  2. Reimbursement path match (MSP vs Private — derived from referral status + urgency + cost-priority answers)
  3. City match (preferred; falls back to Metro-wide if fewer than 3 in-city matches)
  4. Partner tier (Featured Partner → Verified Partner → Listed Only)
  5. DAP accreditation status (DAP-accredited preferred over non-DAP, for medical scans)
  6. Tie-breaker: Google rating × log(review count) — weights average rating by the volume of reviews backing it, so a 4.9★ with 500 reviews ranks above a 5.0★ with 3 reviews

Partner tier is the fourth sort key, not the first. A Listed Only clinic in the user's city with a matching service will always rank above a Featured Partner that doesn't match the user's needs. This is an intentional design constraint — we believe partner advantage in a transparent marketplace should only apply within a tie of equally-relevant choices, never to bend results away from the user's actual need.

Progressive relaxation: if zero clinics match all six criteria, the algorithm relaxes the lowest-priority constraint first (DAP preference, then partner tier, then city) and continues until at least one match is returned. Each relaxed criterion is visually flagged on the result card with a ⚠️ indicator so users see exactly why a non-perfect match was surfaced.

6. Geographic scope

"Metro Vancouver" in our directory means the 21 municipalities of the Metro Vancouver Regional District, plus the City of Abbotsford (Fraser Valley), which is commonly considered part of the regional commute shed for diagnostic imaging. Hospital-based services in adjacent communities (Squamish, Whistler) are referenced where relevant but not enumerated.

7. Refresh and verification cadence

Every clinic detail page displays a last verified date so readers can judge data freshness at a glance. Pages older than 180 days are flagged as stale — re-verification pending.

8. Dispute and appeal process

A clinic that disputes any factual claim on its detail page (classification, accreditation status, services, NAP data, language support, partner tier eligibility) may submit a correction:

  1. Email corrections@vancouverultrasound.ca from a domain verifiably matching the clinic (or with attached letterhead)
  2. Identify the specific claim disputed and the correct fact
  3. Provide a citable public source for the correction (CPSBC license letter, official clinic communication, etc.)
  4. We respond within 5 business days, publish the correction within 14 days, and link the correction from the original page

If we cannot agree on a fact, we publish both positions — the disputed claim with citation + the clinic's stated position — and let the reader judge.

9. Versioning and changelog

This methodology is versioned. Material changes — new sort key, new inclusion rule, new classification category, new data source — are published as a new version with an effective date and a public note. The version of this page (1.0) and its effective date (2026-06-16) are noted at the top.

10. What we explicitly do not do

11. Editorial team and credentials

Editorial direction by Roger Liu, SEO Editorial Director — responsible for clinic taxonomy, content standards, methodology versioning, and YMYL compliance. Engineering and data pipeline by the Vancouver Ultrasound Finder engineering team. All editorial decisions are independent of partner-program revenue (see Disclosure).

12. Open data commitment

The clinic directory we operate is built from public sources. We are committed to: (a) citing every source by name and URL, (b) flagging any inference clearly, (c) publishing corrections promptly, and (d) keeping this methodology page accurate to the system as it actually operates. If you find a gap between what this page says and what the site does, that's a bug — please email corrections@vancouverultrasound.ca.

Last updated: 2026-06-16 · Methodology version 1.0 · Next scheduled review: 2026-09-15 · Editorial content only; not medical advice; consult a licensed physician.