How we verify every clinical claim on spine.co. These rules apply to the Guide, condition pages, procedure pages, the second-opinion surface, and any page displaying medical statistics. They are enforced as launch gates — a page cannot be published until it passes.
Six operating principles
Anchor to published guidelines
Every treatment recommendation on the site — on condition pages, the Guide, and procedure detail pages — must reference a published society guideline (NASS, AAOS, ACP, NICE, ISASS) or a named clinical trial with a stable URL. Unsourced recommendations are not published.
Verify every PubMed ID against NCBI
Every PMID cited on the site is validated against the NCBI API to confirm (1) the identifier resolves, (2) the article title matches what we attributed to it, and (3) the article is actually about spine care. Hallucinated PMIDs — the kind that link to real but unrelated articles and escape simple HTTP 200 checks — are the single most common source integrity failure in medical content online. We run a validation script before every push that contains PubMed references.
No statistic without a citable source
Any numeric medical claim must be traceable to a specific source document with a URL or DOI. If the source cannot be located, the claim is either rewritten without the number or removed. Readers can click through every stat to the originating study or guideline.
Named reviewers only — never "medically reviewed"
When a page is reviewed by a clinician, the reviewer is identified by name, credentials, NPI (for US providers), and review date. We do not publish vague "medically reviewed" or "expert reviewed" attributions. Until a real, named, NPI-verified specialist signs off on a page, the page is transparently marked as awaiting clinical review.
Editorial independence from vendors and sponsors
Treatment recommendations are never influenced by device manufacturer, vendor, or pharmaceutical relationships. Providers who have disclosed financial relationships via CMS Open Payments are listed with their relationships shown on their profiles, but those relationships do not shape the treatment ladders published here. Vendor product pages link into the site but do not dictate what the site says about a procedure.
Re-verify on a published cadence
Society guidelines change. Trials publish new follow-up arms. Procedures gain or lose FDA clearance. Every condition page carries a visible "last updated" date. Treatment ladders are re-validated against current guidelines quarterly. Every cited URL is re-checked monthly by an automated script — broken or moved links get flagged for repair or replacement.
Pages on spine.co pass through four independent gates before reaching a reader:
1Source-integrity check. Every statistic is attributed, every PMID verified, every guideline URL resolved.
2Guideline alignment. Treatment ladders match the current version of the authoritative society guideline. Where guidelines differ, both are cited.
3Editorial review. An internal editor reads the page against a plain-language patient scenario and flags content that is technically correct but not decision-useful.
4Clinical review (optional but auditable). A board-certified spine specialist reads the page and signs off by name and NPI. Pages that have completed this step display the reviewer and review date at the footer. Pages that have not yet completed this step display a clear “Awaiting clinical review” banner — not a vague “Medically reviewed” disclaimer.
The reason for the two-state distinction — reviewed vs. awaiting — is simple: without named accountability, “medically reviewed” is a marketing phrase. We hold ourselves to the standard that every reviewer must be identifiable, verifiable, and reachable.
Are you a board-certified spine specialist?
We are building a named clinical review board across orthopedic spine surgery, neurosurgery, PM&R, pain management, and allied subspecialties. Contributing reviewers receive attribution on every page they sign and a dedicated reviewer profile page that supports their personal authority.
×Unsourced statistics. Every numeric medical claim ties to a named source with a URL.
דMedically reviewed” without a named reviewer. Either a named specialist signed off, or the page is transparently marked as awaiting review. No third option.
×Vendor-dictated treatment recommendations. Product pages can reference procedures, but treatment ladders are independent of any vendor relationship.
×Affiliate-link bias. The site does not contain affiliate links to devices, pharmaceuticals, or treatment centers. Provider and facility listings are free-tier by default.
×AI-generated clinical content without human oversight. AI may assist in drafting or summarizing, but every page is reviewed by a human editor. No clinical page ships without human signoff.