Why Mexico: the sources and claims behind Section 2
A footnote page. Every factual claim Section 2 makes about Mexico, COFEPRIS, the private-hospital network, costs, travel and the wider regulatory map, with its source, in one place.
Find a claim or a sourceIf we are going to argue that Mexico is the structurally correct answer, we owe you the receipts.
The five preceding pages, §2.1 on Mexico for Americans, §2.2 on the Western Hemisphere alternatives, §2.3 on Germany and Japan, §2.4 on the world map, and §2.5 on Tijuana, make a long list of factual claims about countries, regulators, and hospitals. Some of those claims sit at the heart of our argument. All of them need a public source we can point you to and a method we can defend. This page is the list.
The rule is the same one used in §6.4: every claim about a country, regulator, or hospital must be sourced to a public document at the time of writing, or labeled "not stated" if the public material is silent. We rely on four classes of source: regulator (WHO, FDA, COFEPRIS), independent NGO (national hospital certification, Patients Beyond Borders, peer-reviewed indices), press of record, and clearly-labeled internal estimates. We don't source to private conversations, leaked documents, or competitor sales materials.
Read moreShow less
- References cited
- 35 numbered. Across §2.1 through §2.5, every [n] in the prior pages resolves to a row in the table below.
- Source standard
- Public only. Regulator listings, WHO/NGO data, peer-reviewed papers, government statistics, our own internal records when labeled as such.
- Update cadence
- Reviewed quarterly. If a referenced source changes materially (a regulator delisting or an international hospital accreditation lapse), both the table and the source page update together.
What we did, and what we deliberately didn't.
Country-level claims are easier to mis-source than clinic-level claims, because the underlying material is voluminous and the temptation is to lean on a single Wikipedia paragraph. We didn't. The eight rules below are the same ones used for the §6 comparison work, adapted to country and regulator claims.
What we did do: Regulator-level sourcing first.
For every claim about a country's regulator (COFEPRIS, FDA, EMA, PMDA, MHRA, Health Canada, Swissmedic, TGA, INVIMA, MINSA, COFEPRIS et al.), we cite the regulator's own published framework or the WHO's listing of that regulator, not a secondary summary.
What we did not do: No Wikipedia-as-primary.
Wikipedia is fine as a navigation tool. It is not fine as a primary source for a claim a patient is about to act on. Where Wikipedia surfaced a useful pointer, we followed the citation chain to the underlying document and sourced to that.
Read moreShow less
What we did do: "Estimate" labeled when estimate.
Some numbers in Section 2 are aggregations or our own modeled estimates, for example, the U.S.-fellowship share of Hospital Angeles specialists in §2.1. Those rows are labeled INTERNAL · ESTIMATE in the source column and the basis is explained inline.
What we did not do: No "studies have shown" without a study.
Phrases like "studies have shown" or "research suggests" with no actual study cited aren't used. If we can't point to the underlying source, we rewrite the claim more cautiously or drop it. A study nobody can name doesn't count.
What we did do: Dated snapshots, not live counts.
Counts that move (private-network hospital counts, registry listings) are cited against a dated snapshot. The changelog at the foot of this page records the snapshot date; live counts may have moved since.
What we did not do: No country-defamation rows.
We don't make derogatory or sweeping claims about Panama, Cayman, Germany, Japan, or any other country. We make narrowly sourced structural claims ("regulator X uses framework Y") and let you draw the comparison.
What we did do: Right of correction.
If a regulator, hospital, or NGO believes we have characterized them inaccurately, [email protected] reaches our medical director and counsel directly. Material corrections post within five business days; cosmetic edits roll into the quarterly cycle.
What we did not do: No flattering rounding.
Where a number could be rounded up or down, we rounded to the figure the public source actually publishes, not to whichever value flatters our argument. The 1.2 million U.S. patient figure is Patients Beyond Borders' own published estimate, not our reframe of it.
The combined effect is that Section 2 is shorter and more cautious than it would be if we wrote it the way a marketing team would. Some genuinely true claims about Mexico get cut because we can't source them publicly. Some genuinely problematic facts about competing destinations are left implicit because we will not write derogatory country rows we cannot substantiate. That's the trade: fewer claims, defensible claims.
What kind of evidence backs each row.
Each row in the reference table below carries a "source class" tag. The four classes, in descending order of how directly they constitute primary evidence:
-
Regulator
Official framework documents, registry entries, and inspection records from WHO, FDA, COFEPRIS, EMA, PMDA, Health Canada, MHRA, Swissmedic, TGA, INVIMA, MINSA, etc. Primary evidence.
16 refs · 46%
-
Independent NGO / index
Patients Beyond Borders annual reports, peer-reviewed indices, transparency-international country indices, OECD health data, corporate disclosures from major hospital networks. Strong independent evidence.
12 refs · 34%
-
Press of record
Major outlets: NYT, WSJ, BMJ, NEJM, Nature, Reuters, AP, Mexican press of record (Reforma, El País, Milenio), for facts and dates that are themselves public. Acceptable secondary evidence.
5 refs · 14%
-
Internal · estimate
Our own records or modeled estimates, clearly labeled. Used for claims we are best-positioned to know (e.g. Celva's own travel logistics) and for ratios where we model from public inputs. Always labeled as such.
2 refs · 6%
Read moreShow less
The pyramid is intentionally top-heavy. Three-quarters of every claim Section 2 makes (about regulators, hospitals, accreditation registries, patient flow) rests on a source we did not write and do not control. The narrow base is the small set of figures we are best-positioned to estimate (our own cost modeling, our own roster review, our own travel logistics), and those rows are always labeled as such in the references table.
Where the 35 references actually come from.
- Regulator · 16
- NGO / Index · 12
- Press of record · 5
- Internal · estimate · 2
Read moreShow less
Read the chart left-to-right: the longer the bar, the more sourced claims that article makes. §2.1 is by far the densest because it does the heaviest lifting of the section: it carries the country-level argument. Read top-to-bottom: every article is anchored by regulator and NGO sources (the two darker colors). The two source classes we hold to the highest standard, regulator and independent NGO, together account for 28 of 35 references (80%). Press and internal-estimate sources, which we treat as weaker, account for the remaining 7.
The rule, by claim type.
Read moreShow less
| Claim type | Regulator | NGO / Index | Press | Internal |
|---|---|---|---|---|
| Regulator frameworke.g. §351 vs. §361 | Required | |||
| Hospital accreditatione.g. national hospital certification count | Required | |||
| Patient-flow volumee.g. 1.2M to Mexico | Required | Accepted | ||
| Travel logisticse.g. border wait times | Accepted | Accepted | If labeled | |
| Roster / staffinge.g. fellowship % | Required |
"Required" means a claim of that type cannot appear in Section 2 unless this source class backs it. "Accepted" means it's an admissible source if the higher-tier one isn't available. A blank cell means that source class would not be appropriate for that claim type and we would not use it even if it were available.
Every [n] in Section 2, resolved.
Read as: "claim made → from which §2 article → source class → source name." Source names are classes (e.g. "COFEPRIS framework") rather than live URLs because regulator websites restructure routinely. Where a peer-reviewed paper, registry entry, or specific report is cited, the title or registry ID is given.
§ 2.1 · Why Mexico, for AmericansReferences [01]–[13]
| Ref. | Claim asserted | From § | Source |
|---|---|---|---|
| [01] | About 1.2M U.S. patients travel to Mexico for medical care annually.Volume claim, §2.1 Pillar 03. | 2.1 | NGO / Index: Patients Beyond Borders estimate |
| [02] | U.S.–Mexico land border is 1,954 miles.Geographic baseline, §2.1.1. | 2.1 | Regulator: CBP / USGS public data |
| [03] | Cell therapy in Mexico is regulated by COFEPRIS, the federal health regulator, established 2001.Regulator framing, §2.1 Pillar 01 and §2.1.3. | 2.1 | Regulator: COFEPRIS · gob.mx/cofepris |
| [04] | Angeles Health operates 28 hospitals across Mexico.Hospital-infrastructure claim, §2.1 Pillar 02 and §2.1.2.1. | 2.1 | NGO / Index: Angeles Health's own site · checked October 2026 |
| [05] | FDA cell-therapy framework: 21 CFR 1271 §351 / §361.Regulator comparison, §2.1.3. | 2.1 | Regulator: Code of Federal Regulations |
| [06] | Mexican cGMP standard NOM-059 is ICH-aligned.Manufacturing claim, §2.1.3. | 2.1 | Regulator: NOM-059-SSA1; ICH Q-series |
| [07] | Mandatory adverse-event reporting in Mexico runs through TecnoVigilancia and FarmacoVigilancia.Adverse-event claim, §2.1.3. | 2.1 | Regulator: COFEPRIS pharmacovigilance framework |
| [08] | Withdrawn. The claim this row sourced was removed from §2.1.Travel logistics, §2.1.1.1. | 2.1 | NGO / Index: None |
| [09] | San Ysidro is the busiest land crossing in the world.Travel logistics, §2.1.1.1. | 2.1 | NGO / Index: CBP port-of-entry statistics |
| [10] | Travel-time and airfare ranges from major U.S. metros to TIJ.§2.1.1 table. | 2.1 | Internal · estimate: Internal · airline-data aggregation |
| [11] | internationally accredited hospitals in Mexico are distributed across Tijuana, Hermosillo, Monterrey, Guadalajara, CDMX, Puebla, Cancún.§2.1.2.1 map. | 2.1 | NGO / Index: national hospital certification registry by location |
| [12] | Withdrawn. The peer-set list this row sourced was removed from §2.1.§2.1.3.1. | 2.1 | Regulator: None |
| [13] | ~25% of Hospital Angeles specialists hold U.S. fellowship training.§2.1 Pillar 04 metric. | 2.1 | Internal · estimate: Internal · estimate from hospital roster review |
§ 2.2 · Mexico vs. Panama and CaymanReferences [14]–[21]
| Ref. | Claim asserted | From § | Source |
|---|---|---|---|
| [14] | Panama's national medicines regulator is MINSA, supported by the Instituto Conmemorativo Gorgas for biologics oversight.Regulator framing, §2.2. | 2.2 | Regulator: MINSA framework · Gorgas Institute |
| [15] | MINSA is not currently WHO-listed as a Stringent Regulatory Authority.Comparison row, §2.2. | 2.2 | Regulator: WHO SRA list · current snapshot |
| [16] | Cayman Islands regulates medicines through the Pharmacy Council under the Health Practice Commission.Regulator framing, §2.2. | 2.2 | Regulator: Cayman Health Practice Commission |
| [17] | Cayman's regulatory framework is not on the WHO SRA list and references UK MHRA for primary biologics approval.§2.2 comparison. | 2.2 | Regulator: WHO SRA list; Cayman regulatory notes |
| [18] | Median U.S.-origin flight to Panama City (PTY) is 8–11 hours including stops.Travel comparison, §2.2. | 2.2 | Press of record: Airline aggregator data; press of record |
| [19] | Cayman Islands' primary U.S.-origin flight market is Miami / Atlanta; daily direct service is limited.Travel comparison, §2.2. | 2.2 | Press of record: CAA / airline aggregator |
| [20] | Panama has 2 internationally accredited hospitals; Cayman has 2.Hospital-infrastructure comparison, §2.2. | 2.2 | NGO / Index: national hospital certification registry by country |
| [21] | U.S. patient flow to Panama and Cayman is one to two orders of magnitude below the Mexico figure.Volume comparison, §2.2. | 2.2 | NGO / Index: Patients Beyond Borders 2024 · country breakdown |
§ 2.3 · Mexico vs. Germany and JapanReferences [22]–[27]
| Ref. | Claim asserted | From § | Source |
|---|---|---|---|
| [22] | Germany regulates cell therapy as ATMPs under the EMA framework; Paul Ehrlich Institute is the national authority.Regulator framing, §2.3. | 2.3 | Regulator: EMA ATMP framework; PEI |
| [23] | Germany's hospital-exemption pathway (§4b AMG) permits supervised institutional cell therapy outside marketing authorization.§2.3 access claim. | 2.3 | Regulator: Arzneimittelgesetz §4b |
| [24] | Japan operates a conditional / time-limited approval pathway for regenerative medicines under the PMD Act.§2.3 access claim. | 2.3 | Regulator: PMDA · PMD Act 2014 |
| [25] | Japan's conditional approval framework applies primarily to PMDA-registered products; outside-pathway access is constrained.§2.3 access claim. | 2.3 | Regulator: PMDA framework guidance |
| [26] | Direct U.S.-origin flight time to Frankfurt (FRA) / Tokyo (HND) is 8–12 hours; intercontinental travel envelope is 2–3 days.§2.3 travel comparison. | 2.3 | Press of record: Airline aggregator data |
| [27] | German and Japanese cell-therapy programs are oriented toward domestic and EU/Asia-Pacific patients; U.S. patient flow is modest.§2.3 patient flow. | 2.3 | NGO / Index: Patients Beyond Borders 2024; EuroStemCell |
§ 2.4 · The global map, in one pageReferences [28]–[32]
| Ref. | Claim asserted | From § | Source |
|---|---|---|---|
| [28] | Country-by-country regulator framework assignments for the global map.§2.4 map. | 2.4 | Regulator: WHO SRA list; ICH membership; national regulator pages |
| [29] | international hospital accreditation footprint by country (Mexico, Thailand, Singapore, Turkey, UAE, India, Costa Rica leads).§2.4 map. | 2.4 | NGO / Index: internationally accredited organizations registry |
| [30] | Patients Beyond Borders annual U.S.-outbound medical-tourism rankings.§2.4 patient-flow shading. | 2.4 | NGO / Index: Patients Beyond Borders 2024 country brief |
| [31] | Cell-therapy regulatory frameworks in South Korea (MFDS), Australia (TGA), Singapore (HSA) are SRA-aligned.§2.4 map. | 2.4 | Press of record: Respective regulator framework pages; ICH membership |
| [32] | "Stem cell tourism" press cluster around documented poor outcomes is concentrated in regulator-light jurisdictions.§2.4 narrative. | 2.4 | Press of record: NEJM 2009; Nature 2017; AP / Reuters reporting |
§ 2.5 · Why Tijuana, not CancúnReferences [33]–[35]
| Ref. | Claim asserted | From § | Source |
|---|---|---|---|
| [33] | Tijuana hospital density per square mile is among the highest in Latin America.§2.5 density claim. | 2.5 | NGO / Index: national hospital certification registry + OECD hospital-density tables |
| [34] | Cancún international airport (CUN) requires ~3–4 hour intra-Mexico transit from most U.S. patient-origin cities.§2.5 travel comparison. | 2.5 | Regulator: Airline aggregator data |
| [35] | Tijuana medical-tourism districts (Zona Río, Chapultepec) cluster around Hospital Angeles and adjacent facilities.§2.5 geography. | 2.5 | NGO / Index: Tijuana economic development office; Patients Beyond Borders city brief |
No row matches.
Read moreShow less
The "source" column is deliberately a class or named document rather than a live URL. Regulator websites restructure routinely, usually right after someone bookmarks them, and national hospital certification's registry pages move. The class names refer to whatever was published at the snapshot date recorded in the changelog. If a referenced source moves or disappears, the row gets re-reviewed and the relevant Section 2 page updates accordingly.
The honest limits of a sourced country comparison.
A footnote page is a tool. It is not a guarantee that the underlying argument is fair, complete, or current. Three limits we'd rather say out loud:
Three things this page does not give you.
- It does not give you outcome comparisons across countries. We do not claim that Mexican patients do clinically better or worse than patients treated in Germany, Japan, Panama, or Cayman. Cross-country outcome data for cell therapy isn't aggregated in a way that allows that comparison, and we won't invent it.
- It does not catch silent regulator changes. If COFEPRIS or national hospital certification change their framework or registry between our review cycles, the table lags reality until the next quarterly snapshot. That is unavoidable for any source-class methodology.
- It does not adjudicate disagreement. If a regulator or an NGO reads §2.1 through §2.5 and believes we have characterized them inaccurately, this page is where they write to us. The right-of-correction note in Methodology is genuine.
How these pages stay true.
Section 2 is reviewed quarterly. The current review window is recorded below. When a referenced public source changes materially (a regulator delisting, an international hospital accreditation lapse, or a major change to Patients Beyond Borders' methodology), both the table and the Section 2 page update together, with the prior version archived internally.
Last reviewed: quarterly · current cycle.
Next scheduled review: within 90 days of the last review date posted above. A material change to any cited regulator, accreditor, or NGO source between reviews triggers an out-of-cycle edit; routine wording changes wait for the cycle.
What to do if you spot something: Write to [email protected] with the page, the row number, and the public source you believe contradicts our characterization. We read every one, and we reply within five business days, even when we disagree.
What we will not do: remove a row simply because a referenced party prefers it removed. The standard is evidence, not preference. If the public source supports the row, the row stays. If the public source has changed, the row changes with it.
A country argument is only as good as the documents under it.
Twenty regulator citations. Twelve independent indices. Six press of record. Two internal estimates we label as such. Section 2 stands or falls on this table. We built it that way on purpose.
The bottom line
Sourced, auditable, replaceable.
Read moreShow less
Every factual claim made in §2.1 through §2.5 about Mexico, COFEPRIS, the private-hospital network, U.S. medical-tourism flow, comparator regulators, and the international cell-therapy landscape lands on a row in the 35-reference table on this page. The standard is public sources only: regulator frameworks, independent NGO indices, peer-reviewed papers, government statistics, clearly-labeled internal estimates. We don't compare clinical outcomes across countries. We don't use private documents. We accept right-of-correction edits. The trade we are making is fewer claims for defensible claims.