Corrections
Corrections log
Last updated 3 August 2026
This page holds every correction HSREP has issued, with the date, what changed and where the change sits. It is kept separate from the rest of the site so that anyone checking a figure can find the record in one place.
How a correction is handled
Where HSREP wrote the figure, the text is corrected in place. Where it sits inside an article reproduced from the outlet that published it, the published wording stands and a dated note is attached to the piece. HSREP does not hold the right to rewrite another publisher’s copy, and a silent edit would leave no record of what changed.
The article PDFs reproduce each piece as it was published, so they still carry the original wording. Corrections live on the article pages and here.
Source check, 3 August 2026
Every figure, date and citation on this site was checked against its primary source. Ten items did not hold. Each is listed below with what changed and where the correction sits.
The stat panel said 1 in 3 public-health workers were considering leaving, and 20%+ vacancy in epidemiology and lab science. The first is the 2021 PH WINS figure (32%); the 2024 wave reports the measure the other way round, as 75% intending to stay. The second could not be matched to a primary source. Both now read from published 2024 data: 75% intending to stay (PH WINS 2024), and 44% more epidemiologists needed than jurisdictions employ (CSTE Epidemiology Capacity Assessment 2024). The explainer and the social card were re-rendered to match.
The figure was labeled cost of inaction from one major Dhaka earthquake. It is the sum of two components: more than $1B in transport-sector losses and $887M in water and wastewater damage. It covers lifeline infrastructure only, and excludes buildings, health facilities and economic disruption. The arithmetic was right. The label was not. The stat card, the accessibility description, the FAQ answer and the social card now state what the number covers and what it leaves out.
The article described a newly identified 400-kilometer fault line. That characterization has not appeared in peer-reviewed literature. The page now describes what was reported in late 2025, a roughly 400-kilometer structural trend running from the Mymensingh-Jamalpur area toward Kolkata, broadly following the long-recognized Calcutta-Pabna-Mymensingh Hinge Zone, and says plainly that it is not yet in the literature.
2.36% of GDP is Bangladesh’s total health expenditure for 2021, public and private combined, per the WHO Global Health Expenditure Database. The article describes it as spending by a government. Domestic general government health expenditure is well under one percent, which makes the article’s own argument stronger. A dated note sits on the piece.
The article says regulators proposed removing medical debt from credit reports in January 2025, and that a court struck it down six months later. The CFPB proposed the rule in June 2024 and finalized it on 7 January 2025. It never took effect. The Eastern District of Texas vacated it on 11 July 2025 in a consent judgment the bureau itself agreed to, not over its objection. A dated note sits on the piece.
The article attributes to the 2022 KFF Health Care Debt Survey a finding that about one in three adults, 35%, delayed or skipped care. That survey publishes no all-adults total for the question. What it publishes is a split by debt status: 64% of adults with medical debt put off needed care, against 28% without. The note records the split. The 64% quoted elsewhere in the article is correct.
The article gives more than 13,200 community clinics. No official source matches that number. The DGHS Annual Health Bulletin records 13,442 in operation and 13,079 reporting at the end of 2017, and WHO Bangladesh cited roughly 14,000 in 2019. No verified 2024-2026 count exists. The note says exactly that.
The page and its creatives said 61% of U.S. health professionals. The 61% is the share of engagement coming from health-domain industries on the best-performing post, not a share of professionals. The page, five social posts, three covers and both live site images now read health-domain industries.
The demonstration dashboard attached mortality language to Corley et al. (JAMA 2017;317(16):1631-1641). That study measures colorectal cancer incidence and stage at diagnosis, never death. The mortality finding belongs to Zorzi et al. (Gut 2022;71(3):561-567, hazard ratio 2.03), which now carries it. Source numbering was corrected to match.
A 52.7-57.9% range for Rajshahi was presented as screening completion. The source reports colposcopy positivity by division, which is a different measure. The range is removed. The national figure stands alone at 40.4% (Nessa A, et al. BMC Global and Public Health 2025;3:34, covering January 2018 to May 2023).
Found something that does not hold
Email contact@hsraep.org with the page and the figure. If it is wrong, it gets corrected and listed here. The standards this log sits under are set out in HSREP editorial standards.