How factual errors are identified, reviewed, corrected, and noted across the Virus Questions network — and every correction made so far.
The log
A corrections policy is a claim about behaviour. This is the record behind it:
38 entries covering every factual change made to live content across the network,
each traceable to the commit that made it. It includes the entries that are least
flattering — 1 case where a correction was itself wrong and had to be
undone, and 1 where a reported error was investigated and found not to
be an error at all.
What was published: The site stated in several places that no hantavirus vaccine is approved for human use anywhere in the world, and elsewhere described the Korean and Chinese products only as having "mixed effectiveness data."
What it says now: Inactivated whole-virus vaccines against Hantaan or Seoul virus are licensed for human use in the Republic of Korea and in China; published reviews describe their protective efficacy as uncertain (Liu et al., Front Microbiol 2020, doi:10.3389/fmicb.2019.02989; Chai et al., Vaccines 2025, doi:10.3390/vaccines13020198). Every instance now states the accurate position: no hantavirus vaccine is licensed in the United States or Europe, and none exists anywhere for Sin Nombre or Andes virus. The instruction block in the site's automated monthly-review script was carrying the same wrong assertion and was updated in the same change.
Corrected
Environmental survival times for hantavirus (transmission and FAQ pages)
hantavirusquestions.com
What was published: A table gave survival times by environment — days to weeks indoors, hours in direct sunlight, minutes after disinfection — with no source, and the FAQ repeated them.
What it says now: Replaced with the published laboratory measurements and their stated conditions: Puumala virus infectious in soiled rodent bedding for 12-15 days at room temperature, 5-11 days in cell-culture fluid at room temperature and up to 18 days at 4C, inactivated within 24 hours at 37C (Kallio et al., J Gen Virol 2006, doi:10.1099/vir.0.81643-0); Hantaan virus infectious for 96 days in wet suspension at 4C and sharply reduced by drying (Hardestam et al., Appl Environ Microbiol 2007, doi:10.1128/AEM.02869-06). No equivalent study exists for Sin Nombre virus, and the pages now say so.
Corrected
Americas case figures and U.S. state counts (outbreaks page)
hantavirusquestions.com
What was published: Country rows for the Americas gave approximate ranges with no source (for example "about 100 to 150" cases for Argentina), and the U.S. table gave rounded state figures against a cumulative total described as "over 850."
What it says now: Replaced with PAHO's published 2025 figures through epidemiological week 47 (229 cases and 59 deaths across eight countries, 25.7% regional case-fatality rate) and with our tabulation of CDC's patient-level case file (890 cases, 309 deaths, state-level counts and rates). CDC's weekly NNDSS tables were added for 2024-2026, which the consumer page does not cover.
Corrected
El Nino research citation (outbreaks page)
hantavirusquestions.com
What was published: The link between El Nino events and subsequent hantavirus case surges was attributed to "research published in the journal Science," with no citation.
What it says now: Attributed to Hjelle B and Glass GE, J Infect Dis 2000;181(5):1569-1573 (doi:10.1086/315467), which reports a five-fold rise above baseline in Four Corners cases in 1998-99 following the 1997-98 El Nino. We could not locate a paper in Science supporting the original attribution.
Corrected
Respirator recommendation attributed to OSHA (prevention page)
hantavirusquestions.com
What was published: The occupational-safety section stated that OSHA recommends employer-provided N95 respirators or better for agricultural, forestry and construction workers in endemic areas.
What it says now: We could not verify that recommendation in an OSHA document. Replaced with CDC's own published guidance for occupational rodent clean-up: coveralls, boots or shoe covers, gloves, goggles and a HEPA-filtered half-mask or powered air-purifying respirator, with the state pulmonary-function and fit-testing requirements that apply before respirator work begins.
What was published: The page stated that "published clinical guidance recommends follow-up renal function monitoring for at least one year post-hospitalization for severe HFRS cases."
What it says now: No CDC, WHO or ECDC guideline making that recommendation could be located. The sentence was removed rather than left standing on an unverifiable attribution, and replaced with CDC's published statement that complete recovery from HFRS can take several weeks to months.
Corrected
Revenue disclosure (How We Research page)
hantavirusquestions.com
What was published: The page stated that the site "generates revenue through two disclosed mechanisms" and described Amazon Associates commissions and Google AdSense advertising as live.
What it says now: Both are switched off network-wide and have been since before September 2026; the About page already said so and this page had not been updated. It now states that the site earns nothing, that both mechanisms are registered but inactive, and that any change will be published on the day it happens.
Corrected
Zabdeno/Mvabea vaccine status (treatment, prevention, outbreak, about and home pages)
ebolaquestions.com
What was published: The site described Zabdeno and Mvabea (Janssen) as a currently approved Ebola vaccine regimen alongside Ervebo — "Two vaccine products are currently approved."
What it says now: The European Medicines Agency records that the European Commission withdrew both marketing authorisations on May 1, 2026 at the request of Janssen-Cilag, "for commercial reasons," and that the product "had not been marketed in the EU." Ervebo is the only licensed Ebola vaccine. WHO's fact sheet, updated May 2026, still lists the regimen as approved; the pages now state the EMA register's position and note the disagreement.
Source: European Medicines Agency, Zabdeno and Mvabea EPAR pages (withdrawal of marketing authorisation, 01/05/2026); WHO Ebola disease fact sheet, May 2026
Corrected
Second U.S. citizen evacuated to Germany (Ebola in the U.S., 2026 outbreak and home pages)
ebolaquestions.com
What was published: The pages said one U.S. aid worker who tested positive in DRC had been medically evacuated to Germany.
What it says now: WHO reported a second U.S. citizen — a humanitarian worker — medically evacuated from DRC to Germany on July 13, 2026, before the pages' previous review date. Both were treated in Germany, both recovered, and both are counted among DRC's cases.
Source: WHO Disease Outbreak News 603 (May 22, 2026), 613 (July 17, 2026) and 617 (September 10, 2026)
Corrected
CDC travel notice level (prevention and travel-risk pages, related-topic cards)
ebolaquestions.com
What was published: The prevention page said CDC "has issued a Level 3 advisory" for the outbreak; the travel page's schema description and FAQ described a Level 3 warning for Ituri Province recommending avoiding nonessential travel; several related-topic cards repeated it.
What it says now: CDC's notice for Ituri and North Kivu provinces is Level 4, "avoid all travel"; Level 3 applies to Haut-Uélé and Tshopo; Level 2 to the rest of DRC and Uganda. The pages had not been updated when CDC raised the level.
Source: CDC Travel Health Notices and CDC Ebola Outbreak: Current Situation (page dated September 14, 2026)
Corrected
Bundibugyo candidate vaccine described as "preclinical" (prevention, treatment, research and 2026 outbreak pages)
ebolaquestions.com
What was published: Four pages said that "the first Bundibugyo-species-specific candidate vaccine entered preclinical testing" in July 2026, in the UK and Canada, under the WHO R&D Blueprint.
What it says now: The statement could not be verified against any WHO report or registry. ClinicalTrials.gov records NCT07737717, a Phase 1 placebo-controlled human trial of Moderna's mRNA-1469 in 84 adults at three Canadian sites, begun July 31, 2026 — a later stage than preclinical and a different claim. The pages now describe the registered trial.
Source: ClinicalTrials.gov NCT07737717 (record updated September 10, 2026); WHO Disease Outbreak News 615–617
Corrected
Fever threshold for early symptoms (symptoms page)
ebolaquestions.com
What was published: The dry-phase symptom list and FAQ gave the fever threshold as "≥38.6°C / 101.5°F", a figure from a 2014-era clinical case definition.
What it says now: CDC's current guidance for travelers returning from the outbreak area uses 38°C (100.4°F). A reader with a fever between the two values would have been told, in effect, to wait. The page now gives CDC's threshold.
Source: CDC, Information for Travelers Returning from Ebola-Affected Areas (September 11, 2026)
Corrected
PALM comparison-arm figure and Ervebo confidence interval (treatment page)
ebolaquestions.com
What was published: The August 7, 2026 correction gave the PALM trial's ZMapp comparison arm as "≈49.4%", a value that matches neither of the two comparisons the paper reports; the Ervebo card gave the ring-trial confidence interval as "63.5–100%" against "23 cases in controls".
What it says now: PALM (Mulangu et al., NEJM 2019): mAb114 61/174 (35.1%) vs ZMapp 84/169 (49.7%), P=0.007; REGN-EB3 52/155 (33.5%) vs the concurrent ZMapp subgroup 79/154 (51.3%), P=0.002. Ervebo: 0 vs 16 cases in the randomised rings, efficacy 100% (95% CI 68.9–100); 0 vs 23 across all 117 rings (95% CI 79.3–100).
Source: Mulangu S, et al. N Engl J Med 2019;381:2293, doi:10.1056/NEJMoa1910993; Henao-Restrepo AM, et al. Lancet 2017;389:505, doi:10.1016/S0140-6736(16)32621-6
Corrected
Duration of Ebola virus persistence in semen (transmission page)
ebolaquestions.com
What was published: The page said the virus "can persist in semen for up to 12 months after recovery," conflating WHO's 12-month safer-sex period with the measured duration of persistence.
What it says now: Viral RNA was detected in semen of 4% of men tested 16–18 months after discharge (Deen et al.), and WHO records sexual transmission documented up to fifteen months after recovery; the cohort data are now given month by month.
Source: Deen GF, et al. N Engl J Med 2017;377:1428, doi:10.1056/NEJMoa1511410; Thorson AE, et al. PLoS Med 2021, doi:10.1371/journal.pmed.1003273; WHO Ebola disease fact sheet, May 2026
Corrected
Wrong and non-existent citations (research page)
ebolaquestions.com
What was published: The 1977 Pattyn paper carried PubMed ID 985414, which resolves to an unrelated biochemistry article; the Urbanowicz, Bah, Uyeki, Henao-Restrepo and Gire entries linked to the wrong articles; and a citation to "Tully CM, et al., NEJM 2021" for the Janssen vaccine regimen does not correspond to any published paper.
What it says now: Every citation on the page was re-verified against PubMed and now carries its correct DOI and PubMed ID (Pattyn 1977 is PMID 65663). The non-existent citation was replaced with the regimen's actual phase 2 report, Pollard AJ, et al., Lancet Infect Dis 2021, doi:10.1016/S1473-3099(20)30476-X.
Source: PubMed E-utilities esummary/efetch, September 14, 2026
Corrected
World Bank GDP-loss figure misattributed (economic impact page)
ebolaquestions.com
What was published: The page attributed a "$2.2 billion combined" two-year GDP loss for Guinea, Liberia and Sierra Leone to a January 2015 World Bank update.
What it says now: The $2.2 billion figure is the World Bank's October 8, 2014 "Low Ebola" scenario estimate for West Africa as a whole for 2014 alone; the same scenario adds $1.6 billion for 2015, and the "High Ebola" scenario totals $32.6 billion. The page now gives the press release's own figures and Huber et al.'s $2.8–32.6 billion range.
Source: World Bank press release, October 8, 2014; Huber C, Finelli L, Stevens W. J Infect Dis 2018;218(suppl 5):S698, doi:10.1093/infdis/jiy213
Corrected
World Bank GDP-loss figure misattributed (pandemic economic impact page)
virusquestions.com
What was published: The Ebola card described a "$2.2B cumulative GDP loss" for Guinea, Liberia and Sierra Leone through 2015 — the same misreading corrected on the Ebola spoke on September 14, 2026, carried on the hub unchanged.
What it says now: The $2.2 billion figure is the World Bank's October 8, 2014 "Low Ebola" scenario estimate for West Africa as a whole for 2014 alone; the same scenario adds $1.6 billion for 2015 ($3.8 billion over two years), and the "High Ebola" scenario totals $32.6 billion. The card now gives those figures and Huber et al.'s ~$53 billion comprehensive burden.
Source: World Bank press release, October 8, 2014; Huber C, Finelli L, Stevens W. J Infect Dis 2018;218(suppl 5):S698, doi:10.1093/infdis/jiy213
Corrected
Revenue disclosure (How We Research page)
ebolaquestions.com
What was published: The page stated that the site "generates revenue through two disclosed mechanisms" and described Google AdSense advertising and Amazon affiliate links as live.
What it says now: Both were switched off network-wide and the About page already said so; this page had not been updated. It now states that the site earns nothing as of September 2026 and will say so on the day either changes.
Source: Site configuration (MONETIZATION_ENABLED = false); About page disclosures
Corrected
Outbreak counts on the history timeline
ebolaquestions.com
What was published: The 2022 Uganda entry gave "164 confirmed cases, 77 deaths", mixing confirmed and probable cases into one number and probable deaths into the other; the 2018–20 DRC and 2014–16 West Africa entries gave only WHO's totals; the 2007 Bundibugyo entry gave one of four published case counts as if it were the only one.
What it says now: Uganda 2022: 142 confirmed and 22 probable cases (164), 55 confirmed deaths, per CDC. Each timeline entry now names the source of every figure and shows CDC's and WHO's totals where they differ.
Source: CDC History of Ebola Outbreaks (reviewed May 29, 2026); WHO Disease Outbreak News 617; Wamala JF, et al. 2010; Roddy P, et al. 2012
Corrected
Symptom frequencies by variant (symptoms page)
coronavirusquestions.com
What was published: The symptoms-by-variant table gave the original-strain column as 83% fever, 68% cough and 38% fatigue with no source named, and listed loss of smell or taste as a "hallmark symptom (30–60%)" of the original strain.
What it says now: The figures closest to those in the literature are the WHO-China Joint Mission report of February 2020, based on 55,924 laboratory-confirmed cases: fever 87.9%, dry cough 67.7%, fatigue 38.1%. That report does not record loss of smell or taste at all — the symptom was not yet recognised when it was compiled. Every quantified cell in the table now names its study, and the Delta/Omicron columns carry the matched-cohort figures from Menni et al. (The Lancet, 2022) with their odds ratios and confidence intervals.
Source: WHO-China Joint Mission report (February 2020); Menni C, et al., The Lancet 2022, doi:10.1016/S0140-6736(22)00327-0
Corrected
The RECOVER twelve-symptom list (Long COVID page)
coronavirusquestions.com
What was published: The Long COVID page presented a twelve-item list as "the 12 symptoms most predictive of Long COVID" identified by the NIH RECOVER Initiative. Five items on it — shortness of breath, headache, sleep disturbances, joint or muscle pain, and anxiety or depression — are not in the published PASC score, and five that are — gastrointestinal symptoms, changes in sexual desire or capacity, thirst, chest pain, and abnormal movements — were missing.
What it says now: The page lists the twelve symptoms exactly as Thaweethai et al. published them, states that the score was built from 37 symptoms with adjusted odds ratios of 1.5 or greater across 9,764 participants, and shows the wrongly included symptoms separately with the explanation that they are common in uninfected people too and so carry little discriminating signal.
Source: Thaweethai T, et al. "Development of a Definition of Postacute Sequelae of SARS-CoV-2 Infection." JAMA 2023, doi:10.1001/jama.2023.8823
Corrected
Revenue disclosure (How We Research page)
coronavirusquestions.com
What was published: The How We Research page stated that the site "generates revenue through two disclosed mechanisms" and described Google AdSense advertising and Amazon affiliate links as live.
What it says now: Both were switched off network-wide and the About page had already been updated to say so; this page had not. It now states that the site earns nothing as of September 2026, that no advertising and no affiliate link appears on any indexable page, and that the disclosure will be updated on the day either changes.
Source: Site state verified in this review — the monetization flag is off and no ad or affiliate markup renders on any built page.
Corrected
Outbreak-setting percentages (3 pages)
norovirusquestions.com
What was published: The transmission, outbreaks, and cruise-ship pages attributed specific outbreak shares to CDC — "~65%" for long-term care facilities, "~15–20%" for restaurants, and "~10%" for schools.
What it says now: CDC’s Norovirus Outbreaks page says "over half of all norovirus outbreaks reported in the United States occur in long-term care facilities" and publishes no percentage for restaurants or schools; it does state that norovirus causes about 50% of foodborne-illness outbreaks and that cruise ships are about 1% of reported outbreaks. The three pages now carry CDC’s actual wording and figures.
What was published: The FAQ named Takeda’s TAK-214 as among the most advanced norovirus vaccine candidates.
What it says now: Out of date. The most advanced candidate on the public record as of September 2026 is Moderna’s mRNA-1403, in a Phase 3 trial of 37,864 adults (ClinicalTrials.gov NCT06592794, primary completion estimated February 2027). No norovirus vaccine is approved.
What was published: Four citation-level errors found in the September 2026 depth review: the symptoms page cited an unverifiable review via a PubMed search URL; the outbreaks page gave the wrong PubMed ID (26978248) for Bartsch et al. 2016; the homepage attributed the 685 million global illness figure to WHO; the treatment page attributed a "no BRAT diet" recommendation to CDC guidance, which addresses fluids rather than diet.
What it says now: The symptoms reference was replaced with the specific studies cited in the text (Atmar 2008, Atmar 2014, Burke 2021); Bartsch 2016 is PubMed 27115736; the 685 million figure is attributed to CDC’s Facts and Stats page, which states it; the diet paragraph now cites WHO’s diarrhoeal-disease guidance ("nutrient-rich foods") and says plainly that CDC’s norovirus page does not address diet.
Reported error rejected
Heat inactivation temperature
norovirusquestions.com
What was published: An internal audit flagged the stated 145°F cooking temperature as a self-contradiction against other guidance on the site.
What it says now: No change. On review, CDC’s Norovirus Fact Sheet for Food Workers supports the figure as published. The flag was a false positive and was recorded as one rather than quietly dropped.
Source: CDC — Norovirus Fact Sheet for Food Workers
Corrected
MV Hondius outbreak figures (5 pages)
hantavirusquestions.com
What was published: The 2026 Andes virus cruise-ship figures were attributed to WHO Disease Outbreak News 2026-DON599 — the initial notice, which does not carry those figures.
What it says now: Re-attributed to the WHO Disease Outbreak News edition that actually publishes them. The figures themselves were already correct; the citation pointed at the wrong document.
What was published: The 2026 DRC outbreak was described as being caused by a "novel variant." The error had propagated across roughly eleven pages including the homepage.
What it says now: Identified as Bundibugyo virus — an established ebolavirus species first identified in Uganda in 2007, not a new one.
What was published: Mortality figures from the PALM trial were misstated, including a 29-day mortality window.
What it says now: ZMapp comparison arm ≈49.4%; REGN-EB3 33.5%; mAb114 35.1%; the window is 28-day mortality (PALM trial, Mulangu et al., NEJM 2019). See the September 14, 2026 entry, which refines the comparison-arm figure to the published counts.
What was published: The 2026 outbreak section placed the outbreak in Uganda.
What it says now: Centred in the Democratic Republic of the Congo. The honest "no aggregate cost estimate yet" framing was kept rather than filled with an estimate.
What was published: Lagevrio (molnupiravir) was listed as FDA "Approved."
What it says now: Available under Emergency Use Authorization only — not FDA-approved. Paxlovid’s full approval was also scoped to adults, with ages 12–17 remaining under EUA.
What was published: The dilution was published as 1 part bleach to 9 parts water, then changed to 1:10 on the basis of a CDC citation.
What it says now: Reverted to 1:9. Re-checking found the current CDC consumer cleanup page specifies 1 part bleach to 9 parts water, so the "correction" had introduced the error. The live pages state 1:9.
What was published: ECMO centres were described as seeing ">70%" survival in severe hantavirus pulmonary syndrome.
What it says now: Approximately 60–66% at near-100% predicted mortality, per Dietl et al., J Thorac Cardiovasc Surg (2008) — 60.5%, PubMed 18329474. The overstatement recurred in several places and each instance was tracked down.
Change: PRs #158–160
Corrected
MV Hondius case count
hantavirusquestions.com
What was published: The page reported "six confirmed cases across 23 countries." Both numbers were wrong — the "23" was a corruption of the 23% case-fatality rate.
What it says now: 13 cases (12 confirmed, 1 probable) and 3 deaths, per WHO.
Change: PRs #158–160
Corrected
Third-party news summaries
All four spoke sites
What was published: Stored news items reproduced verbatim summary text from New York Times and BBC articles.
What it says now: All 42 stored summaries were removed retroactively across the news and archive sections of every spoke, and a headline-only policy adopted for those publishers going forward.
Entries begin in July 2026, when the network moved to a single repository and changes
became individually traceable. Corrections made before that date were not logged in a form
that can be cited here, and are not reconstructed from memory.
Our commitment
Accuracy is the core obligation of a health information site. When a factual error is
identified — regardless of how it was found — it is corrected promptly. We do not let
errors stand because correcting them is awkward or inconvenient.
How to submit a correction
If you believe a factual claim on any Virus Questions site is incorrect, use the
contact page and include:
The URL of the page containing the error
The specific claim you believe is incorrect (quote or describe it)
The primary source that contradicts it (CDC, WHO, FDA, peer-reviewed journal), with a direct link if possible
Corrections submitted with primary-source evidence are reviewed within 48 hours.
Corrections without supporting primary-source evidence are reviewed within one week.
How corrections are made
Verification: The cited primary source is reviewed to confirm whether the current page content is indeed contradicted.
Correction: If confirmed, the specific claim is corrected. The correction is limited to the inaccurate passage — we do not rewrite surrounding accurate content.
Date update: The page's "Last reviewed" date is updated to reflect when the correction was made.
Acknowledgment: Where a reader's submission led to a factual correction, this is noted in the page's revision history (see below). We do not publish the submitter's name without permission.
What counts as a correctable error
A correctable error is a factual claim that is directly contradicted by current primary
source guidance (CDC, WHO, FDA, NIH, or peer-reviewed literature). Examples:
A stated case fatality rate that has been revised in a CDC update
A treatment described as "investigational" that has since received FDA approval
An outbreak status that has changed since publication
The following are not corrections under this policy: differences of interpretation,
preferences for different sources, or additions of new information not yet contradicting
existing content. These may be addressed in regular content updates.
Transparency
Significant factual corrections are noted at the bottom of the corrected page, including
what was changed and when. Minor corrections (e.g., fixing a broken citation link, updating
a date that changed) are made without a formal note but are reflected in the updated
"Last reviewed" date.
This site uses cookies for analytics (Google Analytics). Privacy Policy