'American Covid' Isn't New, and So Far Its Autumn Is Smaller Than 2025's
The headline figures are real UKHSA numbers, but read against last autumn and their own definitions they tell a calmer story. What they count, how to read them, and who should book a jab.
The page is UKHSA's Covid-19 data dashboard, and its first three tiles hold the week's headlines: 2,127 cases, 41 deaths, 1,168 patients admitted to hospital. Right after them, a fourth tile reads "Autumn booster uptake" (people aged 65 and over), dated "Up to 29 Jan 2025", at 59.3%. That one has not moved since.
The figures behind the "American Covid" stories come from those top tiles, and they are real. What most pages never say is what each number counts, or what it is measured against.
Is it new? No. UKHSA's blog says the XFG variant, nicknamed "Stratus", "is not new". Is it more dangerous? UKHSA says current data does not indicate that. Are levels high? The agency's report published Thursday, Oct. 15 calls Covid "circulating at low levels". Measured against the same week last year, this autumn is smaller so far, though it is rising.
A 230% rise that starts from 1,390 cases in August
The dashboard tile for cases shows 2,127 for the week to Oct. 7, "an increase of 253 (13.5%)". The 230% comes from a different comparison. Coventry Live, AOL (carrying the Daily Express) and LADbible all reported cases rising from 1,374 in August to 4,589 in September; Coventry Live wrote that they "rose dramatically".
The sum gives 234.0%, so "230%" is a loose rounding. The dashboard series, pulled on Oct. 18, now reads 1,390 for August and 5,181 for September, a rise of 272.7%. That sounds alarming until you set it beside last year.
| Aug 2025 | Sept 2025 | Aug 2026 | Sept 2026 | |
|---|---|---|---|---|
| Cases (lab-confirmed) | 6,399 | 10,009 | 1,390 | 5,181 |
| Change, Aug to Sept | +56.4% | +272.7% | ||
| Hospital admissions | 4,617 | 7,122 | 967 | 3,263 |
| Change, Aug to Sept | +54.3% | +237.4% |
The percentage is bigger this year only because the starting line was so low: July 2026 had 594 cases in the whole month. September's 5,181 is about half of last September's 10,009 (0.52 of it), and admissions are at 0.46. A percentage tells you how fast something moved, not how far.
Testing also rose, from 367,031 PCR tests in August to 458,100 in September (24.8%), so part of the rise is more people being swabbed. And a "case" is a positive lab-confirmed PCR result by specimen date; UKHSA says the figure "will underestimate the actual number of COVID-19 infections", and the latest days are incomplete. The dashboard's rolling seven-day positivity, averaged by month in our calculation, is steadier: 2.58% in August and 6.57% in September, close to UKHSA's "around 7% in September".
41 death certificates in England, against 102 a year ago
"41 dead" is the sharpest headline and the easiest to misread. Yorkshire Live, to its credit, carried the comparison with last year: "Covid deaths have yet to reach the same peak as this time last year, when fatalities hit 102 for the week ending October 3, 2025."
The dashboard's series is "weekly deaths with COVID-19 on the death certificate by date of registration". It is not people who died within 28 days of a positive test, and the person may never have tested positive. The Office for National Statistics (ONS) defines the wording:
"Deaths 'involving' a cause include all deaths where the cause was mentioned anywhere on the death certificate, either as a main cause of death or a contributory cause."
Office for National Statistics, weekly deaths notes
It is England, not the UK. ONS, in its release of Wednesday, Oct. 14, counted 41 deaths involving Covid in England for the week ending Oct. 2, among 9,481 from all causes (0.43%). Across England and Wales it was 43. Of England's 41, 26 had Covid as the underlying cause.
The weekly rise is real: 28 the week before. But the earlier weeks were distorted. The week ending Sept. 4 held the Monday, Aug. 31 bank holiday, with 7,557 deaths registered from all causes against 9,644 the next week, so the jump from 6 to 20 is partly a registration effect.
Last autumn's peak week, ending Oct. 17, 2025, reached 159, the highest weekly figure since November 2024. This year's 41 is 40% of the 102 and 26% of the peak. Over the five weeks to Oct. 2, the total was 115, against 446 a year earlier.
None of that makes 41 deaths small to the families concerned. It shows a count of certificates, in England, in a series still climbing. One week can change the picture.
Hospitals are busier, but not by last year's measure
Hospital data are where the rise is most genuine. UKHSA's weekly surveillance report, covering data to Sunday, Oct. 11, put the Covid admission rate at 2.46 per 100,000 people in week 41, up from 2.13. The Daily Mail's "40% in a fortnight" checks out: from 1.75 to 2.46 is 40.6%.
The same week last year the rate was 4.09, when UKHSA described Covid as at "medium levels". This year's 2.46 is 60% of that. In plain terms, about one person in 40,650 is admitted in a week. For people aged 85 and over the rate is 26.69 per 100,000, roughly one in 3,750, up from 24.20. Intensive care admissions stayed at 0.06 per 100,000.
These are admissions of people who tested positive; the report does not separate those admitted because of Covid from those who happened to have it. They are revised too: week 40 moved from 2.04 to 2.13. And they describe a population, not your odds. The 85-and-over rate is why the advice below targets particular groups.
Post by @UKHSA
Hospital positivity was 8.8% to Oct. 11 against 9.0% a week earlier, and cases in hospital settings were 1,848 against 1,847. GP sentinel swabs fell from 10.2% to 6.6%, but that rests on 376 samples, 42.3% from the South West, and UKHSA cautions it may not be representative. Neither reading says the rise has ended.
Wales looks similar. A Public Health Wales spokesperson, quoted by North Wales Live, said:
"As of 11 October 2026, there were 161 patients in hospital in Wales who had tested positive for COVID-19, compared with 415 patients during the equivalent week last year."
A Public Health Wales spokesperson
UKHSA's blog post on Oct. 8 explains the pattern without a new villain: admissions are "rising in line with overall increases in COVID-19 circulation", suggesting "a general increase in COVID-19 activity rather than a variant causing more severe disease". It names colder, wetter weather indoors and immunity that "may be waning" a year after the last significant wave.
A 20-month-old variant, and who should book a jab
UKHSA is blunt about the nickname: "The name reflects where the strain was first sequenced and detected, rather than necessarily where it originated." "Stratus" is a media nickname, not a World Health Organization label.
WHO's variant tracker lists XFG as a variant under monitoring since June 25, 2025, with an earliest documented sample from Jan. 27, 2025: about 20 months old. UKHSA's "around a year" refers to circulation in the UK. The tracker lists no variant of concern, and the European Centre for Disease Prevention and Control says the same, with "no evidence" of any effect on transmissibility, immunity or severity. WHO's 2025 risk evaluation put the added risk at "low at the global level".
"It's completely normal for viruses to change over time. The strains currently going around, including those known as XFG, are closely related to strains that have been with us for over a year, so this isn't something new."
Dr Alex Allen, consultant epidemiologist, UK Health Security Agency (to Newsweek)
Is XFG still the main variant? UKHSA's blog says it "remains the most common current COVID-19 variant". Its fortnightly lineage table, for samples sequenced Aug. 31 to Sept. 13, is more fragmented: TA.1 at 25.8%, RV.2.1 at 16.7%, XFG itself at 9.1%. The table does not say which lines descend from XFG, so the shares cannot be added up, and no other line should be called a new variant. UKHSA warns that low sequencing numbers "will impact the accuracy of the prevalence estimates". A year earlier the table read 81.94% XFG.
Symptoms are the NHS list, from a high temperature and a new continuous cough to a change in smell or taste, which the NHS calls "very similar to symptoms of other illnesses, such as colds and flu". Eye symptoms and back pain are not on it. Newsweek reports no official evidence that XFG causes a different set of symptoms, and doctors cannot tell the variant from symptoms alone.
If you are ill, the NHS advice is to stay home if you have a high temperature or do not feel well enough for normal activities. If you test positive, stay home for five days after the test if you are 18 or over, three if under 18, and avoid people with weakened immune systems for 10 days. Anyone who may be eligible for Covid treatment, such as people with immune conditions, kidney or liver disease or lung conditions, should test as soon as symptoms start. Seek urgent GP or 111 help if symptoms worsen or you are pregnant, 60 or over or immunosuppressed and worried; call 999 for sudden chest pain, breathlessness at rest or collapse.
The vaccine is where readers can act. The autumn 2026 programme covers adults aged 75 and over (including those who turn 75 by Jan. 31, 2027), residents of care homes for older adults, and anyone aged 6 months and over who is immunosuppressed, across all four nations. The NHS lists examples such as blood cancer, transplant, HIV and chemotherapy, and says to ask your pharmacy, GP surgery or specialist if unsure. Healthy under-75s are not included.
In England you can book online or in the NHS App, or use a walk-in site. The NHS says: "You do not need to wait for an invitation before booking an appointment." Appointments run from Oct. 1; the last date for a dose is Jan. 31, 2027 and online booking closes Jan. 30, so 105 days remain from today. It can be given alongside the flu or RSV vaccine.
Uptake is early. By Oct. 11, 23.0% of people aged 75 and over in England and 9.9% of immunosuppressed under-75s had received a dose, up from 10.2% and 4.1% a week before. UKHSA says last year's vaccine was 47% effective for the three months after vaccination, and cites NHS England data suggesting around 16,000 people were admitted with the virus between October 2025 and January 2026. Dr Allen: "If you are eligible for a COVID-19 vaccine, it is because you are at greater risk of severe illness and being hospitalized. Don't risk it; get the vaccine and give yourself the best protection."
For flu, UKHSA says England is at baseline levels overall, with 6.0% laboratory positivity and 1.88 hospital admissions per 100,000. The site's piece on this autumn's early H1N1 flu season and when to get a flu shot is North America-focused, but it is the place to go for timing.
Scroll the dashboard and the vaccine tile still reads 59.3%, dated "Up to 29 Jan 2025". Of everything on that page, it is one of the figures that has stopped (the genomic lineage chart has been paused since April 2024), and it concerns protection. The live number sits in the weekly report instead, 23.0% of over-75s so far, with NHS booking open until Jan. 30, 2027.
