- Design
- Observational cohort study using national administrative and electronic health record data, 2017-2024
- Population
- 16,799 veterans aged 50 or over newly admitted for 90 days or more to long-term care, mean age 74.9, dementia in 44.2%
- Primary outcome
- Cumulative incidence of herpes zoster from 90 days after admission, and vaccination uptake
- Effect
- Incidence 0.8% at 180 days, 1.4% at 365 days, 2.1% at 720 days; 1.2% in recombinant vaccine recipients and 4.1% in those with prior zoster; only 2.9% received any dose within 90 days of admission
Sixteen thousand seven hundred and ninety-nine veterans aged 50 or over, newly admitted for at least 90 days to long-term care between 2017 and 2024, were followed using administrative and pharmacy records. Mean age was 74.9 and comorbidity was what you would expect: dementia in 44.2%, hypertension 37.3%, hyperlipidaemia 30.1%, anaemia 23.8%, diabetes 22.8%.
Before admission, 35.3% had received the live zoster vaccine and 23.1% the recombinant one. In the 90 days after admission, 482 residents - 2.9% - received any zoster vaccine dose at all. Cumulative herpes zoster incidence measured from 90 days after admission was 0.8% at six months, 1.4% at one year and 2.1% at two years. It was lowest among those who had received the recombinant vaccine (1.2%) and highest among those with a previous episode of zoster (4.1%).
The honest reading is that coded zoster is not common, and the authors say so. But 2.1% over two years in a population where 44% have dementia is not nothing, because the burden of zoster in that group is not the rash - it is postherpetic neuralgia in someone who cannot report pain, presenting as agitation, refusal of care or a fall. The intervention is a protocol rather than a clinical decision: vaccination status checked and completed at admission, in the same list as the medication reconciliation and the falls assessment. A 2.9% uptake in the 90 days after admission is not clinical judgement operating, it is nobody having been made responsible.
In Indian long-term care and at home, the constraint is cost and availability rather than protocol, and that is worth naming honestly with families rather than assuming the vaccine is simply on offer.
- Add zoster vaccination status to the admission checklist alongside medication reconciliation
- A previous episode of zoster is a reason to vaccinate, not a reason to skip it - that group had the highest incidence at 4.1%
- Consider zoster in a resident with dementia and new agitation, refusal of care or unexplained distress
- Record the live-versus-recombinant distinction; they are not interchangeable for this population
- Where cost is the barrier, say so to the family rather than leaving the impression it was not indicated
Why it matters
The gap here is not a clinical judgement anyone made; it is a step that no admission process owns.
Don't overread it
Coded incidence in an administrative dataset will undercount zoster, and the vaccinated-versus-unvaccinated comparison does not estimate vaccine effectiveness.
The statistics, in plain English
The incidence figures come from diagnostic codes, which will undercount zoster in a population where 44% have dementia and many cannot describe a burning pain before the rash appears - the true incidence is almost certainly higher than 2.1% over two years. The comparison between vaccinated and unvaccinated groups is not a measure of vaccine effectiveness: people who got vaccinated before admission differ systematically from those who did not, in health engagement and in how much contact they had with services. The higher rate in those with previous zoster (4.1%) is the more robust observation, because prior infection is recorded reliably and recurrence is a known phenomenon.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for geriatrics, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free