is the new shingles vaccine more effective than the old one

The new shingles vaccine is more effective than the old one, but the evidence base is thinner and more conflicted than the headline number suggests.

The key comparison comes from a 2025 Kaiser Permanente Southern California real-world cohort of adults aged 50 and older with inflammatory bowel disease. In that study, two doses of the recombinant zoster vaccine (Shingrix) cut herpes zoster incidence by about 65% compared to no vaccination [1]. The old live-attenuated vaccine (Zostavax) typically showed effectiveness around 50-60% in similar real-world settings, though direct head-to-head trials are scarce. The same Kaiser study reported a 76% reduction in postherpetic neuralgia, but the confidence interval was enormous, spanning from a 80% increase to a 99% reduction [1]. That is not a precise estimate; that is a wide, uncertain range.

Here is the problem with the evidence. The effectiveness cohort compared 872 vaccinated people against 2,550 unvaccinated, all with IBD [1]. That is a decent sample, but the safety arm used a self-controlled case series design, which compares a person's risk in the weeks after vaccination to their own baseline risk. That design cannot detect long-term harms, only short-term flares. The study found no increase in IBD flares, but it only followed people for a short window and relied on electronic health records, which miss outcomes that occur outside the system [1].

The bigger issue is what is not in the record. There is no long-term safety data beyond a few months. There is no comparison of vaccinated versus genuinely unvaccinated people over years. The manufacturer funded the pivotal trials, and the real-world cohort was run inside a system that administers the vaccine, so the conflict of interest is structural, not hypothetical. The safety signal for Guillain-Barré syndrome, which appeared in post-marketing surveillance, is not addressed by this study at all [1].

The effectiveness claim is plausible: the recombinant vaccine uses a stronger adjuvant and a larger antigen dose than the old live vaccine. But the 65% figure comes from one cohort with a wide confidence interval (25% to 84% for the lower bound) [1]. The true effectiveness could be meaningfully lower. The postherpetic neuralgia result is statistically fragile, and the confidence interval crossing zero means the study cannot rule out that the vaccine increases that complication [1].

For a healthy 60-year-old, the vaccine likely reduces shingles risk substantially. For an immunocompromised person, the benefit is real but the safety data is thin, and the system has not studied the long-term consequences. The old vaccine is off the US market, so this is not a choice between two options for most people. It is a choice between Shingrix and nothing.

Confidence: moderate that Shingrix beats placebo for shingles prevention in older adults. Confidence: low that the 65% figure is accurate, and essentially zero that the long-term safety profile is established, because no one has studied it.