Reduced Risk Is Not No Risk

Two-dose JYNNEOS/MVA-BN vaccination does not mean zero risk because protection is not one yes-or-no outcome. It can lower the chance that an exposure becomes mpox, reduce the expected severity when a vaccinated person does become infected, and lower the likelihood of hospitalization in observed populations. A breakthrough infection remains possible because exposure continues, immune responses vary, and the available evidence cannot measure every infection or every relevant condition. The accurate conclusion is therefore not “the vaccine failed” or “the person was invulnerable,” but “the risk was reduced and not eliminated.”

That distinction is the practical value of the evidence. It allows a vaccinated person, clinician, or public-health program to hold two facts together without treating either as a contradiction: a completed series is worth having, and a vaccinated person can still develop mpox. The first fact concerns the average shift in outcomes. The second concerns the remainder of risk after that shift.

The answer to the title is a three-layer relationship

The evidence supports a chain rather than a single endpoint.

First is the probability of infection. In the CDC analysis of U.S. cases reported from May 2022 through May 2024, 271 of 32,819 cases occurred in people who met the report’s fully vaccinated definition. They represented 0.8% of the analyzed case pool. In a separate subset of jurisdictions with usable coverage data, 187 infections were reported among 188,907 fully vaccinated people, an estimated 0.1% infection rate for that defined period and population. The first figure describes the share of reported cases; the second uses a vaccinated-population denominator. Neither is a universal personal probability, but both are compatible with a rare rather than impossible breakthrough pattern.

Second is severity conditional on infection. Once a person has mpox, the question changes from “Was infection prevented?” to “What did the illness do?” The CDC report found less systemic illness and less widespread rash among fully vaccinated cases with the relevant information. The Swedish clinical series found a median mpox Severity Score of 5.0 among 30 patients vaccinated with MVA-BN only, compared with 8.0 among 32 unvaccinated patients. The score combined lesion burden, anatomical spread, mucosal involvement, bacterial superinfection, level of care, and analgesia needs. A positive test could therefore be present in both groups while the average clinical burden differed.

Third is the risk of a severe outcome, particularly hospitalization or death. In the U.S. 2022–2024 comparison, 3 of 212 fully vaccinated patients with known hospitalization outcomes were hospitalized, or 1.4%, compared with 1,662 of 19,716 unvaccinated patients, or 8.4%. In the newer 2024–2025 surveillance analysis, 2 of 118 fully vaccinated patients in the complete-data model were hospitalized, compared with 94 of 646 unvaccinated patients. After adjustment for age group and HIV status, the odds of hospitalization were 9.73 times higher among unvaccinated than fully vaccinated patients, with a 95% confidence interval from 2.93 to 60.65.

These layers are related but not interchangeable. A vaccine can reduce the chance of infection without preventing every infection. It can reduce average severity without making every breakthrough mild. It can be associated with lower hospitalization odds without making hospitalization impossible. A person who develops mpox after two doses is evidence that residual infection risk exists; that person’s illness can still be evidence that vaccination reduced the expected severity of the breakthrough.

This is why “zero risk” is the wrong standard for judging the series. If zero is the only acceptable result, any breakthrough is treated as total failure and all reductions in disease burden disappear from the analysis. If any reduction is treated as proof of complete safety, the breakthrough and the remaining exposure pathways disappear instead. The three-layer model avoids both errors by asking what changed at each stage.

The model also clarifies why a percentage cannot answer every practical question. A population rate describes a group over a defined period. A severity score describes infected patients in a defined clinic. An adjusted hospitalization odds ratio describes a complete-data subset under a specified statistical model. Each can be decision-relevant, but none can be copied into an individual’s life without considering exposure, age, immune status, route, time, and whether an infection would have been detected.

U.S. breadth and Swedish clinical detail point in the same direction—but not to the same claim

The U.S. and Swedish sources are strongest when their different designs are allowed to complement one another. The national CDC report covers a large number of reported cases across jurisdictions. It can show that breakthrough cases were a small fraction of the reported case pool, that fully vaccinated cases generally had less systemic illness and rash spread, and that hospitalization was less common in the available comparisons. Its breadth gives the pattern public-health significance.

The Swedish study covers a much smaller group, but it examines the clinical texture of disease more closely. It reviewed 79 laboratory-confirmed cases at a specialized HIV and sexual-health center in Stockholm. All MVA-BN doses were intradermal. Two clinicians who were blinded to vaccination status applied a validated severity score, allowing the study to describe lesion burden, anatomical spread, mucosal involvement, bacterial superinfection, analgesia, and level of care in one structured assessment. Its value is not a larger estimate of population protection.