There is a crucial ethical distinction between incidental harm and death and mandating harm and death.
When mandates are issued for interventions that carry a non-zero probability of death or severe harm, we can not simply compare statistically the number of lives that one believes will be affected by harm from the disease and those that will be harmed by the imposed treatment.
Imposing a mandate implies an individual is no longer free to select their exposure to possible harm. Someone else is making that decision for them.
If this is justified by appealing to the probabilistic safety of the individual, it is an ethical non-starter. It is never ethical to impose possible harm on an individual by citing protection of that individual (except in the case of guardianship where decisions must be made on behalf of an individual who is unable to make them).
If, however, an imposition is justified by appeal to the protection of others, or of society at large, there may be an ethical case to be made for imposition. That is, in the case of disease, if a treatment that carries lesser risk than the disease reduces the propensity to spread the disease, there is a potential utilitarian gain in the mandated imposition of said treatment.
If the reduction in propagation is enough to eliminate the disease altogether, there may be a strong case for a mandate. If, however, that reduction is not enough for elimination, if the disease continues to circulate, any case made for mandates becomes much weaker.
The ethical situation, albeit not necessarily the statistical situation, devolves to the risk of incidental exposure and that of forced exposure.
In the first case, incidental exposure, agents retain volition in terms of what they choose to engage in and how they may be exposed. In the latter case, forced exposure, all choice is eliminated.
The reduction of choice to zero is not something that should be done without strong justification. How many lives, statistically, are worth this tradeoff? The answer is not clear. And that is an essential point: we should not posture as if it is clear. A crucial component of ethical behavior is in acknowledging uncertainty.
But I am confident in saying the ethics of this uncertain situation are not exhaustively covered by utilitarian statistical calculations. They may be used to build a case, but not to determine an answer. We enter here the realm of questions such as ‘what is good?’, ‘what makes life worth living?’. Ultimately, these are philosophical and theological questions. Certainly most would agree a marginal increase in lifespan would not justify being enslaved — but where do the thresholds lie?
When one supports mandating what amounts to guaranteed harm and death for some number of other people, even if one makes a case for the protection of others in so doing, one must tread very lightly. Else one flirts with evil.