Very often in articles related to nuclear disasters are totally unaware of the correct unit to use for human casualties and risks.
The correct unit is Micromort not how many people died. Or to be more precises the accumulation of them.
If from an accident one 40 year old dies, he lost half his life. If through another accident 4 people will die 10 years earlier, both events have the same mort amount, just differently spread, but the impact on society is roughly the same in sense of human-alive-duration.
As nuclear catastrophes usually affect millions, even just some years of reduced lifespan accumulate to many morts and not just the 0 the nuclear industry tries to sell us.
Actually, you need to look closer, how the accident affected the quality of life in general. ECRR is a good source for that.
If the birth defect rates rise as it did in Chernobyl, the amount of suffering increases, for the child, for the supporting families, friends,...
The immune system in younger generations is often reduced near the fallout zones. This will impact their way of living as well - nobody enjoys beeing sick, sick days are mostly wasted days. If you give these effects a quantifier and start to accumulate those effects, the cost of a nuclear disaster will most likely not be comparable to anything else.