I think the following will tend to be the best to maximise the cost-effectiveness of saving a human life:
Accounting solely for the benefits to the person saved, saving human lives in countries with low, but not too low, real gross domestic product (real GDP) per capita. Saving a human life is cheaper in lower income countries, but self-reported life satisfaction and life expectancy decrease with income. So saving human lives in the lowest income countries may be suboptimal.
Accounting solely for the benefits from economic growth, saving human lives in countries with high real GDP per capita. Saving a human life is more expensive in higher income countries, but these have greater productivity and life expectancy. So paying more to save human lives there may be worth it.
In terms of effects on animals, I consider:
It is unclear whether saving human lives has a positive/negative impact on nearterm welfare accounting for effects on both humans and animals.
Saving human lives in high income countries is better than in low income countries if the benefits from economic growth dominate. The former increases animal suffering nearterm more, but makes it peak and end earlier, such that there is a greater overall reduction.
Eating less animals and more whole-food plant-based increases economic growth via decreased mortality.
My overall view is that:
If improving nearterm welfare is the best proxy to maximise future welfare, helping animals seems better than saving human lives in low income countries.
If boosting economic growth is the best proxy, saving human lives in high income countries seems better than in low income countries.
I guess improving nearterm welfare is a better proxy than boosting economic growth. Nonetheless, I am mainly in favour of research on whether indirect longterm effects dominate, and, if so, on which are the best proxies for them.
Relatedly, I have just published a post on Helping animals or saving human lives in high income countries seems better than saving human lives in low income countries?.