You really need to look at trends more than absolute numbers, or you will run into the issue of actually verifying that the numbers are comparable.
E.g. death due to X tends to depend greatly on what the policy is on autopsies and other checks, and whether or not there's any pressure on coroners to speculate about cause of death vs. just pronounce a generic reason.
In Norway for example, an autopsy only happens rarely unless there is a reason to expect foul play, mostly because it's seen as expensive and unnecessary. As a result, a lot of "natural causes" or "heart failure" can obscure more specific causes that will only be recorded if accompanied by e.g. police observations.
Then, even if you do tests, there's the question of how you record deaths where there are multiple possible causes. E.g. you might have heart failure possibly due to a combination of alcohol and other drugs. Depending on your reporting system you might put it down in any number of different ways.
Then on top of this, you have to consider what make people use, not just what potentially make people not use. It could be the differences in legal system, or just that other societal differences makes fewer people predisposed to want to use in the first place.
All of this basically makes it impossible to compare country to country and conclude that changes in a single area such as law enforcement makes a difference.
The reason we look to Portugal is not primarily because it looks good compared to other countries, but because it performs well compared to itself prior to decriminalisation, and because it was a recent change.