Study Puts Puerto Rico Death Toll from Hurricane Maria Near 5,000 | Hacker News Reader

Study Puts Puerto Rico Death Toll from Hurricane Maria Near 5,000

npr.org

16 pointsrcarndrums4 comments

>"A research team led by scientists at the Harvard T.H. Chan School of Public Health didn't simply attempt to count dead bodies in the wake of the powerful storm. Instead, they surveyed randomly chosen households and asked the occupants about their experiences."

I expected the death toll to refer to the former. Whatever was measured here was more like "estimated deaths due to the stressful seqeulae of Hurricane Maria."

>"The researchers calculate there is a 95 percent likelihood the death toll was somewhere between about 800 and 8,500 people."

This actually interprets a confidence interval correctly. Something is getting through to NPR. EDIT: Wait, no it is still wrong. (If assumptions are ok, etc) there is a 95% chance any given CI contains the actual mean, we can't say anything about the chance this particular CI contains it.

>"This household-based survey suggests that the number of excess deaths related to Hurricane Maria in Puerto Rico is more than 70 times the official estimate... surveys can provide an independent estimate of mortality that does not rely on death-certificate data" https://www.nejm.org/doi/full/10.1056/NEJMsa1803972

Maybe, but then you need use the death certificate data as a gold standard (somewhere where those records are good) to see if your survey was measuring the same thing. In this case though, like mentioned above, they measured something different from what people usually mean by "killed by hurricane".

The article discusses the methodology in-depth including the meat of the specific points you raise. The study was published in a peer-reviewed journal by professionals who understand the subject intimately and whose conclusions carry overwhelmingly more weight than your off-the-cuff analysis, one which misses the overarching point that based on available data, it is a statistical near-certainty that hundreds or thousands of people died as a direct result of the hurricane. Had you been a victim, I sincerely doubt you'd be so concerned about the distinction between being drowned right away or dying of a medical issue that would have been preventable with care that would have otherwise been accessible before the hurricane.

>"The article discusses the methodology in-depth including the meat of the specific points you raise."

The meat of my methodological issues is that the survey has not been validated in any way as a measure of mortality. Where do they deal with that? From the supplement, it looks like this is a custom survey that has never been used before to me.

>"The study was published in a peer-reviewed journal by professionals who understand the subject intimately and whose conclusions carry overwhelmingly more weight than your off-the-cuff analysis, one which misses the overarching point that based on available data, it is a statistical near-certainty that hundreds or thousands of people died as a direct result of the hurricane."

Anyone who has looked at enough studies like this knows the results are typically highly questionable. My comment was more to alert them to which of the usual issues seem to be present. If you are still reliant on the "peer-reviewed" heuristic, then I agree that you should not listen to random people on the internet, or even in person, like me.

>"Had you been a victim, I sincerely doubt you'd be so concerned about the distinction between being drowned right away or dying of a medical issue that would have been preventable with care that would have otherwise been accessible before the hurricane."

Taken literally, this is true. If I had been a victim I would be dead, so I wouldn't care about anything at all. I think that is the more generous interpretation of you just wrote. If a family member had been a victim, I would definitely want such studies to distinguish between dying from drowning and dying due to eg lack of access to healthcare a week later. I don't see how it is helpful to conflate these two and confuse the public since they require different solutions.

This is incredibly tragic, and equally, if not moreso, tragic that this is not in the least bit surprising given that the administration's response was to belittle and attack the people who were on the ground trying to save lives and crying out for help. There is palpable irony too in the administration's relentless crowing and insistence on massive increases in military funding, and yet when it comes down to brass tacks in a situation where the military could have been leveraged in an all-out, no-holds-barred operation to save the lives of thousands of American citizens, they chose not to and instead focused on pathetic, flimsy excuses ("big water, ocean water", heartless arguments about PR's financials, etc.) instead of demonstrating true leadership. These are truly dark times we are living in, and it is tremendously unfortunate that the poorest and most vulnerable of our citizens bore, and continue to bear, the ultimate and final costs in the form of their lives.

Reply on news.ycombinator.com