250,000 Jehovah's Witnesses have died refusing blood

by nicolaou 739 Replies latest watchtower medical

  • slimboyfat
    slimboyfat

    It matters if sensationalism and a disregard for the facts discredits legitimate complaints about the JW blood doctrine in general.

  • Marvin Shilmer
    Marvin Shilmer

    -

    “erm, but it's OK for you to transpose this risk to 7.5 million people?!?!?”

    The 7.5 million match the sample set of JWs in New Zealand. The 10x multiplier is not presented of all patients with anemia. In other words, if we look at the 10x as a “sample set” it does not align (match) with all patients with anemia. It only aligns with patients with conditions that align with the matched comparison groups.

    “Anyway, the questions I've asked are easy to find. I thought you'd be able to find them by clicking on the page numbers but obviously not. I know you're trying to bait for insults so you can storm off but I'd like to thrash this out until you see your own error.”

    I’ve failed to address several questions and comment in this discussion because they didn’t have any merit worth answering for. I said this already.

    “How come your numbers don't change if the rate of deaths does? If there are 1000 JWs instead of just the 103 who refused treatment ... the rate is the same as regular people (1.9%) but your method comes up with 50,000 "extra" deaths. Are you claiming that *no* JW ever accepts blood?”

    My numbers are tied to hard data and not rates. The number of preventable deaths is hard. The number of JWs in New Zealand is hard. Various mortality rates will not change these hard numbers. These hard numbers present a fixed ratio.

    “Your mistake through all this is to treat "JWs who chose to refuse blood" as synonymous with the entire population of JWs. Any stats to do with deaths from refusal to accept blood is already based on a smaller and more specific sample which is NOT representative of JWs in general. In fact, all JWs who do not refuse to accept blood are in the stats for, well, everyone else.”

    The data set from Beliaev includes JWs who accepted blood product forbidden under Watchtower doctrine. Hence the issue you raise is present and accounted for.

    Regardless, we still have a hard number of preventable deaths and we still have a hard number of JWs in New Zealand. Also, the data set at issue is specifically about red cells transfusion. Those who died over and above the norm refused red cell transfusion. If there were JWs in New Zealand suffering severe anemia who accepted red blood cell transfusion all that means is those JWs are not in the data set and whether they lived of died does not affect the ratio mentioned above because the ratio is not of JWs who die and those who do not die based on accepting red cells transfusion but, rather, the actual number of preventable deaths versus the JW New Zealand population.

    “It's why your numbers don't tally with simple sanity checks and experience / observation.”

    The number of 50,000 over a 50-year period is imperceptible unless a person has figures before them to do the math.

    “If he'd had more patients in the study but the same mortality rate, then what would the extra deaths do to your numbers? What effect would fewer numbers (but the same rate) have?”

    My extrapolation is based on hard numbers and not a rate. However, if more patients died and the rate of mortality remained the same among the JW patients compared with the matched group then nothing would have changed in terms of JWs at risk for premature mortality.

    “If your method was based on the rate it wouldn't change ... but it does. Almost as though it has nothing to do with the mortality rate.”

    Again, look at the hard numbers. The mortality rate in the Beliaev study is not stated in a way to apply it to a larger sample that does not align with the matched group. But the hard numbers are stated in a way to apply it to a larger group.

    Marvin Shilmer

  • Simon
    Simon

    Yes, if a journalist gets burnt reporting data that turns out to be nonsense then chances of issues and news items being picked up again in future diminish dramatically.

    Wrong numbers, whether accidentally or intentionally can do great damage. It's all about credibility and trust and ultimately can undermine other issues that have heavy basis in fact and evidence. If we get blood death figures so wrong then what about child abuse figures too? "Oh, those people are just bitter and make up false claims - look, here is the proof I can point to". It matters.

    I really don't see any problem with someone saying 'oh yeah, I hadn't taken those factors into account ... thanks' which garners more respect than stubbornly sticking to a debunked theory and insulting or belittling anyone who questions it.

  • Simon
    Simon

    Marvin: The only hard number is 19. Stop trying to make out you have more facts than you do.

    All you have is a very small sample from a very small corner of the world by people looking for that outcome.

    I am happy that you are so convinced but I think you're numbers don't make sense and your attitude and approach to people asking questions reduces confidence still further.

  • Marvin Shilmer
    Marvin Shilmer

    -

    “Marvin, kindly show me the actual finding that shows "there is no reason to think mortality due to severe anemia patients (for all causes) refusing blood was more prevalent in New Zealand than any other nation".

    “Your statement looks suspiciously like an assumption and not a finding at all. Or do you use the words "finding" and "assumption" interchangeably. Your responses do not inspire confidence I'm afraid.”

    For statistical purpose I did assume an equal prevalence of severe anemia between New Zealand and the rest of the world and have said so numerous times in this very discussion. I looked for reasons why we should treat the New Zealand incidents of severe mortality as either larger or smaller. What I found is that, particularly in the case of JWs, the evidence points to higher incidents on average outside New Zealand because by comparison the healthcare system in New Zealand prevents patients suffering anemia from getting the point of severity of patients in the Beliaev study. But for JWs this is not the case among those who refuse transfusion of red cells. Hence my assumption for purpose of statistical extrapolation.

    “Furthermore, your statement assumes (1) you have supportive data that shows the numbers of JWs with serious blood disorders (requiring transfusions) in other countries around the world is on a par with New Zealand JWs and (2) that the availability of Western medical treatment (especially blood transfusions) is uniformly equally available elsewhere in all other countries where there are Witnesses.”

    My extrapolation is not of New Zealand versus Western nations, though I don’t see how that would change anything given a general parity of healthcare. My extrapolation is of New Zealand versus all other nations, and in this comparison, if anything, we should expect incidents in New Zealand to be less because of the state of healthcare in that nation.

    “You unknowingly pile assumption upon assumption upon assumption which bodes poorly for your conclusions - and you appear genuinely baffled by the robust criticism.”

    Assumptions I’ve made for statistical purposes are either implicitly or explicitly stated on my blog. In this discussion I’ve shared them in explicit terms and repeatedly. The primary assumption affecting my extrapolation is my proration of incidents from 19 to 33 over the 10-year period. But for reasons given in this discussion not only do I see this proration as reasonable I see it as conservative.

    Marvin Shilmer

  • Marvin Shilmer
    Marvin Shilmer

    -

    “All maths and extrapolations and assumptions aside, the implications of the 50k and 250k number don't make sense.

    “Marvins 'conservative' 50k implies 30 JW's a year bleeding to death in UK hospitals - 250k implies...well you can do the maths.”

    besty,

    No. Not necessarily bleeding to death.

    There is a difference between severe anemia and bleeding to death. One could lead to the other but one is not necessarily related to the other. Apparently you don’t know the difference.

    When it comes to recognition of the number of JWs who suffer premature death due to refusing blood it is often difficult then-and-there to say the reason was refusal of blood product. This realization is something that emerges after analyzing data and making matched comparison. Hence not only is the statistical number of deaths diluted to a point of about 1 per year per 60 congregations, but this 1 per year may be unrecognizable until (if ever!) an overall matched comparison is made. Again, this is something you don’t seem to understand.

    “Marvins numbers implies that for every JW death that makes the national and local news there are another 30 or so deaths that go unreported....”

    Yes. Because of what I write above. The lion’s share of this mortality goes unreported because it’s only recognizable in hindsight and then only when/if someone conducts a matched comparison as Beliaev did in New Zealand.

    Marvin Shilmer

    Marvin Shilmer

  • Marvin Shilmer
    Marvin Shilmer

    -

    “Marvin: The only hard number is 19. Stop trying to make out you have more facts than you do.

    “All you have is a very small sample from a very small corner of the world by people looking for that outcome.”

    Simon,

    Respectfully, you don’t know the difference between a sample size and a population of that sample. It’s apparent by what you write above in concise terms.

    Yes. The hard number is 19. But this is only 19 deaths among a minority of trauma centers in New Zealand. This is something you fail to account for, and so far even talk about in terms of how it impacts the discussion.

    When someone raises something worth responding to I'll do so. Otherwise I think I’m done here.

    Any specific questions?

    Marvin Shilmer

  • Simon
    Simon

    Yes. The hard number is 19. But this is only 19 deaths among a minority of trauma centers in New Zealand. This is something you fail to account for, and so far even talk about in terms of how it impacts the discussion.

    Wha... ? That's what this whole discussion has been about. How truly represantative of the ENTIRE JW POPULATION is this SMALL SAMPLE and can you be justified applying it to all JWs.

    It's a pre-selected sample of JWs who refuse blood. Not all JWs do. That's one reason why your figures are off. The other is extrapolating from a tiny number of questionable provenance.

    It's like coming up with numbers for global warming by taking one thermometer reading each year for a decade in one place and then expecting people to believe the line you draw going off the chart that appears from practically nothing.

    You have 'hard numbers' for something like 0.000000006% of JWs and think you can work out what they all had for breakfast from it.

    Finally, you ignore the maori element which I seem to remember someone saying made up 20% of the people and have a predisposition to anemia - but you want to make out this small group had a lower than average rate so that you can slap a "conservative" label of confidence on it.

    Your numbers and methodology are not convincing. I'd use the word laughable but you'd do your prima-dona routine and start claiming you'd been insulted again.

    When someone raises something worth responding to I'll do so. Otherwise I think I’m done here.

    Any specific questions?

    That's the ticket ... declare anything you don't want to answer as not worthy of answering.

    "Thus solving the problem once and for all"

    "but ..."

    "ONCE AND FOR ALL!!"

    (Futurama)

  • adamah
    adamah

    Marvin said-

    My extrapolation is not of New Zealand versus Western nations, though I don’t see how that would change anything given a general parity of healthcare. My extrapolation is of New Zealand versus all other nations, and in this comparison, if anything, we should expect incidents in New Zealand to be less because of the state of healthcare in that nation.

    More arguments from (personal) ignorance? You're right: YOU don't see that which you don't know, since you don't even KNOW that you don't know. Your ignorance (honest or feigned) is no excuse, since it's tantamount to arguing that an ostrich is correct in concluding a threat goes away if he doesn't think the threat exists by not seeing it.

    (And before anyone argues the point, yes, I am fully aware that ostriches don't bury their heads in the sand: it's an old-wives tale).

    For ONE, it's not about "general parity of healthcare", but the likely invalidity of the assumption that all humans respond to anemia treatment in the same exact manner. They DON'T. Even in severe anemia, many patients who refused treatment SURVIVED (although generally at a greater cost, with more complications). Some died, and some survived: WHY? Is whatever factor that allowed some to survive appicable to other patient populations?

    THAT'S merely ONE factor you're willing to accept by extrapolating, when the truth is YOU DON'T KNOW; hence science errs on the side of safety, and doesn't allow such assumptions to be made without supportive evidence.

    Fact is, it's an utterly absurd and non-scientific assumption to make, since IT IS KNOWN that there's greater biodiversity amongst various ethnic groups and peoples, eg the Navajo tribes are simply more prone to diabetes, and they respond quite differently to standard treatment than other groups. In fact, the connection between disease and ethnicity is so strong, it has spurred a separate sub-discipline within medicine called 'epidemiology'. You really should look into it if you want to write on medical topics, as a familiarity with it's principles can only help you.

    Simon said-

    Wrong numbers, whether accidentally or intentionally can do great damage. It's all about credibility and trust and ultimately can undermine other issues that have heavy basis in fact and evidence. If we get blood death figures so wrong then what about child abuse figures too? "Oh, those people are just bitter and make up false claims - look, here is the proof I can point to". It matters.

    What's a bit worrisome is Marvin's dogmatic stance, publicly refusing to see what many have pointed out is clearly a problem from a biostatistical standpoint (which IS the only thing that matters in the real World, since someone's wants, desires, and wishes doesn't count for a hill of beans: everyone needs to be able to PROVE their CLAIM(S) WITH SUPPORTIVE DATA, since 'positive' claims require proof to convince others as to WHY they should accept (and not reject) a claim).

    The other intellectually-honest option is to simply weaken the claim, by couching it in 'weasel words' (as suggested most-recently by Ruby456). This is a skill, in itself, as anyone who's taken a course in writing for medical literature knows. Language matters, and yes, standards exist for writing for journals, just as standards exist for handling of data.

    Yet Marvin stubbornly refuses to even acknowledge that a problem exists, as if he's seemingly hell-bent on living up to the GB's claim of "evil apostates who are willing to mislead".

    A dozen of such wild-ass fabrications later (eg Anon's hacking into JW systems, report of JW being bankrupt (!), AAWA outing apostates via FaceBook, blaming WT for attacking sites (when it was due to the admin's refusal to patch a KNOWN vulnerability), and you can see why some JUST MIGHT start to wonder what's going on?

    Who's ignoring the evidence, again?

    Simon said-

    I really don't see any problem with someone saying 'oh yeah, I hadn't taken those factors into account ... thanks' which garners more respect than stubbornly sticking to a debunked theory and insulting or belittling anyone who questions it.

    Heck, that's exactly WHY I've started threads here on my blog articles: knowing that there's SOME mistakes I've made, and someone else (hopefully) will catch them and more importantly, let me know. I've changed and added, on the basis of comments. To do anything OTHER than that is foolish arrogance, and only harms the author.

    Anyone who's had to defend their thesis before a panel of professors is familiar with this level of academic scrutiny, and MOST candidates are smart enough (with healthy egos, i.e. non-paranoids) to not interpret it as a personal attack against THEM (as if the big meanie professors were attacking their thesis to get THEM, rather than to challenge their IDEAS).

    Unfortunately, many are quite unable to discriminate between their ego and their beliefs, hence react to ALL challenges in a paranoid fashion as if it's a personal attack.

    Adam

  • Marvin Shilmer
    Marvin Shilmer

    -

    “I am happy that you are so convinced but I think you're numbers don't make sense and your attitude and approach to people asking questions reduces confidence still further.”

    Simon,

    My attitude and approach to people asking questions is to cull through the mire and respond to what appears to me as substantive issues. I don’t waste my time responding to anything else unless someone points to something specific and requests response. I’ll respond to these inquiries as time and circumstance allows. If you or anyone else finds that offensive or somehow putting-off, I’m sorry. But this is how I use my time. If it upsets you then you’re upset. I’m happy for readers to make of this whatever they will.

    When someone raises something worth responding to, I think I’m done here.

    Any specific questions?

    Marvin Shilmer

Share this

Google+
Pinterest
Reddit