250,000 Jehovah's Witnesses have died refusing blood

by nicolaou 739 Replies latest watchtower medical

  • besty
    besty

    wheres Nicolau - this is his bloody thread after all

  • bigmac
    bigmac

    oddly enough--i found him on fb this morning and sent friend request---no reply yet

  • Simon
    Simon

    wheres Nicolau - this is his bloody thread after all

    Nice pun

  • steve2
    steve2

    Marvin, your acknowledgement that you simply assumed the incidence of anaemia in New Zealand matched that of other countries shows how simplistic your research endeavors are. New Zealanders have ready access to the health care system for serious medical conditions and - I would venture far readier access than many other countries, including the USA. Besides, it is also highly likely that the incidence of blood disorders from country to country - and even within countries - varies significantly across socioeconomic groups. It is untenable basing hard numbers research findings on the kinds of assumptions you have made. BTW, whether you openly acknowledge your assumptions or not, unsupportable assumptions remain unsupportable.

    This kind of "research" would never make it into a peer-reviewed publication -although there are very likely vested interest publications that would snap it up and publish it - the kinds of publications that are hungry for sensational claims and thatcare little for substantiating those claims.

    You will have a ready market for your preposterously unsupportable claims - just not among those who have a basic knowledge of sound research design and accountable reporting.

  • slimboyfat
    slimboyfat

    Marvin, this should be a fairly straightforward question which you've danced around:

    How many JWs have died? (using known populations and normal mortality rates)

    What proportion of those deaths are you claiming are due to refusing blood?

    Good question I hope Marvin answers.

  • adamah
    adamah

    Finkelstein said-

    Fact is there is no obtainable and accurate way to determine the total deaths occurred by people who were JWS on not taking Blood transfusions.

    That is the final and only truthful indisputable answer.

    No, it isn't; eg how about if someone conducts a study AFTER ARMAGEDDON™ in the New System™ and simply asks the Resurrected Ones™ if they died from refusing a blood transfusion?

    What? Did I say something wrong?

    Adam

  • steve2
    steve2

    I have worked in a metropolitan hospital in New Zealand for over ten years and can categorically state that preventable deaths due to blood-transfusion refusal are extremely rare. All hospitals in New Zealand have a statutory obligation to formally report "sentinel events" which includes deaths due to treatment refusals (although the reporting does not include personal identifiers). Again, the numbers stand out as very low. I personally am very angry about the extrapolations Marvin and his supporters are making. These claims detract from the impressive body of reasoning that exposes the Watchtower's blood doctrine as scripturally and medically unsuppprtable.

  • AndersonsInfo
    AndersonsInfo

    Your comments please on this article:

    USA Today - 5/22/2007

    http://www.usatoday.com/news/health/2007-05-22-death-rates-side_N.htm

    "Grading hospitals is anything but easy, experts say, and sometimes the information can be challenging even for experts to interpret. For example, Florida's health information website, www.floridahealthstate .com, notes that Shands Hospital at the University of Florida has a higher bypass surgery death rate, 6.31%, than the state rate of 2.51%.

    "'Florida's an unusual state," says Curt Tribble, Shands' chief of cardiovascular surgery. "We're one of four transplant centers in the state, the fifth-busiest heart and lung transplant center in the country. We do almost all the Jehovah's Witnesses in the state, and they refuse transfusions. '"

  • adamah
    adamah

    Hi Barbara,

    The article is about how transparency in reported outcomes allows consumers to make better decisions, eg where to get bypass heart surgery. But even there, the article points out that it's not so simple as comparing death rates:

    Grading hospitals is anything but easy, experts say, and sometimes the information can be challenging even for experts to interpret. For example, Florida's health information website, www.floridahealthstate .com, notes that Shands Hospital at the University of Florida has a higher bypass surgery death rate, 6.31%, than the state rate of 2.51%.

    "Florida's an unusual state," says Curt Tribble, Shands' chief of cardiovascular surgery. "We're one of four transplant centers in the state, the fifth-busiest heart and lung transplant center in the country. We do almost all the Jehovah's Witnesses in the state, and they refuse transfusions. We're also a site for people with extraordinary obesity. One guy we did last week weighed 500 pounds."

    So even there, by leaving off the trailing bit about how Shands Hospital also treats obese patients, you've engaged in a bit of JWesque "quote mining" (of course, you provided a link, but simply running it in an article would properly leave you open to some criticism, since everyone agrees that "quote mining" is an intellectually-dishonest practice, attempting to mislead the readers).

    The article continues:

    Most hospital grading systems attempt to balance out these factors. But Tribble says that can be difficult. "The tricky part is that risk-adjustment systems don't perform well at the extremes of risk," he says. "We're a magnet for the sickest patients."

    This is a well-known phenomena in the medical field, where the specialists often handle the most-challenging cases that are referred to them by their non-specialist colleagues, and hence their outcomes will drop; it doesn't mean they're "bad doctors", only that they're willing to deal with the more-challenging cases. That's what specialists are, by definition: doctors who take on advanced and difficult cases.

    But even aside from that, many of the difficult cases are also poor, and hence most reliant on public hospitals; the patients on the "rich side of town" contains heathier patients, and their outcomes are typically better.

    Just saying that all of these are factors, and hence why trying to force the NZ data to extrapolate it Worldwide is questionable, and likely constitutes 'biostatistical malpractice' (that is, if such a tort existed! I don't think practicing amateur biostatistics actually requires a license, LOL!).

    Adam

  • Simon
    Simon

    I have worked in a metropolitan hospital in New Zealand for over ten years and can categorically state that preventable deaths due to blood-transfusion refusal are extremely rare. All hospitals in New Zealand have a statutory obligation to formally report "sentinel events" which includes deaths due to treatment refusals (although the reporting does not include personal identifiers). Again, the numbers stand out as very low.

    I imagine hospitals and surgeons would be keen to report those numbers too as otherwise their death rate rises which is unfair if it was due to the patient refusing treatment.

    Also, direct first hand knowledge is a lot more valuable and convincing.

    The hospitals remind me of school ratings - at some point you are not measuring the ability of a school to educate but actually to attract the brightest potential and reject the lowest performing. Specialist hospitals that take on the most challenging cases could have higher death rates than others but be doing a much better job. Taking figures from one would be inconclusive.

Share this

Google+
Pinterest
Reddit