Posts by Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“Does he say what the outcomes for the 3,000 patients were? How 'increased' was the risk to JWs when they were taken into account? Aren't you talking about 19 out of 3,000 and not 103?”

Simon,

Outcomes for the 3,000? He does not give this because his study was of and only of outcomes and costs in relation to treating patients suffering severe anemia.

Increased risk when the 3,000 are taken into account? What does that have to do with our discussion? We are talking about increased incident (not rate) of mortality of patients with severe anemia refusing blood.

We can talk about rate or rates of mortality until the cows come home and it won’t change the number (the incident) of statistical deaths among JWs in New Zealand with severe anemia refusing blood for the years of 1998-2007.

“It sounds like his model was perfectly designed to give the answer he was looking for which is basically that patients with severe anemia that are urgently in need of a transfusion are better off having a transfusion.”

The authors were looking for a means and method to compare outcomes and cost between treating severe anemia patients with red cell transfusion versus not treating them with red cell transfusion. The only means and method available was to dig into patient records for individuals with severe anemia who refused blood transfusion. There was no way to conduct a controlled trial because it would pose an unethical threat to patients. This left JWs.

Hence the authors scoured patient records from 4 trauma centers and extracted all incidents where patients suffered severe anemia and then separated these into 2 categories. 1 group that accepted red cell transfusion and 1 group that refused red cell transfusion. Will call this Group A and Group B respectively.

Then they examined Group B (the group that refused transfusion) and used it as a model to segregate patients in Group A based on things like general characteristics (i.e., gender, ethnicity, comorbidities, hospital admission type and treatment type). Once this segregation was made the authors randomly selected from the segregated patients of Group A to achieve a matched comparison to Group B so that both groups were very close in all comparisons accept one, and that one comparison was whether the patient accepted or rejected blood.

From that point forward it was only a matter of math, including making adjustment for things like Maori ethnicity given severe anemia was the subject.

The thing about math is that it speaks for itself.

This is what the authors did.

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“So you think they have a big wall or something? They don't have planes? That other countries allow hoardes of blood-refusing-near-death JWs to flood across their borders each day? Pretty much all nations run themselves so I'm not sure what that 'sovereign' part has to do with the price of milk.”

Simon,

What you raise is an important factor to examine. It goes to the question of “What is the population of JWs serviced by healthcare facilities in New Zealand?” This value is important in my extrapolation because it’s a direct corollary with the number of statistical deaths due to refusing blood.

A large part of my research on this article was about this very issue. I didn’t share all this research in order to avoid burying the main point (numbers of needless deaths) in a lot of tedious background information. When I began this project my impression was that surrounding island communities were probably serviced by the New Zealand healthcare system with the effect that I should include these JWs in my calculation. The information I gathered suggested my impression was not the case, and was probably not much of a factor anyway.

What I found was that most JWs in regions surrounding New Zealand who needed healthcare where the “no blood” position was likely to be an issue were not being serviced in New Zealand but, rather, either in their own communities or else Australia. When I looked at the number of JWs in the vicinity of New Zealand that were not being serviced in Australia, the number was so low that it wouldn’t have made much difference anyway. In the end, as stated in a post above, I used the number of JWs in New Zealand to compare against statistical deaths among JWs in New Zealand due to refusing blood.

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“And that is why your application of the results to ALL JWs is giving you an unbelievable answer.

“You are only looking at results from a small subset of the JW population.”

Simon,

Wrong. Records of more than 3,000 JW patients were found by Beliaev. 103 of these fit the model for his study of costs. It just happens to be the case that the model he pulled for his review is the best possible model for determining deaths (not a statistical mortality rate but, rather, statistical deaths!) among JWs for refusing blood because a primary criterion was hemoglobin concentration of =/< 8 grams per dL.

It is the statistical number of deaths over and above the norm among these patients who refused blood that establishes a statistical death per capita value for JWs in New Zealand during years 1998 to 2007. It was 1 JW for every 3848 JWs.

“That means nothing. Are you claiming it's a model? Most likely from what I've seen so far all it really does is automate the calculation.

“Why even do that? I mean, if you have the numbers and the numbers are so hard then you wouldn't be re-calculating anything, ever ...

“Unless you want to play around and do "what ifs" until you get the answer you want. A round 50,000. Nice.”

The process is much simpler than you think.

To extract the number of statistical deaths over and above the norm during years 1998-2007 of patients who refused blood versus those who accepted blood in Beliaev’s study requires a calculation. This calculation gives an aggregate value for years 1998-2007 among JWs in New Zealand. That value is 19.

To determine an aggregate value to compare these deaths against requires establishing a fixed method of determining the number of JWs serviced by the same facilities whose patient records were part of the study. It does not matter the method so long as it is used equally to establish the number of JWs in New Zealand and, later on, the number of JWs in the world. By “method” I mean a rule or set of rules. The method I used was to use the published average number of JWs in New Zealand by Watchtower. To build this as an aggregate value requires a calculation. This calculation gives an aggregate value for years 1998-2007. That value is 126,989.

One problem with the comparison of the 19 with the 126,989 is that the 19 value is gathered from patient records in a minority of healthcare institutions in New Zealand whereas the 126,989 value is of JWs in all of New Zealand. Hence another method (rule) is necessary to make these dissimilar comparisons similar. An examination of demographics in New Zealand and its distribution of healthcare facilities against Beliaev’s data set shows a regionalization that is useful for this purpose. New Zealand has 4 regions (Northern, Midland, Central and Southern) of population and each region has sets of healthcare facilities serving the population. In Beliaev’s study the 4 institutions’ patient records he used were trauma centers. 2 of the 4 had advanced trauma services, 1 of the 4 had district trauma services and 1 of the 4 had basic trauma services. But all 4 were hospitals with trauma services. Hence the rule (method) I applied ignored any and all hospitals in New Zealand that did not have trauma services by assuming none of those hospitals had any deaths of JWs refusing blood. Though the 4 hospitals in Beliaev’s study represented a disproportionately small number of trauma service centers in the 2 healthcare regions in my method (rule) assumed the other trauma centers in the same regions had no deaths at all among JWs for refusing blood. The final rule (method) for finding the corollary of JWs with the hospital patient records used by Beliaev was to use the regional percentage of total population as an indicator of JWs serviced in each region. This is the 57% factor you find in my blog presentation. This factor gives a direct method of adjusting the value of 19 deaths in 2 regions of New Zealand to compare against the total population of JWs in New Zealand in all 4 of its regions. This requires a calculation, and it gives an aggregate value for years 1998-2007 among JWs in New Zealand. That value is 33.

Once you have 2 relevantly similar values determining the per capita deaths is simple division. You can annualize the result to get an annual per capita value. The annual number of statistical deaths among JWs in New Zealand during years 1998-2007 is 1 JW per every 3848 JWs.

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“LisaRose did you notice earlier in the thread where Marvin insisted the statistics were especially reliable because "New Zealand is an island"? What was that all about?”

The fact that New Zealand happens to be an island is only one thing that helps statistically capture the population its healthcare system cares for. This is a factor to consider when attempting to use numbers of deaths to achieve a per capita ratio.

The fact is that New Zealand is a sovereign nation and the Beliaev study presents a data set representing a very large percentage of the nation’s overall population. This is something that’s rare (nearly unique) in data presentations of mortality among Jehovah’s Witnesses refusing blood. It’s useful for constructing an excellent picture of deaths per capita among New Zealander JWs refusing blood over the period of 1998-2007. The annual rate is 1 JW for every 3848 JWs during the period of 1998-2007.

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“I take it to mean he did an excel spreadsheet? So what, so did I. But that means nothing if you are using the study to prove something it doesn't.”

LisaRose,

Yes. The calculator I constructed for this extrapolation is built in Excel.

The study by Beliaev was conducted for a purpose other than how I’ve used the finding. But that does not make my finding weak. Whether my finding is weak depends on how I used data presented by Beliaev.

Beliaev conducted his study to examine cost efficacy between treating severely anemic patients with and/or without red cell transfusion.

I used hard findings published in the Beliaev study against the 10-year population of JWs in the same area to identify the ratio of annual deaths-per-JWs statistically the result of refusing blood. The annual ratio ended up being 3848-to-1. That is to say, based on hard numbers of mortality in the Beliaev study, over the 10-year period of 1998-2007, each year there was 1 death for every 3848 JW publishers. This rate of mortality is not the overall mortality of JWs suffering severe anemia. It is the rate of mortality of JWs suffering anemia over and above the rest of the same population.

Because my presentation is based on actual (statistical) deaths over and above the rest of the population due to refusing blood, and because this mortality is well-documented in the Beliaev study then the statistical annual ratio of 3848-to-1 is beyond dispute.

After establishing the above we have to ask this question: On average is the rest of the world’s population more or less likely to suffer severe anemia (for all causes)?

I don’t see how or why the rest of the world would, on average, be less likely than New Zealanders to suffer severe anemia (for all causes). I looked hard to find this, and could not. The closest I came to finding such a cause is the Maori ethnicity factor. But the Beliaev study made a statistical adjustment for this.

I do see how and why the rest of the world would, on average, be more likely than New Zealanders to suffer severe anemia (for all causes).

So how did I treat this? I assigned the best case scenario to the rest of the world’s population by assuming it not more likely than New Zealanders to suffer severe anemia (for all causes).

This leaves hard numbers that just about anyone can crunch to obtain a statistical number of deaths per annual number of JWs at any time during the period of 1961 to 2011.

Hence the only bad inputs possible are of hard numbers. What hard numbers have I wrongly used as inputs?

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“And this study doesn't address deaths to to accidents where the patient might have lived l if they had accepted blood. I would think the cause of death would be listed as accidental injury, even though the blood issue was a contributing factor. That would, in effect, hide other cases of deaths because of the blood policy.”

LisaRose,

The Beliaev study was not conducted based on “cause of death” on a death certificate. Inclusion was based on patients who presented at or below a certain hemoglobin concentration regardless of why they were experiencing this hemoglobin concentration.

“Marvin seems to think that because he is making conservative estimates in certain areas, he is justified making assumptions in other areas, and I just don't get that. Potentially under estimating in one area does not make it OK to make huge assumptions in other areas, it just makes you wrong.”

What assumptions do you think I’ve made that are likely to statistically drive upward the figure of JWs who died due to refusing blood over the years 1961-2011? Please specify.

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“One question is, how were these patients selected.”

LisaRose,

A set of inclusion criteria was constructed and all patients who refused blood transfusion were identified for inclusion in the study. The JWs who met the criteria but were in palliative care were excluded. For years 1998-2007 there were 10,786 hospital admissions of 3,529 JW patients admitted to 4 hospitals. After reviewing these records it was found that 103 patients met the criterion.

The same criterion was applied to all patients who accepted blood transfusion over the same period admitted to the same 4 hospitals. Patients in this group were selected based on a random drawing at a 1-to-1 ratio with the model group (above) from the same diagnose related sub-category as the model group. This resulted in a randomly selected group that was very well matched across a spectrum of variances to narrow the distinction down to one thing: accept blood or refuse blood.

“Was it all severe anemia cases in those four hospitals, or were there cases of severe anemia in those hospitals that were not selected for some reason?”

All cases of severe anemia were evaluated for inclusion. The only instances of exclusion were patients whose presentation met inclusion criterion but were in palliative care.

“Were patients from other areas moved to these hospitals because of the study, or for some other reason?”

We must assume patients are transferred to hospitals better equipped for critical conditions. For this reason my review ignored any possibility of mortality at a hospital without trauma service because all of the hospitals in this study had facilities and staff to offer trauma services. Because my values are achieved against a hard number of mortalities then, if anything, this method has the effect of understating the actual mortality of JWs in New Zealand with severe anemia refusing blood.

“What has changed due to the increased use of blood substitutes and fractions since the study ended?”

The change you mention occurred in year 2000. Most of the years included in the Beliaev study fall after the change. Hence retrospectively any values achieved would be conservative by comparison.

“What is the death rate in New Zealand for severe anemia overall? That would be a great sanity check to see if we are in the ballpark.”

Based values expressed in the Beliaev study, and after prorating based on assigning mortality values to 57% of the population of the 2 regions his data came from, I get an annual mortality of 0.0000086% of the population. This does no adjust upward for the increased rate of mortality among JWs in New Zealand but because JWs are such a small minority an adjusted value would not add much.

“Why do four hospitals out of eighty treat 57% of the population?”

The 57% figure does not represent 4 out of 80 hospitals.

The 57% figure represents the population of New Zealand serviced by the 4 hospitals whose patient data was included in Beliaev’s study.

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“I looked at CDC numbers for severe anemia, for the US, there were 4,501 deaths in 2010, for a population of 281 million, that seems much lower than your New Zealand anemia rates.”

LisaRose,

I’d like to see the precise source for those figures to see how “severe anemia” is defined by the source. Can you share a link?

The 2 values you offer express an annual mortality of 0.0016018% of the population for “severe mortality.

Based values expressed in the Beliaev study, and after prorating based on assigning mortality values to 57% of the population of the 2 regions his data came from, I get an annual mortality of 0.0000086% of the population. This does no adjust upward for the increased rate of mortality among JWs in New Zealand but because JWs are such a small minority an adjusted value would not add much.

This would make mortality related to “severe anemia” much higher in the United States than in New Zealand by a factor of nearly 200x. This is why I’d like to see the source you cite.

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“Are you saying that out of 80 hospitals, the four hospitals in the study treat 57% of the population? That would mean the other 76 hospitals treat 43% of the population. That doesn't make any sense to me, can you explain?”

LisaRose

No. I am certainly not saying that out of 80 hospitals, the four hospitals in the study treat 57% of the population.

New Zealand has 4 healthcare regions. The 4 hospitals in Beliaev’s data set are in 2 of these 4 regions. The population of those 2 regions is 57% of the total population of New Zealand.

My extrapolation assumes that not one single death of a JW occurred at any other hospital in the same regions of New Zealand as the 4 hospital are in. Mathematically this makes my usage of Beliaev’s data conservative, and this is underscored by the fact that there are several additional hospitals in these 2 regions with trauma services.

Marvin Shilmer

739

250,000 Jehovah's Witnesses have died refusing blood

by nicolaou in

80+ branch davidian cult members died in waco, texas.

250,000+ #jehovahswitnesses have died refusing blood.

stop this #cult!

Marvin Shilmer

-

“Marvin, many many thanks for your well-researched and reasoned article. I am employed by one of the District Health Boards in New Zealand in a community-based mental-health team. I have no statistics to hand but based on anecdotal reports from colleagues who work in the Emergency Department, JWs with sudden massive blood loss (as occurs in motor vehicle accidents) are more likely to die compared to others with similar trauma who are given immediate blood transfusions.

“If it were possible to get statistics on these accident "victims", we would undoubtedly find that the extrapolated 50,000 dead worldwide from non-emergency blood refusal since 1961 would be a low estimate when the blood-refusal deaths from emergency departments are added to the estimates.”

steve2,

Thanks for sharing your view and experience.

Marvin Shilmer