Friday, November 11, 2011

Prohibitionists Drug Free Australia well rebutted

The usual prohibitionists recently commissioned a critique of the Vancouver injecting centre, Insite, but the critique has been shown to be without scientific merit and dependent on false methodology.

Gary Christian
Drug Free Australia (DFA) and its honorary secretary Gary Christian passionately oppose harm reduction measures such as injecting centres. Mr Christian has lately turned his attention to Vancouver's Insite, a facility comparable to the Medically Supervised Injecting Centre in Kings Cross, which itself was previously targeted in a DFA attack based on a particular interpretation of statistics.

This time, DFA's ally the Drug Prevention Network of Canada (DPNC) critiqued a peer-reviewed study published in Lancet that had demonstrated significantly reduced overdose deaths from illegal drugs in Insite's local area. This was unacceptable to the prohibitionists. Their critique, which was not peer-reveiwed,  claimed the original Lancet study was flawed and denied that the centre had saved lives, a claim DFA had also made in Kings Cross. This is remarkable as what these centres DO is professionally treat people who have overdosed, immediately and on site. Getting to the OD victims so quickly means nearly all can be treated simply with oxygen.

It's a bit like saying that people who suffer heart attacks in an emergency ward have worse outcomes than those who have heart attacks at home.

Now the critique itself has been critiqued, and found wanting. The response, by the authors of the original study, maps the many flaws in the DPNC work, starting with the following:
Using BC Vital Statistics data, they argue that overdose deaths increased rather than decreased during the period considered in our study. This apparent discrepancy is explained by several flaws in their analysis. First, our study in the Lancet focused on a defined area of interest in close proximity to Insite that included 41 city blocks... However, the data considered in the... DPNC report examined the entire Downtown Eastside Local Health Area (LHA)—an area that is much larger and includes approximately 400 city blocks.
That's comparing 400 city blocks with 41, a 10x difference. If you ignore the tyranny of distance you can come up with all sorts of wonderful conclusions. The DNPC critique seems to assume that addicts will travel up to 200 blocks or more before they inject. From what I've seen, addicts desperate for a hit can't wait to inject - in Kings Cross they almost run to the MSIC the second they score.

And that's only point 1.

I won't quote the whole document here - if you want to read the full story you can download the easy-to-read 5-page pdf at the link above or, for your convenience, here. If DFA or its allies have a credible rebuttal to this rebuttal, feel free to comment. They seem to have gone pretty quiet on this one though.

8 comments:

Gary Christian said...

Michael

Drug Free Australia has not gone quiet at all. Here is the answer for the edification of your readers.

In the interests of what is now an international drug policy issue I will make some observations on Kerr's and Montaner's misleading reply.

1. Our analysis, despite Kerr and Montaner's erroneous response, does indeed separate the 5 (yes, only 5) intentional/other deaths out of the 155 deaths that Vital Statistics' coroner's data cites for the Downtown Eastside (DTES) sub-area of Vancouver, from the 150 non-intentional deaths which Kerr et al. calculated from in their study for the same sub-area. Anyone can go to the Vital Statistics data we cited in our analysis, see that the 5 years' overdoses of all kinds add up to 155, and then go to Kerr and co's own data which we reference in our analysis and see that they were themselves citing 150 non-intentional deaths for exactly the same sub-area in their study. With the 155 total deaths in the DTES increasing from 27 in 2002 (the year before the Sept 21, 2003 opening of Insite) to 37 in 2005, a startling 37% increase in raw community overdose deaths in the 400 block area around Insite, those 5 intentional/other deaths cannot in any way change the fact that there was an increasing trend of deaths for the DTES. You can slice and dice those 5 intentional/other deaths in any way you like - even if all 5 intentional/other deaths happened in 2005 alone, there STILL remains an increasing trend even after population adjustments. So Kerr and Montaner's claim about our analysis not separating intentional/other deaths from non-intentional deaths is both fatuous and very, very easily disproven (see page 3 par 1 of our analysis for proof). They are left with every sub-area of Vancouver outside their 41 block focus, including the rest of the DTES, with increasing trends in overdose deaths when they told us there were decreases. This by itself makes their study invalid, because they must use the inflated numbers of deaths in 2001 to create their false 'decreases'. It is no small thing to create the appearance of mortality decreases in the larger Vancouver area which was compared in the Lancet study. A reported 9% decrease in overdose mortality for Vancouver could promote the false misconception that Insite also influenced these (false) decreases. So . . . their claim that we did not account for intentional/other deaths is FALSE, and demonstrably so. The only thing we do not know is the year in which each of the 5 intentional/other deaths occurred, and that is inconsequential to our analysis. Their claim that we used raw data is FALSE, and demonstrably so. And Montaner and Kerr's use of such palpably false claims in any debate forum would amount to a clear misrepresentation of the other side's argument.

2. Montaner and Kerr cite daily heroin use figures to say that there was no real change in heroin use amongst all Vancouver heroin users between 2001 and 2005. This is like saying that our plummeting overdose mortality in Australia in 2000-2001 had no connection wih heroin supply, whereas the connection between plummeting deaths in British Columbia, from 417 in 1998 to 170 in 2002 (down from 246 in 2001) was exactly the thesis in two previous journal studies by three of the Lancet article's researchers. Appealing as they do to daily use figures, where many heroin users might be using 4 times daily in1998, but with a 50% reduction in supply be using once or twice daily in 2002, is misleading. The 50% reduction in use will make no difference whatsoever to the graph Montaner and Kerr produced in their response, but it will most certainly roughly halve overdoses and deaths in the same way as overdose deaths here were linked to proportionally reduced supply in Australia. In a debate, the daily use graph would be considered nothing other than a misleading red herring.

Gary Christian said...

Continued from previous comment . . ,
3. Drs Kerr and Montaner criticise our analysis’ failure to adjust for population increases. However, as we have noted elsewhere, “When these increases in overdose deaths are compared against population growth in both Vancouver and the DTES the increases in deaths well overwhelm any changes in population. The Lancet study, at Table 2, calculates a 3% change in Vancouver’s population between 2001 and 2005, yet drug deaths increased by a much greater 14% between 2002 and 2005. The Lancet study calculated an 8% increase in population for the DTES, yet drug deaths increased by 37% between 2002 and 2005. In the scenario where all 5 suicides or unexplained deaths, as discussed previously, occurred in the DTES in 2005 alone, the increase in drug deaths would still be 18%, well beyond the 8% population increase for that sub-area of Vancouver.” These observations appear in the peer-reviewed Journal of Global Drug Policy and Practice (JGDPP)

4. Drs Kerr and Montaner set up a straw man of wonderful devising, falsely alleging that our analysis used raw coroner’s data for a 400 city-block area of Vancouver to question their conclusion that overdoses reduced by 35% in the 41 block area around Insite. We did nothing of the kind. Drs Kerr and Montaner are demonstrably and clearly wrong on two counts: a. Our analysis only used raw coroner’s data to demonstrate that there was an increasing trend in overdose mortality for British Columbia, Vancouver and the DTES from 2002 to 2005, and that therefore any claims to 9% decreases in overdose deaths for Vancouver were false. or that the use of 2001 in a pre-Insite comparison was invalid. In regards to the 41 block area immediately surrounding Insite we in fact used no raw data at all. Paragraph 4 of page 5 of our analysis CLEARLY says:

"However, despite the errors of the Lancet article discussed thus far there is every likelihood that overdose fatalities close to Insite since 2003 could have decreased relative to the previously graphed increases for the rest of the DTES and for the rest of Vancouver."

Our analysis further attributes these expected increases to the changed policing. We stated in the same paragraph as quoted above that:

"The reason for some confidence in this assertion is the major changes in policing and police officer numbers in the immediate area around Insite instituted 6 months before Insite officially commenced operations, with these changes continuing to this day."

Drs Kerr and Montaner are demonstrably incorrect in their accusations about how raw data was used, and failed to note that our analysis contains a map with the exact location of all 89 deaths within the 41 block area of their focus, but also showing that the vast majority of these deaths fall within the 12 block area patrolled by 48-66 extra police deployed since April 2003, indicating that most of these deaths likely happened in the pre-Insite comparison period when these blocks were an open drug scene.

In a debate, these claims by Montaner and Kerr would amount to obfuscation and misrepresentation.

Gary Christian said...

Continued from previous comment . . .

5. Drs Kerr and Montaner infer that they did not know that policing was expanded up to 66 police when Insite opened, up from the 50 police that were operative during the pre-Insite 6 month crackdown they studied in their 2004 journal article (where they reported a 46% reduction in disposed needles in the crackdown area, which they cited as evidence of a clear displacement effect of users and dealers away from the policed area. That they knew of the policing changes is clear from evidence that they had read the police statements of this continuing but ramped-up crackdown post-Insite-commencement recorded in Colin Mangham's JGDPP article from 2007 - see http://www.breitbart.com/article.php?id=cp_f1elbr55d2&show_article=1&cat=0). In a debate this tactic is called . . . . We also note the ad hominem attacks on Dr Colin Mangham. In a debate that is called . . . 'ad hominem attack' when you have no argument.

6. We would love to see what documentary evidence from 2010 Montaner and Kerr could produce to back their inferences that the Australian Medical Association (NSW chapter actually) or Royal Australian College of Physicians had ever sighted, critiqued or rejected our previous analyses of the Sydney injecting room evaluations. In a debate this would be called fabrication if no such evidence eventuated.

7. We find it nigh on inconceivable that while British Columbia had 909 non-intentional overdose deaths between 2001 and 2005, of which 41% were non-injection-related, the DTES had NOT ONE non-injection-related death amongst its 150 non-intentional overdoses. For all of Montaner's and Kerr's assertions that they studied every one of these deaths, our lack of confidence in their ability, on grounds already expressed, to produce a defensible study precludes a ready acceptance of their claim. Explanations about coroners having incomplete information do not properly address the high percentage of the 909 deaths that were non-injection-related in BC.

Gary Christian said...

5. Drs Kerr and Montaner infer that they did not know that policing was expanded up to 66 police when Insite opened, up from the 50 police that were operative during the pre-Insite 6 month crackdown they studied in their 2004 journal article (where they reported a 46% reduction in disposed needles in the crackdown area, which they cited as evidence of a clear displacement effect of users and dealers away from the policed area. That they knew of the policing changes is clear from evidence that they had read the police statements of this continuing but ramped-up crackdown post-Insite-commencement recorded in Colin Mangham's JGDPP article from 2007 - see http://www.breitbart.com/article.php?id=cp_f1elbr55d2&show_article=1&cat=0). In a debate this tactic is called . . . . We also note the ad hominem attacks on Dr Colin Mangham. In a debate that is called . . . 'ad hominem attack' when you have no argument.

6. We would love to see what documentary evidence from 2010 Montaner and Kerr could produce to back their inferences that the Australian Medical Association (NSW chapter actually) or Royal Australian College of Physicians had ever sighted, critiqued or rejected our previous analyses of the Sydney injecting room evaluations. In a debate this would be called fabrication if no such evidence eventuated.

7. We find it nigh on inconceivable that while British Columbia had 909 non-intentional overdose deaths between 2001 and 2005, of which 41% were non-injection-related, the DTES had NOT ONE non-injection-related death amongst its 150 non-intentional overdoses. For all of Montaner's and Kerr's assertions that they studied every one of these deaths, our lack of confidence in their ability, on grounds already expressed, to produce a defensible study precludes a ready acceptance of their claim. Explanations about coroners having incomplete information do not properly address the high percentage of the 909 deaths that were non-injection-related in BC.

The Editor said...

The above comments from Gary Christian are too technical for my poor brain. It seems we have two positions here in flat contradiction about the facts and figures.One of them is right and I suspect it is the version that has been peer-reviewed and upheld in the courts. Applying Ockham's Razor, I still can't accept the DFA position that supplying clean needles does not prevent the transmission of disease, and treating people immediately for overdoses doe not save lives. Better heads than mine are tackling this debate: see http://luckylosing.com/2011/11/16/drug-free-australia-manipulate-misrepresent-data-to-discredit-insite/

Paul Cubitt said...

It is very sad how some organisations gain attention and funding by confusing an issue, that with a little investigation and courage could be properly addressed. The really sad part is that people are dying or having their health and Human Rights compromised because of moral driven ideologies, not science.

It would be nice to see a drug free society, the reality is that every culture through history has used psycho-active drugs and they did not cause their downfall. Plus our life expectancies would be a lot less than they are now.

Paul Cubitt
President
Law Enforcement Against Prohibition (LEAP) Australia

Anonymous said...

The response Michael has linked to already addresses all of gary's points. For example;
Reduced overdoses allegedly caused by policing?

<<>>

Gary borrows heavily from an earlier DPNC report written by his co-author Colin Mangham.
<<< It should be noted, however, that the organization that commissioned Dr. Mangham’s report, the Royal Canadian Mounted Police,
later acknowledged that Dr. Mangham’s work “did not meet
conventional academic standards.”
Not surprisingly, Health
Canada’s Expert Advisory Committee on Supervised Injection Site research chose to ignore Dr. Mangham’s report in their review of existing research on supervised injection sites.

Further, during a recent Supreme Court of Canada hearing focused on Insite, the lawyers representing the Government of Canada were forced to admit that they did not have any credible research to suggest that Insite was not working (i.e., they did not offer Dr. Mangham’s work as evidence).>>>

You can find all of Gary's other arguments addressed and rebutted here; http://www.myphotoart.com.au/downloads/insite-response-to-allegations.pdf

Here's the rub;

<<< In summary, we welcome academic debate, but we stand by
the data presented in our Lancet paper and note that, unlike the report prepared by Mr. Christian and colleagues, our data and methodological approach were subjected to extensive peer
review and published in one of the world’s leading medical journals. The results of our study demonstrate that Vancouver’s
supervised injection facility appears to have had a localized yet
significant effect on overdose mortality in the area of densely
concentrated injection drug use where the facility is located. >>>

Anonymous said...

Hi,

One of my quotes dropped out of the above message.

<<< The impact of policing: The REAL Women/DPNC report suggests that a large-scale policing operation may account for the reported decline in overdose deaths in the Lancet study, and Mr. Christian alleges that our team was aware of this crackdown and committed an act of omission by failing to mention this crackdown in our study.

We are confused by this suggestion for several reasons.

First, documents on the City of Vancouver’s website and a published evaluation of the police crackdown reveal clearly that this policing initiative ended within weeks of Insite’s opening and was not ongoing throughout the study period, as the authors of the
REAL Women/DPNC report suggest.

If this crackdown was the cause of the decline in overdoses after Insite opened, this would imply that:
(1) the police crackdown led to an increase in overdose deaths in the area where Insite would later open;
and
(2) the subsequent decline in overdoses reported in the Lancet paper occurred because this policing initiative ended.

However, various reports and our published study of this crackdown (Wood et al., Displacement of Canada’s largest public illicit drug market in response to a police crackdown.
CMAJ, 2004; 170(10): 1151-1156) demonstrate that this police initiative displaced drug users away from the area where Insite was subsequently located.

Hence, it could be argued that the displacement of drug users away from where Insite was located, prior to its opening, could have served to create a conservative bias in the Lancet study by reducing overdoses in this geographic area before the facility’s opening.

This in turn would have made it MORE difficult to demonstrate a decline in overdose deaths after the facility opened. >>>