Sunday, July 23, 2006

Deaths dive

The Sunday Age presents a dramatic picture of the decline in heroin-induced overdose deaths. Deaths were 359 in Victoria in 1999, 39 last year and 9 so far this year. The reasons:

  • Education about overdoses, including a big campaign this year via syringe programs.
  • Substitution away from heroin may be in favour of amphetamines.
  • The 2001 heroin drought and a heroin shortage in the first three months this year.
  • The strength or purity of the available heroin being lower than a year ago, and way down on its purity at the height of overdose deaths five years ago.
  • A steady rise in the number of people receiving methadone and buprenorphine treatment, with an increase from about 7000 in 2002 to nearly 11,000 this year.
  • The 'war on terror' which has improved surveillance and international police co-operation thereby reducing heroin imports.

Even with substitutions to drugs such as amphetamines and esctacy this is good news since fatalities have fallen net - amphetamines caused about 14 deaths Australia-wide in 2004.


FXH said...

yes good news - 359 a year to 39 is amazing. mind you the 359 was an abberation wasn't it - from memory.

I'd like to have a squiz at some of the detail and test a few thoughts.

I'm still not clear on overdoses.

hc said...

Fxh, You are right, 1999 was a conspicuously bad year though the graph in the article (not provided in the hyperlink) suggests a fairly gradual buildup to that record level.

Me too! I am not clear on overdoses - why do they suddenly drop off so markedly when supplies get low.

None of the theories I have read about explain this in a way that completely satisfies.

davidp said...

Could the reason be that when heroin is expensive users are more careful with it and use smaller quantities? (I remember my housemate from Japan who was always very careful about quantities of meat (and his recipes often featured small quantities) whereas, comparatively speaking, my approach to cooking was on the way to those meat-loving blokes at the BBQ on the Carlton ad! (though funnily enough in 1990 he had eaten Kangaroo in Japan whereas it was pretty rare in Australia)

I am aware that overdoses can result from mixing drugs but maybe smaller quantities of heroin help?

FXH said...

hc - I'd be keen to look at research (or even ideas) about the people who overdose- age, gender, income, educational level, experienced users, prior disease etc .

Cleary those who overdose are outliers compared to most users, i'd be keen to see detail on what the indicators are for a likely overdose death, especially in 1999 and immediately prior.

And some comparision between overdose with survival vs. overdose and death. Obviously it woudl only be overdoses with a formal medical intervention and not ODs with lay intervention.

Overdose(ers)deaths are a clear marker, it is (unfortunately) definite.

I agree none of the theories of fatal OD work too well although I have some sympathy for the education one - but I'd like to see if thee was some evidence like more willingness to call ambos - that should be an easy stat. to get.

david - the rationing idea makes some sense, but..... It's hard to support rational theories on ODs when rationality would suggest people wouldn't get too close to OD state anyway.

ambos will readily tell stories of being abused by ODs, who were on their way out, for spoiling their high by using NARCAN. And of having jabbed someone with narcan and an hour or so later being called again to same person with new OD.

davidp said...

fxh: Thank you for your comments. I will elaborate a bit more on the idea though - my guess is that there is some randomness whether someone ODs (particularly fatally). So each time someone takes (perhaps a certain amount) they are taking a punt that they don't OD. The idea is that with a drought either:
1. They "punt" on less occasions OR
2. By reducing the quantity, the probabilities of ODing drop considerably.

Either would lead to drops in the number of OD deaths.

FXH said...

david - I don't have any problem with your suggestion of the contribution to lower deaths, but like a lot of explanations it isn't sufficient, not that I think you are claiming it is.

It seems as if there is a likely cluster of explanations, but what fascinates me, and I think harry and probably you, is how do they all work together - like this year - it still seems more "spooky" than explained yet.

I am sure there isn't much randomness in ODs - like a lot of "Emergency" demands there is a shocking amount of "predictability" when decent figures are available to be interrogated intelligently.There are users who never OD and those who no one is surprised when they do.

I'm sure it should be possible to build up a 5 or 6 point check list of indicators and say - this kind of person is in high OD danger this one less so.

I'm a great fan of simple check lists quick and dirty based on evidence.