Melbourne Injecting Room - ineptitude or premeditated fraud?
. . . false conclusions that don’t follow from the evaluation's own data
The Melbourne injecting room commenced operations in July 2018, with an evaluation of its first 18 months being released mid-2020.
It is an understatement to say that its conclusions regarding the number of 21-27 lives saved in those first 18 months did not follow from the data. It would be more correct to say that the estimates were pulled out of thin air while being simultaneously entirely falsified and made laughable by the data carefully collated in the body of the evaluation - after all, all their relevant population data showed actual increases in deaths in the vicinity of this injecting room.
So here is the whole sordid story on ‘saved lives’ and if this were an isolated incident you could almost laugh it off. But the very same kind of gross error (are they just inept?) also happens with each of the major evaluations and studies done on the Sydney and Vancouver injecting rooms, which along with Melbourne are the most studied facilities in the world. So why is there this pattern of ineptitude, or is it premeditated fraud? And why is it in common across studies done on these facilities?
The 2020 evaluation analysed the MSIR’s performance against 6 legislated objectives (see pages x-xiv)
1. Reduce heroin deaths
2. Referrals to treatment and other services
3. Reduce ambulance and hospital attendances
4. Reduce discarded needles on streets
5. Improve public amenity
6. Reduce the spread of blood-borne viruses
Addressing the first objective of reducing heroin deaths, which is arguably of greatest concern for the public, the reviewers wrote:
“The MSIR trial has supervised 116,802 injections (96.6 per cent of which involved heroin) and responded to 2,657 overdoses, with no fatalities. Compared with other people who inject drugs, MSIR clients are significantly more likely to have recently injected in high-risk settings, as well as to have recently experienced a non-fatal overdose, a known predictor of fatal overdose. Prior to registering, more than half of MSIR clients had overdosed and nearly half had witnessed an overdose.
Of those who do attend the service, the nature of the overdoses is significant, and without intervention it is likely that many would have died or been permanently injured.
In the first 18 months of operating, there were 271 extremely serious incidents that required the opioid reversal agent naloxone. Many more required oxygen and measures to keep the airways open, potentially saving additional lives and avoiding harms associated with lack of oxygen to the brain. Advice provided to the Panel from an experienced medical practitioner consulted for the review was that ‘the [overdoses] are at least as acute an emergency as those we receive in an [emergency department]’. Of those who attend the service, the nature of the overdoses is significant, and, without intervention, it is likely that some would have died or been permanently injured.
The harms associated with overdoses can be profound; some are permanent. The facility has the appropriate equipment and MSIR staff are well trained and clearly demonstrate the capacity to respond, manage and administer interventions required to avoid death or further harm. Staffing levels ensure timely responses.
The MSIR has advanced its critical objective to save lives. While these results are not observable in coronial data, the Panel assesses that without responses to overdoses provided by the MSIR, the number of deaths could have increased during the trial period.
Modelling allows an estimate of the number of lives that the MSIR may have saved and, while there are different ways to model this, using conservative estimates, these data suggest that between 21 and 27 deaths were avoided over the 18 months of this review. This does not include the prevention of permanent disability including acquired brain injury.”
This analysis focuses on the MSIR’s first objective, that of reducing heroin-related deaths.
ESTIMATES OF DEATHS AVERTED DERIVED FROM OVERDOSE NUMBERS
The MSIR reviewers calculated that the facility had averted 21-27 deaths, according to their modelling, which is accorded no explication in their report.
It is highly likely that their estimates of averted deaths have been derived from the number of overdoses in the MSIR – whether they have been derived from the total number of overdose interventions or from the lesser number of naloxone administrations is not apparent.
It is beyond dispute that the estimates bear no relation to the realities of deaths averted at the community level, as they have acknowledged above. Figure 17 on page 45 of the review reveals no impact by the MSIR on heroin-related deaths at the community level. Focusing on heroin deaths within 1 kilometre of the MSIR, the results are graphed below.
Broadening the focus to the entire host Local Government Area, Figure 15 from the MSIR Review shows the following:
Broadening the scope to those suburbs from which the highest number of MSIR registrants were drawn:
It should be noted that the latter figure excludes two quarters’ data for St Kilda, which is not shown.
The MSIR review, on page 44, records 176 deaths for the rest of Victoria in the 12 months prior to the opening of the MSIR, and 175 in the year following.
Clearly, at the community level, the MSIR failed to make any observable impact on heroin-related deaths.
It can be concluded that estimates of 21–27 deaths averted by the MSIR are based on indefensible and inept assumptions, most likely on bloated overdose numbers within the facility. These unprecedented numbers of overdose in the facility thereby must necessarily be analysed, which will be done in a later Substack post.
We further note that numerous government-funded evaluations of the Sydney MSIC failed to find any positive impact on overdoses at the community level. For example, p 55ff of the MSIC Evaluation 1, completed in 2003, found no positive impact on overdose deaths in surrounding postcodes, and pp 54,55 found no reductions in presentations for overdose at nearby hospitals.
The MSIR review found no reductions in hospital presentations for heroin overdose (see p xi).
SO WHAT IS A DEFENSIBLE ESTIMATE?
Australian overdose statistics indicate that the MSIR is not even capable of averting one death per annum.
That is not to say that supervised injecting facilities cannot save any lives – the MSIR did host enough injections in its first 18 months to avert one death, but not enough in 12 months to do the same. This explains why no effect is seen for the MSIR at the community level.
The maths is very straightforward.
It derives from the European Monitoring Centre’s (EMCDDA) 2004 Review of Drug Consumption Rooms http://www.emcdda.europa.eu/html.cfm/index54125EN.html. Their method avoids the error of many other studies which have made the simplistic error of calculating averted deaths from the raw number of overdoses in the supervised injection facility assessed. If the safety of the room is vastly and unnaturally elevating overdoses, artificially inflated overdoses within a facility cannot possibly be defended as the starting point for calculating averted deaths. Yet too often, enthusiastic researchers seeking to promote injecting rooms, see the high overdose figures as an opportunity for demonstrating that many lives have been saved, casting proper methods to the wind.
There is well-utilised Australian data indicating that one in every 100 dependent heroin users die each year from an opiate overdose. So well established is this ratio it has been used to officially back-calculate the number of Australian heroin users in a given year using the number of heroin fatalities for that year.
From this ratio we know that on any of the streets of Australia, one heroin user on average will die for every 109,500 opiate injections. It is calculated as follows.
Dependent heroin users, the ones most at risk of overdose, inject ‘at least’ 3 times a day. This was a fact that formed the backbone of ‘deaths averted’ calculations in the 1st MSIC Evaluation as seen on page 58 of that document. One user will inject 3 times daily for 365 days in the year, or 1,095 times in a year, just as 100 users will inject 109,500 times (3 injections per user per day x 365 days in a year x 100 users) of which one injection will be fatal. The 112,830 heroin injections hosted during the MSIR’s first 18 month evaluation period is minimally more than the 109,500 injections which would have normally been associated with one death.
The aversion of just one death in 18 months accords well with the community-level deaths observed within 1 kilometre of the MSIR, in the Yarra LGA, in the city of Melbourne as well as the State of Victoria, where no observable difference was made at any of these levels.
Calculating deaths averted from overdose numbers within the MSIR - without first comparing MSIR overdoses to other known overdose rates in the community – leaves only two invidious options for reviewers - the claimed 21–27 deaths averted is inept or possibly fraudulent. Governments need to be made very aware of this fact.
For the $6 million spent by the MSIR to save one single life, the Victorian government could provide 73 optimally-funded residential rehab beds for a full year.[1] $6 million can statistically save one life in the MSIR (which can nevertheless be lost tomorrow injecting elsewhere) or alternatively make many users drug-free, given that a residential rehab bed is filled by more than one opiate user in a year. Successfully rehabilitated users reduce the need for police expenditures and ambulance interventions as drug use is ceased altogether.
[1] In August 2018 the NSW Legislative Council’s Portfolio Committee No.2 (Health and Community Services) Report 49 recommended “That the NSW Government significantly increase funding to drug and alcohol related health services” (Recommendation 2). The NADA submission recommended $224.95 of funding per bed day for residential rehabs, which equals $82,106 per annum or 73 bed years for the $6 million to save one life in an injecting room. If patients are offered 6 months of rehab each over 140 users will have been assisted towards being drug-free, freeing them from the morbidity of non-fatal overdoses and freeing the community of crime and public nuisance.