King’s Cross “Safe” Injecting Centre
Fails to Reduce Drug Overdose Deaths

Lucy Sullivan

 

The overwhelming conclusion to be drawn from the four official Reports on the Sydney Medically Supervised Injecting Centre (MSIC)1, which together cover its operation from its opening in May 2001 to April 2007 (six years), is that the unavailability of heroin is of far greater significance in preventing drug overdose deaths than the availability of a “safe” injecting facility — in fact the latter shows no measurable effect at all.
This outcome should be of considerable importance to the future of the Centre, and for any plans for replication elsewhere, in that preventing heroin deaths was a dominant political argument for its establishment, and the one that held most sway with the public. It is now clear that the Sydney MSIC was established on false premises.
 The reports appear meticulous and trustworthy in their collection and analysis of their data. Less trustworthy than the full reports are the Executive Summaries which, without actual misrepresentation, avoid citing the most telling findings of ineffectiveness, and paint a rosier picture than is warranted.

Clients, Attendance and Injecting Habits
In order to understand the failure to reduce drug overdose deaths, one needs to take into account the general pattern and reach of client registrations and attendance at the Centre, and also the characteristics of client injecting behaviour.
From May 2001 to April 2007, 9,778 injecting drug (ID) users registered at the MSIC. This makes an average of 138 registrations per month. The rate was higher in the early days of operation — 210 per month from May 2001 to October 2002, but only 86 from January 2005 to April 2007. Visits for injections over the six year period totalled 391,170, making a daily average of 181 visits. In the initial reporting period — May 2001 to October 02 — the daily average was 105, but doubled to about 200 thereafter.
While these figures sound impressive, they represent only a tiny percentage of the clients’ drug use. Thirty-eight percent of total clients reported on registration that they injected at least once daily, with 47% less than daily (10% quite infrequently); for 5% these data were missing. More detailed information on injecting frequency is provided in Report No 1: Among registrants from May 2001 to December 2004 (75% of the total), 9% mostly injected more than three times daily, 15% two to three times daily, 15% once daily, and 36% less than weekly; 6,379 of the then 7,379 (86%) registrants had injected in the last month.
Although not highlighted in the Reports, these figures indicate that most clients did most of their injecting elsewhere than the MSIC. For the period of the first Report, the average of 164 visits per day averages at only 29 total visits per client over a period of approximately 1,300 days, which represents a visiting rate of once per 44 days. As 38% of clients reported injecting at least once daily, and 61% at least weekly, obviously most client injecting went on elsewhere.
For the MSIC to fulfill its aim of preventing overdose deaths, it would seem important that clients should do the major part of their injecting at the Centre. If injecting at the Centre is for the most part episodic, then the major part of their injecting practice will continue to lay them open to the risks they suffer if no such Centre is in existence.
The final Report states that only 23% of clients lived locally. This fits the picture created by the attendance figures, that the majority of usage was of an ad hoc convenience nature, serving the ID user passing through and preferring somewhere private to inject; and that it was not integrated into the injecting lifestyles of the majority of clients in such a way that it could be expected to impact significantly on their health and safety.

Prevention of Drug Overdose Deaths
Kings Cross, the chosen site for the MSIC, by no means hosts the majority of opioid deaths in New South Wales. Over the period May 1998 to April 2006, there were 1,652 morphine deaths in New South Wales, of which 211 (13%) occurred in the Kings Cross vicinity and 1,441 (87%) in the rest of New South Wales.
Over the full six years of the four Reports, 2,106 “overdose-related events” were managed at the Sydney MSIC, the majority (93%) involving heroin or another opioid. There were no deaths. This, however, does not mean that the MSIC actually prevented this number of drug overdose deaths, as ambulance and hospital emergency departments are also called on in overdose emergencies. To be convincing on this count, a fall in the rate of overdose deaths in the area of the MSIC from before to after its establishment is required, and further, this needs to be compared with trends in non-MSIC areas, lest some more general factor determining a decline had come into play subsequent to its establishment. This the Reports very properly do.
A significant complicating factor of this nature was the advent of a “heroin drought” at the end of the year 2000, which had the effect of reducing opioid-related deaths in New South Wales from a rate of 31 per month for the four years prior to May 2001, the opening date of the MSIC, to 9 per month over the six succeeding years.
The Report offers three measures of the effectiveness of the MSIC in combatting overdose deaths, one direct and two indirect: trends in opioid-related deaths in the Kings Cross area as compared with trends in the rest of New South Wales from May 1998 to April 2006; trends in ambulance attendances for suspected opioid overdose in the Kings Cross vicinity (postcodes 2010 and 2011) as compared with trends in the rest of New South Wales also over the period May 1998 to April 2006; and trends in opioid presentations at the Emergency Departments of St Vincent’s Hospital and Sydney Hospital, which serve the Kings Cross area, over the same period.
Opioid-related Deaths: There were 1,652 opioid-related deaths in the whole of New South Wales in the period May 1998 to April 2006. Of these, 211 (13%) occurred in the Kings Cross area and 1,441 (87%) in other parts of New South Wales. Of this total, 1,104 occurred in the three years prior to the heroin drought and the opening of the MSIC, and 548 in the five years thereafter.
The MSIC opened just when the drought was at its most extreme. It briefly eliminated deaths in the Kings Cross area entirely, where the rate had hitherto varied between 9 and 1 per month, and in the rest of New South Wales they plunged to a low of 6 in May 2001 after varying between approximately 45 and 15 over the previous three years.
In the pre-MSIC period, deaths in the Kings Cross area averaged 4 per month, and this fell to 1 per month in the post-MSIC period. The equivalent figures in the rest of New South Wales were 27 and 8, respectively. The falls in average monthly deaths were statistically significant in both cases, being by about 70%, and did not differ significantly in the two locations.
This means that, on the basis of these statistics, the presence of the MSIC in Kings Cross cannot be credited with any preventative effect on overdose deaths subsequent to its establishment there. The fall was due to the heroin drought.
Ambulance attendances: Trends in ambulance attendances for suspected opioid overdoses during MSIC opening hours for the period May 1998 to May 2006 for the Kings Cross area (postcodes 2010 and 2011) are compared with those in the rest of New South Wales. From a total of 20,409 attendances in New South Wales, 12,646 occurred during MSIC opening hours, 1,485 (12%) in the Kings Cross area, and 11,161 elsewhere in New South Wales. Thus attendances in the rest of New South Wales dwarfed those in Kings Cross by a factor of about 8 : 1.
There were statistically significant decreases in average monthly attendances both in the Kings Cross area and in the rest of New South Wales between the years before the opening of the MSIC and thereafter. In Kings Cross they fell from an average 27 per month to 9, and in the rest of New South Wales from 188 to 73. The former represents a decline of 68% and the latter of 61%, and although the difference is not great, it is statistically significant. However, because the falls in overdose death rates were not significantly different, this means that ambulance attendance was as effective in preventing death as was immediate attention in the Injecting Centre.
Emergency Department Presentations: Of a total of 1,558 opioid presentations at St Vincent’s Hospital (82%) and Sydney Hospital (18%) in the period May 1998 to April 2006, 834 occurred in MSIC subsequent operating hours, 75% of which arrived by ambulance. From before to after the opening of the MSIC, there was a decrease in average monthly presentations from 11 to 7. As there are no comparison figures for the rest of New South Wales this can only be taken as an effect of the heroin drought. In any case, it does not bear directly on actual deaths.

Conclusions
The Reports were unable to identify any preventative effect of the Sydney MSIC on drug overdose deaths. While this outcome is clearly stated in the body of Report 4, it is omitted from the Executive Summary, which is all that will be read by most people, including politicians and policy makers.
The Executive Summary misleadingly reports only the finding of a slightly greater fall in ambulance call-outs to overdose events in the MSIC area than in the rest of New South Wales, with the implication that rapid treatment on site in the MSIC would be more effective in preventing deaths than when treatment is delayed till an ambulance arrives (or till transport by whatever means to an Emergency Department). As the actual death rate figures reveal, this is not the case.
The assessment is made that, “It is likely that substantial proportions of overdoses managed at the site would have resulted in significant morbidity had they occurred elsewhere, and that approximately half would have otherwise occurred in public places.” Morbidity, of course, is not mortality, and as regards morbidity, “likely” and “would have” will not do as hard evidence. A little later in the Summary it is suggested that on-site management is more effective in preventing mortality also, although the Reports produced evidence to the contrary.

Why is the Injecting Centre ineffective?
As already pointed out, the statistics of total registered client numbers and total visits in six years of operation suggest a very random and occasional usage of the MSIC which it would be hard put to have any significant effect. They indicate that the MSIC was largely used by clients on an ad hoc or purely convenience basis, and not as a means of improving their health and safety as ID users.
Support for this interpretation is provided by the statistic that only 23% of registrants lived locally (in postcodes 2010 and 2011) and that 72% were not users of local ID user services (e.g. free needle distribution points). Thus 75% lived elsewhere, were only passing through, and used the MSIC as a stop-gap which would have no permanent effect on their ID habits.
At the time of the opening of the MSIC, a significant factor for ID users welfare came into play, namely, the extreme fall in availability of heroin at the end of 2000, and an only partial recovery thereafter. As a result, it has become absolutely clear that overdose deaths fall when there is a fall in heroin use, not as a result of the occasional use of safe injecting centres — and it seems that the voluntary use of such centres will only ever be occasional. This outcome effectively disproves the much vaunted claim of the drug legalization lobby that making the illegal drugs freely available would dissipate their harmfulness.
Given the belief that withdrawal of a heroin habit is a dangerous and difficult process, on the face of it, it is surprising that it was apparently achieved so readily by so many when heroin became unavailable; and the rebound, with a return of supplies, was to much less than the former level. The so-called trial of the Sydney Medically Supervised Injecting Centre supports the common sense view that restriction of access to drugs, in which illegality plays a major role, is crucial to the prevention of drug overdose deaths.

ABOUT THE AUTHOR:
Lucy Sullivan, PhD, has written on families, taxation, child-rearing and education for the Centre for Independent Studies.

ENDNOTES
1. The Sydney MSIC Interim Evaluation Report No.1: Operation and Service Delivery; Interim Evaluation Report No.2: Evaluation of Community Attitudes towards the Sydney MSIC;         Interim Evaluation Report No.3: Evaluation of Client Referral and Health Issues (March 2007); Evaluation Report No.4: Evaluation of service operation and overdose-related events (June 2007)