NDARC Monograph No. 31 (1997)
In 1995, the National Drug and Alcohol Research Centre was commissioned by the Commonwealth Department of Health and Family Services to revise the Illicit Drug Reporting System (IDRS) first implemented in Australia in 1989. The revised system aimed to provide a co-ordinated approach to the monitoring of data associated with the use of opiates, cocaine, amphetamines and cannabis. This information was to act as an early warning indicator of the use, availability and related health problems of the main drug categories so that responses could be implemented before significant problems developed. Data needed to be sensitive enough to alert the existence of emerging problems of national importance rather than describe phenomenon in detail, given that the main audience was the Ministerial Council on Drug Strategy and the National Drug Strategy Committee. In addition to suggesting areas for more detailed data collection, it needed to provide data in a timely manner, collect comprehensive, comparable national data, include representative coverage of the relevant populations, be simple to operate, be linked to a mechanism which could commission the collection of more in-depth data and be cost effective.
NDARC conducted a 12 month trial in Sydney of four methods for a revised IDRS. These included: (1) qualitative key informant interviews with professionals working in the drug field and illicit drug users; (2) a quantitative survey of injecting drug users; (3) ethnographic research among heroin users in southwest Sydney; and (4) an examination of other early warning indicators. Issues such as drug of choice, route of administration, type and number of illicit drug users, intensity of illicit drug use, drug-related problems, manufacture and distribution of drugs, price and purity, and reactions to government strategies were considered. In addition, a number of methodological issues, such as the degree of convergent validity, feasibility, cost and time were examined to determine the most appropriate ways to measure trends in illicit drug use.
Key informant study
This component involved interviews with key informants who were illicit drug users both in and out of treatment, and professionals recruited from health, law enforcement, research and outreach. All had good knowledge of current trends in illicit drug use and first hand contact with illicit drug users. A targeted sampling framework was used to recruit a broad range of key informants, where information from secondary data sources, previous research on illicit drug use in Sydney and consultation with researchers currently in the field were used to identify areas from which key informants could be selected. Interviews were conducted with key informants in both groups and on an individual basis to compare the efficacy of these techniques. Forty-four key informants participated in one of five group discussions, 12 of whom were also interviewed individually by phone prior to the groups. Most participants (90%) were contacted in the two week period following the groups for evaluation purposes. In addition, a summary of the group findings were sent to participants to indicate the extent of their agreement with the comments made. After revision, a second trial of key informant methods occurred from March to April 1996, to test the final methodology. It should be kept in mind that while attempts were made to substantiate key informant reports, these results are still a subjective profile of drug use and availability based on the perceptions of key informants.
There was general agreement among most key informants that the following changes in illicit drug use had occurred in Sydney during the previous 12 months:
Survey of injecting drug users
Injecting drug users (IDU) were targeted in the survey as they are a sentinel group for drug trends. Quantitative interviews with 152 injecting drug users from the inner city (IC) and south western Sydney (SW) were conducted. Entry criteria was having injected at least monthly in the 6 months prior to the interview. Subjects were recruited using multiple methods including advertisements in rock magazines and via needle exchange outlets. They were interviewed at places convenient to them, such as coffee shops and hotels.
Several points arise from this survey that deserve mention. First, subjects from both the inner city and southwest Sydney were clearly polydrug users, familiar with a wide range of drugs and drug trends. The mean age of both samples was lower than in previous studies, which is consistent with the perception of subjects that there are more younger heroin users entering the market. The higher proportion of female injectors in the inner city sample also merits mention. Traditionally, samples of IDU have been approximately two thirds males. The data may indicate a trend towards more female injecting, which is consistent with the perceptions of the inner city sample themselves.
A finding that has major implications is the large proportion of subjects from both areas who had made a transition from amphetamine injecting to heroin injecting. This may well be the source of the new, younger heroin users. The high rates of the injection of "non-injectables", methadone syrup and benzodiazepines, among both samples should also be noted.
In-depth ethnographic interviews and observational fieldwork designed to elicit information in relation to drug use patterns, local drug market conditions and emerging trends were undertaken in Cabramatta, Sydney, over a three month period between September and December 1995. A total of forty subjects participated in a tape-recorded interview and observational data in the form of field notes were collected on each subject and on the nature, type and level of interactions between subjects in the study. In addition, subjects were required to complete a short structured questionnaire on local drug market conditions.
The principal findings concern a relatively hidden group of young, recent initiates to heroin use, the emergence of a street-based injecting culture in the study area and the apparent resilience of the local drug market to pressures from law enforcement. Specifically, the results of this preliminary study suggest that heroin users in Cabramatta may be significantly younger, have lower levels of education and higher levels of unemployment, be more likely to be female, less likely to be Anglo-Australian, more likely to have initiated heroin use by smoking rather than parenteral use, more likely to be involved in crime (including drug distribution and sales activity), more likely to engage in high risk injecting episodes and to have little or no experience of treatment, than those encountered in the literature.
A range of early warning indicators available on an annual basis were sought which would complement and validate the original data, including general and special population survey data, and health and law enforcement data. Ideally, these indicators also needed to: be nationally available, be in an accessible format, not require any special collections, include 50 or more cases, be brief, be collected in the main study site, and include details on the main illicit drug types. Except for AOD telephone advisory data and law enforcement statistics, few indicators were collected annually, satisfying the early warning criteria. The following summary of indicators was thus expanded to include data which met the remaining criteria and improved our understanding of the illicit drug situation.
In terms of previous survey findings, the national household surveys indicate that cannabis and amphetamine have remained the most popular illicit drugs during the 1990s, particularly among males and young adults. Cocaine and heroin were less commonly used by the general population. An increase in MDMA use was noted between 1991-1993. Targeted surveys of homeless and detained youth found that they were large consumers of illicit drugs, more so than secondary school students. Persons from non-English speaking backgrounds were unlikely to use most illicit drugs. Cannabis use was higher among Aboriginal persons compared to the general population. Injecting drug users were consistently found to be large polydrug users.
Data on the characteristics of clients presenting to drug treatment agencies in NSW and nationally show significant increases for those with primary cannabis and amphetamine problems. However, opiates remained the most frequently reported illicit drug problem. Heroin overdose, both fatal and non-fatal, was relatively common and was often related to concurrent polydrug use. The Alcohol and Drug Information Service received the most number of phone inquiries relating to cannabis during 1995, followed by heroin then amphetamine. The number of phone calls by clinicians to the Specialist Advisory Service fluctuated during 1995, although cannabis, heroin then amphetamines were again the most common purpose of calls. A national review of risk behaviours for HIV infection found that the self-reported rate of IDUs sharing needles remained consistently low (below 50%) since 1989, dropping to below 20% in 1994. The proportion of prisoners reporting a history of injecting was around 50%, and the considerable risk behaviour that occurs in these environments was not declining. Higher rates of HCV and HBV were found, compared with HIV, particularly among IDU populations.
From NSW police statistics, large fluctuations in the quantity of cannabis leaf and heroin seized by the DEA were recorded during 1995. Fewer and more consistent seizures were made of cocaine, amphetamine and MDMA. Mean purity levels of around 50% for both cocaine and heroin were recorded during 1995, with some fluctuations. Purity levels were lower and more stable for MDMA (28%) and amphetamines (5%). According to DEA price statistics, decreases in gram and ounce purchases of heroin, and gram purchases of cannabis heads occurred during 1995. Increases in the price of gram purchases of cocaine, street gram purchases of amphetamine, and ounce purchases of cannabis heads and leaf were also noted.
Areas requiring further investigation
Overall, the four studies in the IDRS trial highlighted the need for further investigation of the main illicit drug types and several high risk populations. These included:
Excluding research currently in progress, the following topics still require further investigation: transitions from amphetamine to heroin injection; HCV transmission; patterns and harms associated with the use of MDMA; psychostimulant drug markets; developing interventions for psychostimulant problems; developing interventions to reduce benzodiazepine and methadone injection; and examining illicit drug use and related harms among inner city ATSI and street youth.
A comparison of methods
The feasibility of the four methods tested in the trial differed, with each having various strengths and weaknesses. For the key informant study, individual interviews worked better than groups in that they allowed greater flexibility, participation, time for discussion and substantiation of information. Data from individual interviews were also easier to transcribe and to analyse. Collecting trend data from key informants who were illicit drug users proved problematic. They were less able to conceptualise drug trends than professionals working in the area, less likely to substantiate sources of information, more difficult to recruit, to organise into groups and to follow-up over time. The most efficient and valid way to collect information from illicit drug users was to individually ask them about recent, personal behaviours, as was done in the survey component. The key informant study was useful in that it allowed valuable and timely input on drug trends from a wide range of professionals working in the drug field. It utilised rapid qualitative research methods which provide an overview of issues requiring more in-depth research.
Problems with the IDU survey involved inadequate sampling of some populations due to time and cost constraints, issues easily resolved in future monitoring studies of IDU. As expected, IDU appeared well placed as a sentinel group for trends over a wide variety of drug classes. In particular, accessing nontreatment populations of IDU as was done in the present study will allow appropriate responses to be implemented before serious problems develop. This component represents a central element of a monitoring system as few comparable surveys of IDU are undertaken on a regular basis in Australia.
The ethnographic research tapped into an existing ethnographic study of drug use and economic behaviour among heroin users in southwest Sydney, thereby minimising both the usual delays involved in gaining access and establishing relations of trust and credibility with subjects, and costs involved in conducting such research. While in-depth information on drug practices in this area was obtained, overall the study was time consuming and expensive, and required skilled field staff dedicated to penetrating and immersing themselves in this environment, issues not insignificant in assessing the feasibility of an ethnographic monitoring component to an IDRS.
Few additional indicators were useful from an early warning perspective. Some work is needed to establish the regular collection of other early warning indicators, such as methadone and arrestee urinalysis data, emergency room drug mentions, ambulance data on overdoses and some police data.
Some types of IDRS data were best collected in particular formats. Overdose data, treatment admissions, drug prices and purity data were more reliable from existing indicator sources, specialist studies and the survey of IDU, rather than from key informant reports. The ethnographic component and some key informants (such as covert police officers) were able to provide most information on the manufacture and distribution of illicit drugs. Ethnography provided the most detail on reactions to government strategies such as law enforcement.
Finally, a degree of convergent validity was noted between the four methods, particularly on heroin trends in southwest Sydney, an area measured by all methods. Overall, the IDU survey, key informant study and analysis of existing indicator data were the cheapest and easiest to conduct and covered a broader range of illicit drug use in comparison with the ethnographic study.
Recommendations for a revised IDRS
Given that multiple methodologies are preferable for measuring drug trends, the following qualitative and quantitative measures combining drug user, health, outreach, law enforcement and research perspectives are recommended for a revised IDRS:
Ethnographic research is optional depending upon the need and the availability of resources.
Results from the trial suggest that both the survey of injecting drug users and key informant study could be conducted consecutively on an annual basis. This would provide greater confidence in the identified drug trends, ensuring that they are not just transitory observations and are potentially of national significance. It would also allow information from a range of sources to be obtained. Quarterly summaries of existing health and law enforcement data would allow trends to be tracked during the year. A national IDRS could be implemented in a selection of different sized capital cities, as these are places where illicit drug users tend to congregate. A national IDRS report would summarise trends in each State, providing individual State reports as attachments, and be presented annually to the NDSC and MCDS.
In addition, National Drug Trends conferences need to be resumed to allow further discussion of the issues raised in the IDRS, including prioritising future research topics, identifying areas of national significance and discussing other potential responses.
Overall, the revised IDRS would provide a more accurate, efficient and standardised system of data collection than has previously occurred. While the present trial represents a simplified version of an early warning monitoring system, these methods can be further expanded and refined over time to achieve a greater level of sophistication. While similar data collection systems currently exist in North America, it is imperative that a comprehensive early warning monitoring system be established in Australia to allow prompt responses before the development of serious drug problems. Such a system would also provide a more cost-effective mechanism for addressing Australia's drug problems.