Rural Wastewater Monitoring: The Australian Experience

As flagged at the outset of this chapter, the value of concentrating our discussion upon America and Australia lies in the fact that both these large countries use traditional research methods to monitor substance use across very large geographical areas. The WWA monitoring system that Australia has recently established would be applicable for the examination of rural drug-related issues not only for the American context, but also for other large nations, such as Canada, India, China, Russia and Brazil.

Australia's monitoring system has only just begun to demonstrate the potential utility of WWA in rural areas. But it has made more progress on rural monitoring than the European system operated by SCORE under the aegis of the EMCDDA. This is mainly because of necessity: Australia's 24 million people are spread over 7.7 million square kilometres, which comprises 75% of the landmass of Western Europe.

With sufficient funding it is feasible that SCORE could increase its representation of rural areas in future years to increase confidence that its current samples reflect national, rather than urban trends. Given that the Australian monitoring system is national and longitudinal, it can be differentiated from other multi-sites studies that have been conducted in other regions, such as France (Nefau et al. 2013).

The Australian Criminal Intelligence Commission commenced funding for the National Wastewater Drug Monitoring Program (NWDMP) in 2016. This program builds on the larger-scale sampling undertaken by SewAus,3 an Australian Research Council Linkage Project led by researchers at The University of Queensland. The NWDMP is a collaboration between the University of South Australia, Adelaide, and the University of Queensland, Brisbane that draws on the strengths of both pioneering teams of WWA in Australia. Both the NWDMP and SewAus are supported by local water authorities, which collect and transport samples to the university laboratories, following accepted protocols (see Chapter 2).

Findings of the Australian National Wastewater Drug Monitoring Program in 2017

The NWDMP reports are published three times each year on data from approximately 50 sampling sites (the number fluctuates slightly) that collectively service over 12 million people, more than half the Australian population (ACIC 2019). The number of samples collected from each site in each data collection wave ranges from four to seven. The analysis of the samples and writing of the reports takes three to four months. This demonstrates the speed with which WWA data can be collated, analysed and reported - astonishing for research conducted at a national level.

In terms of national trends between 2016 and 2017, the NWDMP estimated that 8.4 tonnes of methamphetamine was consumed annually in Australia. This is in addition to 3.1 tonnes of cocaine, 1.3 tonnes of ecstasy and 0.8 tonnes of heroin (ACIC 2018). Comparing this against national seizure data, the ACIC (2018) calculated that law enforcement agencies seized approximately:

  • • 25% of the heroin estimated to meet national demand; and
  • • 40% of the methamphetamine needed to meet national demand.

The amount of ecstasy seized equalled annual consumption. And seizures of cocaine exceeded the total weight required for national demand.

What did the NWDMP analyses reveal about rural substance use? The NWDMP uses a binary variable for remoteness: capital cities and all other areas, which are defined as 'regional areas' (ACIC 2018, 1). Because this categorisation is so broad - for example, encompassing large suburbs - it is minimally useful in examining rural trends. We will return to this limitation presently.

The NWDMP reports identify the trends of the capital cities. The locations of the other sites are de-identified, a practice that is consistent with WWA ethical guidelines (Prichard et al. 2016). However, the results from each location are revealed in confidence to the participating municipalities. This means that local authorities have access to trends in their own community.

Figure 4.1 presents findings from the NWDMP on methamphetamine and fentanyl for the Australian States and Territories.4 We chose these substances as examples because of the harms associated with their misuse.

Given the limitations of the wide definition of 'regional', no conclusions can be drawn from these data with respect to rural substance use in Australia. Compared with regional areas, Australian capital cities have a lower average consumption rate of fentanyl. But, within 'regional areas', it is unclear whether fentanyl consumption differs in any way between large suburbs and very remote places.

Nonetheless, the data show the potential utility of WWA for rural areas. First, they show how readily drug consumption trends can be compared across multiple sites within one country. This feature of WWA can assist to identify specific 'hot-spots' or locations where substance use is unusually high.

Two good examples can be seen in Figure 4.1. The first is the estimated consumption of methamphetamine in site 066, in Victoria (VIC). This location recorded methamphetamine consumption that was two to three times

Meso Applications of Wastewater Analysis

103

Estimated Consumption for December 2017 in Mass Consumed per Day (Left Axis) and Doses per Day (Right Axis) per Thousand People, by Jurisdiction

Figure 4.1 Estimated Consumption for December 2017 in Mass Consumed per Day (Left Axis) and Doses per Day (Right Axis) per Thousand People, by Jurisdiction

Capital cities are presented in white and all other areas ('Regional') are presented in red.

Source: ACIC (2018,25,29).

higher than all other Victorian sites. The second example concerns the estimates of fentanyl consumption in New South Wales (NSW). Sites 016 and 082 were estimated to consume in excess of five times the average of all sites nationally.

This level of geographical granularity is not available in other monitoring systems, including the NSDUH and the NDSHS. These surveys also cannot estimate consumption over a specific time period; they respectively record the self-reported retrospective of substance use in the previous month (NSDUH) and the previous year (NDSHS).

A critical point to underscore is that Figure 4.1 does not present any patterns over time, only the results of one wave of data collection (which represents the average of between four and seven samples at each site). Since data collection takes places three times each year for regional sites and six times per year for capital city sites, each site will in time be able to develop relatively fine-grained trend data. If the program continues, within 10 years each regional site could have 30 data points and 60 data points for the capital city sites. By contrast, over the same time period the NSDUH would have ten data points and the NDSHS three or four.

Potential Improvements for Rural Monitoring in Australia

Could the Australian monitoring program be expanded to include more sites and to achieve good representation of rural areas? Managing the monitoring program is more difficult in rural areas because of, among other things, shortages of personnel at WWTPs to conduct sampling and the costs of providing portable sampling equipment. Nonetheless, there are reasons to believe that the monitoring program could be extended to more rural areas. We base this argument on the fact that, at the time of writing, SewAus had sampled from approximately 130 sites across Australia, including the 50 sites already included in the NWDMP.

Tables 4.3 and 4.4 provide information about the numbers of sites in each jurisdiction, the approximate population size of each catchment, and how the sites could be classified according to the Australian remoteness code (major city, inner regional, outer regional, remote and very remote).

Catchment population is estimated by Australian Bureau of Statistics Census 2016. Population estimates are missing for an additional 20 sites. Actual population estimates are not shown to protect the anonymity of the communities where sampling occurs.

Table 4.3 shows that the NWDMP has good, albeit uneven, representation in all jurisdictions in proportion to their total populations. Eighteen of the sites service large catchments of over 200,000 people, the largest serving approximately 2.2 million people. Half of the sites (65) service populations of between 5,000 and 80,000. Sixteen of the smallest sites treat wastewater from less than 5,000 people.

Jurisdiction

(Total Population, Thousands)

Sewerage Catchment Population Bracket* (Thousands)

<5

>5-20

>20-80

>80-200

>200-500

>500-2,200

N

Aus. Capital Territory (397)

-

-

-

-

1

-

1

New South Wales (7,480)

-

3

6

4

2

3

18

Northern Territory (229)

2

1

4

-

-

-

7

Queensland (4,703)

11

12

14

5

3

1

46

South Australia (1,677)

1

3

3

1

1

1

10

Tasmania (509)

2

3

4

-

-

-

9

Victoria (5,926)

-

3

4

1

1

2

11

Western Australia (2,474)

-

2

3

-

1

2

8

Total

16

27

38

11

9

9

110

Source: Adapted from SewAus.

Table 4.4 SewAus Wastewater Sampling Sites 2018, by Remoteness and Average Population

Sites

Missing

Population Data

Average Catchment Population (Excluding 20 Missing Cases)

Major cities

41

2

387,303

Inner regional

32

5

27,622

Outer regional

40

5

19,387

Remote

11

2

8,855

Very remote

6

6

-

Total

130

20

-

Source: Adapted from SewAus.

Table 4.4 aggregates the sites according to their remoteness code, which is the same coding used by the Australian general population survey, the NDSHS.

The 41 sites located in major cities clearly service many times the number of people served by sites in other settings, with an average catchment population of 387,000. Nonetheless, wastewater sampling is occurring in 89 locations in regional and remote parts of Australia.

At the time of writing, population estimates were not available for the six very remote sites and two remote ones. However, the remaining nine remote community's average population was 8,855 people. Reference to raw data shows that three of the remote communities had populations under 2,000 people. A further three have populations between 5,000 and 9,000. The last three service between 10,000 and 20,000 - two sites in Queensland and one in the Northern Territory.

As discussed in Chapter 2, WWA calculations are based on assumptions relating to the 'standard' metabolic functioning of the human body. Although the majority of individuals metabolise drugs in a 'standard' fashion there are some individuals who metabolise drugs to a greater extent than 'standard' and others who metabolise drugs to a lesser extent than 'standard'. In large populations, these individuals do not have a significant statistical impact upon the reliability of drug consumption estimates. In small populations, by contrast, the statistical influence of individuals may lead to greater uncertainty in drug consumption estimates. Therefore, for values estimated from small populations larger ± values are likely applicable. To date the WWA field has not developed clear guidelines about the empirical validity of analysing wastewater data from small communities. Castiglioni and Lor (2016, 36) suggested that 10,000 might represent the lower level for reliability.

Applying this guide to the SewAus sites, only three of the remote locations would be higher than the 10,000-population reliability guideline. All

Meso Applications of Wastewater Analysis

107

of the sites in major cities, inner regional and outer regional places exceed this number.

This means that each of these eligible sites could be compared against each other with an acceptable level of reliability three times per year. This would be extremely difficult to achieve using survey methods in any setting. But the challenges in smaller rural communities would be particularly acute; the problems associated with recruiting participants in such settings would be compounded by survey fatigue. In other words, even if the barriers to recruitment, discussed above in 4.1.1, could be overcome for the first data collection survey, it is unlikely that a sufficient sample size could be recruited ever)' four months or so.

Limitations Relating to Jurisdictional Comparisons of Remoteness

Could Australia's six states and two territories be compared according to the five remoteness categories (major cities, inner regional, outer regional, remote and very remote)? Earlier (4.1.3) we noted that, due to an insufficient sample size in some rural areas, such an analysis could not be achieved even when data from three of Australia's general population surveys was combined (Roche and McEntee 2017). Table 4.5 presents current SewAus sites in each jurisdiction ranked by remoteness categories.

There are good prospects for comparing jurisdictions according to major cities, inner regional and outer regional areas. However, clearly

Table 4.5 SewAus Wastewater Sampling Sites 2018, by Remoteness and Jurisdiction

Major

City

Inner

Regional

Outer

Regional

Remote

Very

Remote

N

ACT*

2

-

-

-

-

2

NSW

14

6

3

-

-

23

NT

-

-

4

3

-

7

QLD

14

10

19

4

-

47

SA

4

2

3

1

-

10

TAS

-

6

3

-

-

9

VIC

4

7

5

-

-

16

WA

3

1

3

3

6

16

Total

41

32

40

11

6

130

* The Australian Capital Territory is solely the location of the national capital, Canberra, and consequently the whole jurisdiction falls into the category 'major city'.

Source: Adapted from SewAus.

WWA data are not collected from many very remote locations. Six are in Western Australia, but as indicated in Table 4.4, the population sizes of these catchments are presently not known so it is unclear whether sampling reliability guideline requirements could be met. Better prospects exist for the comparison of 'remote' locations by jurisdiction. We highlighted above that 3 of the 11 remote sites have sufficient population sizes for WWA.

In short, with publicly available information we were not able to determine whether any very remote communities exist in Australia where WWA could be reliably conducted. It is also unclear whether enough suitable 'remote' sites exist with which to make meaningful comparisons of the jurisdictions.

This area of uncertainty highlights an important potential limitation of WWA. It means that data on substance use in some remote and very remote locations can only be captured with traditional methods. The other implication is that WWA may not be able to circumvent the sorts of problems encountered by Roche and McEntee (2017), which means that jurisdictional comparisons of remote and very remote locations may not be achievable with WWA.

 
Source
< Prev   CONTENTS   Source   Next >