1. Health
Download icon

Building a microsimulation model of heroin use careers in Australia

  1. Alison Ritter  Is a corresponding author
  2. Nagesh Shukla  Is a corresponding author
  3. Marian Shanahan  Is a corresponding author
  4. Phuong Van Hoang  Is a corresponding author
  5. Vu Lam Cao  Is a corresponding author
  6. Pascal Perez  Is a corresponding author
  7. Michael Farrell  Is a corresponding author
  1. Drug Policy Modelling Program, National Drug and Alcohol Research Centre, University of New South Wales, Australia
  2. SMART Infrastructure Facility, University of Wollongong, Australia
Research article
Cite this article as: A. Ritter, N. Shukla, M. Shanahan, P. Van Hoang, V. Lam Cao, P. Perez, M. Farrell; 2016; Building a microsimulation model of heroin use careers in Australia; International Journal of Microsimulation; 9(3); 140-176. doi: 10.34196/ijm.00146
14 tables

Tables

Table 1

Mutually exclusive states and their definition.

State Definition
Abstinence (S1) Individuals in this state are not using heroin but have used heroin at some previous time.
Irregular use (S2) Individuals in this state use heroin irregularly, as defined as 1 to 3 days a month, or a maximum of 1 or 2 days a week. They do not ever develop dependence.
Regular/dependent heroin use (S3) Individuals in this state use heroin regularly (defined as more than 3 days per week in the last month) and they are not in receipt of any type of treatment.
Withdrawal (S4) Individuals in this state receive withdrawal treatment for their heroin dependence. Withdrawal treatment is concerned with neuro-adaptation reversal, involves about 5–7 days care (in an inpatient or outpatient setting) and includes medications to manage symptoms, supportive care and case management. In this model it is characterised as residential withdrawal (for the purposes of unit costing), and excludes long-tern tapered methadone withdrawal.
Residential rehabilitation (RR) (S5) Individuals in this state receive residential rehabilitation for their heroin dependence. RR is concerned with behavioural change across all life areas, including relapse prevention, psychological wellbeing, physical health, nutrition etc. It is provided in residential settings, and an ideal treatment program is 6–9 months long although many will leave in the first week. While some RR in Australia also provide OST as part of the program, as characterised here this RR does not include OST.
Opioid Substitution Treatment (OST) (S6) Individuals in this state receive Opioid Substitution Treatment for their heroin dependence. The provision of a legal, safe opioid (either methadone or buprenorphine), dispensed daily or less frequently with take-away doses; requires a prescriber and attendance at a pharmacy (primary care or clinic settings). The model does not distinguish between methadone and buprenorphine.
Counselling Only (S7) Individuals in this state receive counselling for their heroin dependence. Provision of psychological therapy only, on an outpatient basis (weekly or fortnightly) with case management.
In-prison no treatment (S8) Individuals in this state are in prison; and have a history of heroin use.
In-prison treatment (OST) (S9) Individuals in this state are in prison, have a history of heroin use, and are in receipt of OST in prison.
Table 2

Initial population by state.

State Starting number
S1 abstinent 22,000
S2 irregular ongoing use 53,000
S3 regular, dependent use 16,500
S4 withdrawal 20
S5 Therapeutic community 80
S6 OST 17,500
S7 counselling 100
S8 prison not in treatment 1,100
S9 prison in treatment 1,100
TOTAL starting population 111,400
Table 3

ATOS transition probabilities.

Current state
State 1 State 2+3 State 4 State 5 State 6 State 7 Sum
Previous state State 4 0.1822 0.4186 0.0000 0.1124 0.2209 0.0659 1
State 5 0.4383 0.2222 0.0370 0.0000 0.1358 0.1667 1
State 6 0.4015 0.4891 0.0219 0.0146 0.0000 0.0730 1
State 7 0.4203 0.3188 0.0580 0.0435 0.1594 0.0000 1
Table 4

Final base transition probabilities used in the model.

From: To: S1 S2 S3 S4 S5 S6 S7
S1 Abstinence 1.0
S2 Irregular use 1.0
S3 Dependence 0.04 0.01 0.03 0.90 0.02
S4 Withdrawal 0.18 0.52 0.05 0.20 0.05
S5 Residential rehab 0.40 0.40 0.10 0.10
S6 OTP 0.40 0.55 0.02 0.01 0.02
S7 Counselling 0.30 0.50 0.05 0.05 0.10
Table 5

Summary of data sources for LOS estimation.

State LOS data sources
S1: Abstinence ATOS Dataset, (Hser, 2007; Nosyk et al., 2013; Shah, Galai, Celentano, Vlahov, & Strathdee., 2006; Simpson & Marsh, 1986; Termorshuizen, Krol, Prins, & Van Ameijden, 2005)
S2: Irregular use ATOS Dataset, (Coffin & Sullivan, 2013; Grönbladh & Gunne, 1989)
S3: Regular use ATOS Dataset, (Bell et al., 2006)
S4: Withdrawal ATOS Dataset
Alcohol and Other Drug Treatment Services National Minimum Data Set (AODTS-NMDS) Dataset
S5: Residential Rehab ATOS Dataset
AODTS-NMDS Dataset
S6: OTP ATOS Dataset, (Burns et al., 2009)
S7: Counselling ATOS Dataset
AODTS-NMDS Dataset
S8, S9: Prison states All LOS determined by crime type (probability) and then the average sentence for that crime type (BOCSAR Court Data Report, 2012, see NSW Bureau of Crime Statistics and Research. http://www.bocsar.nsw.gov.au).
Table 6

Crude mortality rates for individuals based on the state and length of stay.

State: To State: From LoS CMR (Crude Mortality Rate) (all cause) Source
S1 Any Fixed rate across LoS 5.3 /1000 PY (5.0–5.6) (Degenhardt et al., 2009)
S2 Any Fixed rate across LOS 5.3/1000 PY (5.0–5.6) (Degenhardt et al., 2014; Degenhardt et al., 2009)
S3 S1, S4, S5, S7 1st week
2nd week
17.4/1000 PY (11.7–25.0)
20.1/1000 PY (13.8–28.4)
(Degenhardt et al., 2009)
S8 1st 2 weeks out of prison 59.5/1000 PY (41.3–83.6) (Degenhardt et al., 2014)
S2, S6, S9 For the rest of LoS 11.5/1000 PY (11.1–12.0) (Degenhardt et al., 2009)
S4 Anywhere Fixed 6.0/1000 PY (5.7–6.4)
S5 Anywhere Fixed 6.0/1000 PY (5.7–6.4)
S6 S3, S4, S5, S7, S8 1st week of LoS
2nd week
39.5/1000 PY (31.9–48.8)
17.0/1000 PY (11.8–23.6)
S9 1st week of LoS 10.9/1000 PY (4.0–23.8) (Degenhardt et al., 2014)
S6 All other times 5.6/1000 PY (5.2–5.9) (Degenhardt et al., 2009)
S7 S3, S4, S5, S6 Fixed 6.0/1000 PY (5.7–6.4)
S8 Any Fixed 2.7/1000 PY (1.2–2.2) (Larney et al., 2014)
S9 Any Fixed 0.7/1000 PY (0.3 to 1.2)
Table 7

Heroin treatment costs.

State tSate description Per day costs (AUD)
S1 Abstinence $0.00
S2 Irregular Use $0.00
S3 Dependent Use No Treatment $0.00
S4 Withdrawal $410.77
S5 Residential Rehab $139.36
S6 Opioid Treatment Program $27.09
S7 Counselling $18.49
S9 Prison treatment $224.81
Table 8

Other healthcare costs.

State State name Per month costs (AUD)
S1 Abstinence $649.00
S2 Irregular Use $649.00
S3 Dependent Use No Treatment $441.00
S4 Withdrawal $648.00
S5 Residential Rehab $392.00
S6 Opioid Treatment Program $860.00
S7 Counselling $566.00
S8 Prison No Treatment $0.00
S9 Prison Treatment $0.00
Table 9

HIV and HCV treatment costs in AUD (per person per year).

HIV treatment costs $20,678
HCV treatment costs $15,526
Table 10

Crime costs (AUD) per offence.

Crime type Police costs Social costs
Homicide and related offences $ 13,484.80 $2,731,649.19
Acts intended to cause injury $3,038.83 $2,650.68
Sexual assault and related offences $ 4,178.39 $3,959.32
Dangerous or negligent acts endangering persons $2,469.05 $2,904.72
Abduction and other offences against the person $4,273.35 $2,904.72
Robbery or extortion and related offences $7,312.18 $5,179.91
Unlawful entry or burglary or break and enter $3,988.46 $1,060.68
Theft and related offences $3,228.76 $2,883.83
Fraud or deception and related offences $3,133.79 $506.05
Illicit drug offences $2,753.94 0
Prohibited and regulated weapons offences $3,228.76 0
Property damage and environmental pollution $2,564.01 $1,060.68
Public order offences $3,703.57 $97.16
Traffic and vehicle regulatory offences $1,139.56 $97.16
Offences against justice procedures $3,703.57 $97.16
Miscellaneous offences $2,012.00 0
COURT COST (Fixed) $7,712
Table 11

Capturing the resources relating to those who are charged and found not guilty and those who are found guilty but do not go to prison.

2012 AUD – weighted by distribution of offences Applied to
Police costs $3,090 • those found not guilty
• those found guilty and not sentenced to prison
Court costs $7,712 • those found not guilty
• those found guilty and not sentenced to prison
Social costs $1,144 • those found guilty and not sentenced to prison
Table 12

Comparison of computed mortality rates from simulation and input rates.

State Mortality rate from simulation Input mortality rate (CI interval 95%)
S1 0.0057 0.005 – 0.0056
S2 0.0060 0.005 – 0.0056
S3 0.0106 0.0111 – 0.0120
S4 0.0047 0.0057 – 0.0064
S5 0.0061 0.0057 – 0.0064
S6 0.0057 0.0052 – 0.0059
S7 0.0056 0.0057 – 0.0064
S8 0.0024 0.002 – 0.0037
S9 0.00038 0.0003 – 0.0012
Table 13

Comparison of computed imprisonment rates from simulation and input rates.

From Simulation From the Input data
To S8 To S9 To S8 To S9
S1 0.0022 - 0.0026 -
S2 0.0286 - 0.0282 -
S3 0.0404 0.0456 0.0408 0.0408
S4 0.0414 0.0429 0.0408 0.0408
S5 0.0400 0.0430 0.0408 0.0408
S6 - 0.0784 - 0.0816
S7 0.0427 0.0395 0.0408 0.0408
Table 14

Comparison of computed imprisonment rates from simulation and input rates.

Simulation output Input rates
S1 S2 S3 S4 S5 S6 S7 S1 S2 S3 S4 S5 S6 S7
S1 - - 1 - - - - S1 - - 1 - - - -
S2 1 - - - - - - S2 1 - - - - - -
S3 0.16 - - 0.02 0.03 0.76 0.03 S3 0.154 - - 0.015 0.048 0.750 0.033
S4 0.18 - 0.52 - 0.05 0.20 0.05 S4 0.18 - 0.52 - 0.05 0.20 0.05
S5 0.60 - 0.33 - - 0.03 0.04 S5 0.60 - 0.34 - - 0.04 0.04
S6 0.39 - 0.56 0.02 0.01 - 0.02 S6 0.40 - 0.55 0.02 0.01 - 0.02
S7 0.30 - 0.50 0.05 0.05 0.10 - S7 0.30 - 0.50 0.05 0.05 0.10 -
S8 0.30 - 0.60 - - 0.10 - S8 0.30 - 0.60 - - 0.10 -
S9 - - 0.10 - - 0.90 - S9 - - 0.10 - - 0.90

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)