Table 2 Summary of characteristics of included studies
Study/references | Cost year/currency | Country | Form of the evaluation | Perspective taken | Treatments evaluated | Model population | Time horizon | Study designa | Outcome measure | Societal costs | Health states |
|---|---|---|---|---|---|---|---|---|---|---|---|
Journal-based publications | |||||||||||
Barnett [11] | 1996 (US $) | US | CEA | US Healthcare provider | METH versus Drug-free treatment | Hypothetical cohort of 1000 25Â year old heroin users | Life-time | Markov | Cost/LYG | No | NR |
Barnett [12] | 1998 (US $) | US | CUA | US Healthcare provider | BMT versus MMT | Hypothetical cohort | 10Â years | Dynamic model | QALY | No | 9 states based on HIV status (uninfected, asymptomatic HIV +ve, AIDS) and drug user status (IDU not on tx, IDU on tx, non-user) |
Masson [13] | NR (US $) | US | CEA | US Healthcare provider | MMT versus Enriched Detox | Based on 179 patients in a RCT | 10Â years | Markov | LYG (base case) QALY (SA) | No | Alive and dead |
Negrin [15] | NR/(Euro (€)) | Spain | CEA | Drug Treatment centres | 3 MMT programmes (high, medium, low intensity) | Based on 586 patients in drug tx centre | 1 year | Bayesian | CEAC & CEAPF | NR | NR |
Schackman [16] | 2010 (US $) | US | CUA | Societal | Office-based BUP/NAL versus no treatment | Hypothetical cohort of stable patients on treatment for 6Â months | 24Â months | Cohort simulation | Cost/QALY | Patient costs | In tx off drugs, Off tx off drugs, In tx on drugs, Off tx on drugs |
Sheerin [17] | 1999/2000 (NZ $) | New Zealand | CEA | New Zealand Healthcare | MMT | Hypothetical cohort of 1000 IDU | Lifetime | Markov | Cost/LYS | No | HCVÂ +Â ve, no HCV, Chronic HCV, HCC, Compensated LC, Decompensated LC, Liver transplant, Death |
Stephen [18] | 2011 (US $) | US | CUA | Societal | MMT versus theoretical course of Deep Brain stimulation | NR | 6Â months | Decision analytical | QALY | Yes (productivity losses, crime costs) | NA (decision tree) |
Tran [19] | 2009 (US $) | Vietnam | CUA | Vietnamese Health Service | MMT versus non-MMT | Based on 370 drug users from a cohort study | 1Â year (5% discounting) | Decision tree | Case of HIV averted QALY of MMT versus non-MMT | No | NA (decision tree) |
Zaric [20] | 1998 (US $) | US | CUA | US Healthcare provider | Expanding MMT programme (HIV prevalence rate of 5% & 40% versus 15% baseline) | Hypothetical cohort | 10Â years | Dynamic model | Cost/QALY & cost/LYG | No | 10 states based on HIV status (uninfected, asymptomatic HIV +ve, AIDS) and drug user status (IDU not on tx, IDU on tx, non-user) and AID death |
Zaric [21] | 1998 (US $) | US | CUA | US Healthcare provider | Expanding MMT programme (HIV prevalence rates of 5,10,20, 40%) | Hypothetical cohort | 10Â years | Dynamic model | QALY and LYG | No | 10 states based on HIV status (uninfected, asymptomatic HIV +ve, AIDS) and drug user status (IDU not on tx, IDU on tx, non-user) and AID death |
Zarkin [22] | 2001 (US $) | US | CBA | Societal | METH | Hypothetical cohort of 1 million adult patients | Lifetime | Monte Carlo simulation model | Cost/benefit ratio | Yes (productivity losses, crime costs) | Heroin non user & not in tx, Heroin user and not in tx, In tx, Incarcerated heroin user, Incarcerated non-user |
Miller [14] | NR/(Canadian $) | Canada | Cost Comparison | Societal | MHPP versus non-MHPP | ≥20 years old with > 5 year history of injecting heroin, to inject heroin at least daily, and to have previously failed MMT | 5 years | Monte Carlo simulation model | Total cost over 5 years | Yes (criminal activity costs) | NA |
HTA-sourced models | |||||||||||
Adi [23] | 2004 (GBP £) | UK | CUA | NHS & Societal | NTX versus standard psychosocial care | Hypothetical cohort | 1 year | Decision tree with Monte Carlo simulations | QALY | Yes, in a secondary analysis | NA (decision tree) |
Connock [1] | 2004 (GBP £) | UK | CUA | NHS & Societal | MMT versus BMT versus Placebo | Hypothetical cohort | 1 year | Decision tree with Monte Carlo simulations | QALY | Yes, in a secondary analysis | NA (decision tree) |
Schering-Plough [24] | 2004 (GBP £) | UK | CUA | NHS & PSS | Maintenance versus no drug tx, BUP versus no tx, BUP versus METH | NR | 1 year` | Decision tree with Monte Carlo simulations | QALY | NR | NA (decision tree) |
SMC [25] | NR/(GBP £) | UK | CUA | NHS & Societal | BUP/NAL versus METH, BUP or no treatment | NR | 1 year | Decision analytical | QALY | NR | NR |
Abstracts only | |||||||||||
NR/(US $) | US | BIM | US Healthcare provider | BUP/NAL film versus BUP/NAL tablets | Patients initiating treatment for opioid dependence | 5Â years | Markov model | Cost impact comparing 100% on BUP/NAL film versus 100% on BUP/NAL | No | NR | |
Fowler [28] | NR/US ($) | US | CUA | NR | MMT versus BMT | Hypothetical cohort of opioid-dependent pregnant women | NR | Decision analytical model | QALY | NR | NR |