Numbers of community based service users were collected from existing reporting systems for NSP sites and MMT clinics in Shuimogou, Shayibake and Tianshan Districts of Urumqi. The costs of implementing harm reduction programmes were collected from financial records and a survey of staff from these same sites. The survey collected information for 2008–2010 for MMT and 2006–2009 for NSP.
Costs were assessed based on establishment costs (equipment, staff training), operational fixed costs (rent, utilities, equipment maintenance, staff salaries) and operational variable costs (in-service staff training and consumables for testing; supplies such as methadone, needles and syringes and condoms). These cost definitions are generally accepted as describing direct costs of services [5
]. Income from national and local government, donor and service user payments covered the costs of providing MMT services.
Costs for MMT were estimated for 2006 and 2007 based on cost per service user derived from the 2008 costs. Costs for NSP in 2010 were estimated based on costs per service user in 2009.
Trends in incidence were measured by BED HIV-1 EIA Assay testing. Blood samples were tested from drug users resident at the Urumqi City Public Security Bureau Detoxification Centre from 2000 to 2010. From 2000–2005, BED HIV-1 EIA Assay testing was carried out on stored blood samples and from 2006–2010 on fresh samples. Obtaining blood samples from all residents in Detoxification Centers is a routine component of admission and required for the regional surveillance system. Residents were informed of the reasons for blood sampling. Informed consent for the use of stored blood samples was not possible to obtain due to the lack of accurate and current contact information for those who had given samples between 2000 and 2005.
To estimate PWID community based population size, study results from previous internal studies were used. In 2010, an internal XUAR Centre for Disease Control research study showed that 68.2% of 2487 drug users in various communities including Urumqi, reported having injected opiates. Therefore, the population size of PWID was estimated based on numbers of registered (10,481) and estimated total drug users (20,000) in Urumqi provided by the XUAR Public Security Bureau. The minimum number of registered PWID was 7,148 (68.2% of 10,481) and the maximum number was 13,640 (68.2% of 20,000). These estimates were used to establish the upper and lower ranges for HIV infections averted in each year of the study.
To estimate HIV infections averted incidence was projected for 2006 to 2010 based on the average rate of increase as measured in samples between 2000 and 2005 using a standard average rate of increase formula [6
–average rate of increase
n–number of years in the series
a0–origin of series (year)
a1–first figure in series (year).
Annual projected incidence results were compared to actual results on tested blood between 2005 and 2010 and the difference between these results enabled estimations of infections averted.
Costs of providing treatment and care were calculated based on personal costs for PLHIV and government supported costs. Personal costs for PLHIV were estimated from survey results of 100 PLHIV in Urumqi (36 participants were using antiretroviral therapy (ART) and 64 were not). Government costs were summated based on testing costs, provision of ART, treatment costs of opportunistic infection and follow-up service delivery. The government supported costs for providing free treatment and other services were added to the personal costs to derive the total direct costs of treatment.
Calculation of annual and cumulative net benefit was based on annual costs of providing harm reduction programmes which were derived from the average costs of treatment services. This provided the basis for the derivation of estimated net financial benefit and a ratio of comparative investment (USD1 spent on harm reduction saves USDX spent on treatment).
All data were input to EpiData (Version 3.1) and analysis was completed with SPSS (Version 11.0). All information used in this paper was provided by the Xinjiang Uighur Autonomous Regional Center for Disease Control. Information ascertained through surveys was provided freely and openly by survey participants.
The XUAR Bureau of Health Ethics Committee was established to support the implementation of the China National Bureau of Health Research Programme. The information collection and methodologies used to support this study were approved by this group.