the important risk factors for HIV transmission

cross-sectional studies from other countries and cities in
China have indicated that IDUs were exposed to many of
the important risk factors for HIV transmission, including
sharing needles with poor sterilisation technique and low
condom use during sexual intercourse (Zhang et al.,
2002; Parviz et al., 2006; Perngmark et al., 2004). Yet
overall results in studies have been inconsistent, for
example condom use ranged between 0 and 90%
(Perngmark et al., 2004; Xinhau News, 2008). Variation
in geographical and cultural settings consequently makes
these studies difficult to generalise to the IDU population
in rural China. Furthermore, the aforementioned studies
do not take account for the recent implementation of
Chinese government HIV prevention programmes such
as needle exchange and condom promotion.
This study aimed to provide an up-to-date investigation
of current HIV risk behaviours amongst heroin users
undergoing rehabilitation therapy in rural China, with the
hope of informing the development of future risk
reduction interventions within this population.
METHODS
Setting and population
Once ethical approval from the Chinese University of Hong Kong
and the University of Birmingham had been obtained, data
collection took place at a rehabilitation centre in Dali, Yunnan
Province between November and December 2006. The detoxification
centre itself has over 2,000 beds and caters for the whole of
Dali Prefecture. For the majority of drug users treatment is
compulsory owing to China’s “Zero Tolerance” Policy regarding the
possession of heroin. Treatment at the centre largely consists of
medical therapy, psychological counselling, physical exercise,
manual labour and vocational training.
Procedures
As a cross sectional study, 1026 questionnaires were administered
to heroin users who had a six month history of intravenous drug
abuse prior to admission. Participants were recruited and briefed by
one doctor. Individuals were excluded if they had mental illness or
were deemed inappropriate to participate in the survey for any other
reason by the doctor. If the participant was illiterate, the questionnaire
was filled in by the doctor in the form of an interview. A selfreporting,
structured, anonymous questionnaire was used in order
to gather information on the demographics of the population, heroin
addiction, needle sterilising and sharing behaviour and sexual
health behaviours including condom use. Participants were asked
to reflect upon their behaviours six months before their admission.
The risk of psychological harm was minimal. Participation was
voluntary and written informed consent was obtained.
Data handling and analysis
Data collected were inputted into a SPSS database by university
students under the supervision of a research assistant at the
Chinese University of Hong Kong.
Data were analysed using the SPSS 15.0. Initially data collected
were cleaned and interrogated. Imputation was not used for missing
values as it was felt that they were small in number and potential
biases could be avoided. Preliminary analysis was descriptive in
nature and scrutinised the key demographic features including age,
sex, ethnicity, marital, education and occupational status, heroin
addiction and HIV risk factors including needle sharing, condom
use, sexual activity, sex workers and knowledge about HIV.
The second stage of analysis involved a Chi-squared test of
difference in the proportions and t-test of difference in means (as
appropriate) to investigate demographic features of the population
and the likelihood of using condoms during vaginal sex and needle
sharing. These two key HIV risk factors were chosen as sexual
transmission and needle sharing as they are the two primary modes
of HIV transmission in China in 2007 (Xinhau news, 2008). In
addition “ever sharing needle in the past” was used instead of
“sharing a needle the last time you used drugs” as it was thought
that this variable gives more indication of general behaviour as
opposed to a one-off occurrence. A logistic regression model was
performed to identify whether any of the significant demographic
factors were independent predictors of having unprotected vaginal
sex or sharing needles.
RESULTS
Demographic background of the population
Between November and December 2006, 1026 drug
addicts were invited to take part in the survey at Dali
Drug Rehabilitation Centre. Of the 993 (90.2%) who
agreed to participate, 848 were male, 130 were female
and 15 did not disclose their sex. The average age was
31.4 years (range 13 - 88 years). The non-response rate
to individual questions was between 0 and 13.2%. Table
1 illustrates the demographic characteristics of the
population stratified by sex.
Heroin addiction, treatment and intravenous status
The survey included drug addicts at all stages of treatment
at the rehabilitation centre with their length of stay
ranging from 1 - 24 months. All inmates were receiving
treatment for heroin addiction and on average had been
resident there for 5.7 months. Over 80% had been in
addiction therapy before with the average respondent
having received treatment twice previously. Treatment
was mandatory for 56.5% of the population. The median
age for participants to have their first contact with heroin
was 22 years old (range 9 - 55 years old). Average daily
consumption of heroin was 0.2 - 0.4 g. 81.2% admitted to
injecting at least once a day and the majority (44.6%)
injected 2 - 3 times per day. The mean length of time
spent injecting heroin was 56 months or 4.7 years.