Long-term Follow-up of Community-based Drug and HIV Prevention Intervention in Yunnan, China Debby Lee Oh UCLA Department of Epidemiology

1
Long-term Follow-up of Community-based Drug and
HIV Prevention Intervention in Yunnan,
ChinaDebby Lee OhUCLA Department of
Epidemiology

• Funded by the National Institute on Drug Abuse
• and the UCLA Global Health Training Program

2
OBJECTIVE

• To investigate the long-term effectiveness
• of a multi-component community-based
• drug and HIV prevention intervention
• implemented in Yunnan, China in 1997.

3
Specific aims

1. Identify and characterize the components of the
   1997 drug abuse intervention that are still in
   operation.
2. Describe the current drug use situation.
3. Measure knowledge, attitudes, and behaviors
   related to drug use and HIV/AIDS.

4
Overview

• PHASE I
• Qualitative study May 2007
• 30 in-depth interviews with village and county
  level leaders
• PHASE II
• Quantitative study June 2008
• Cross-sectional questionnaire of 15-39 year old
  males (n774) and small subset of 15-39 year old
  females (n107)

5
BACKGROUND
6
HIV in China (2007)
700, 000 PLWHA (0.05)
50,000 new infections in 2007
71.3 PLWHA are male
44.7 heterosexual activity
42.0 injection drug use 12.2 homosexual
activity 1.1 mother-to-child
617 methadone clinics 729 needle and syringe
exchange sites
7
Injection drug use (IDU) in China

• Estimated 2.35 million IDU
• 12.3 of IDU infected with HIV
• Concentrated in south and west near Golden
  Triangle and Golden Crescent
• China relatively drug-free 1950s-1980s
• Open door policy opened country to drug trade

8
(No Transcript)
9
Longchuan County
10
Longchuan statistics

• Population 173,000
• Over half are ethnic minorities
• Jingpo
• Dai

11
Longchuan

• Adjacent to Golden Triangle
• Drug trafficking from Myanmar across the Chinese
  border
• First indigenous cases of HIV in China were found
  in 1989, in a county neighboring Longchuan

12
Longchuan statistics (2006)

• 4,600 IDUs
• 59.7 of IDUs infected with HIV
• Of HIV, 75 infections associated with IDU
• 3,700 PLWHA, 89.1 male

13
1994 study

• In 1994, Dr. Zunyou Wu and colleagues conducted a
  cross-sectional study of 1,548 males 18-29 year
  olds in 82 high risk villages.
• Found that initiation of drug use associated
  with
• having premarital or extramarital sex
• having a family member who used drugs
• being unmarried
• Found the incidence of drug use increased from
  10 in 1991 to 30 in 1994.

14
1997 intervention

• In response, local leaders worked together with
  Dr. Wu and Dr. Roger Detels to implement a
  community-based drug prevention intervention.
• 19 high risk villages in one cluster were
  randomly allocated to the intervention, 19
  similar villages were selected to serve as a
  control group.

15
1997 intervention

• Games
• Videos
• Agricultural classes
• School curriculum
• Skits
• Activity center
• Youth work core

16
1998 Follow-up

• In October 1998, all males 15-49 in the 38
  villages were invited to participate in a
  follow-up questionnaire.
• 559 males in control villages (91 participation)
• 748 males in intervention villages (88
  participation)

17
1998 Follow-up

• Intervention villages had a 2.7-fold greater
  decrease in incidence of new drug users compared
  to control villages
• Greatest reductions were observed in those most
  at risk
• Youngest age groups
• Single individuals
• Jingpo minority people
• Illiterate and semiliterate

18
PHASE I - Qualitative studyMay 2007
19
Phase I

• 30 in-depth interviews
• 9 in control villages
• 17 in intervention villages
• 4 with county level government
• Key informants
• Government officials
• School teachers
• Youth leaders
• Womens federation leaders
• Health workers

20
Phase I - Analysis

• Atlas.ti software used for analysis
• Data transcribed in Chinese
• Grounded theory
• Made initial notes
• Open, axial, and selective coding
• Conditional matrix of data built
• Visual model created
• Key quotes translated into English

21
(No Transcript)
22
Knowledge

• At that time, there were a lot of drug users and
  a lot of people were dying--there were too many
  people dying of AIDS. If there were no
  prevention activities, we wouldnt know the
  danger of these things to families and
  individuals.
• -45 year old woman, intervention village

23
Knowledge

• At the time, we were all scared of AIDS. We
  didnt really understand how HIV was transmitted.
  If you just said HIV people would get scared.
  But after the training, we gradually understood
  that HIV infection is preventable.
• --52 year old woman, intervention village

24
Motivation

• Through the HIV/AIDS education, everyones
  knowledge about prevention increased. We also
  treated those who were infected with HIV with
  more loving care. The knowledge touched us
  deeply--we also realized how precious life is.
  We all became motivated to do something.
• --43 year old man, intervention village

25
Motivation

• in 1997 we decided to form the Village
  Protection Committee to carry out prevention
  activities--we thought of this solution
  ourselves. At the time we thought, if our group
  of leaders could take the initiative on
  controlling the drug situation, it could make the
  drug users quit. While we were developing this,
  every household gave five yuan.
• --39 year old male, intervention village

26
Motivation (lack of)

• Relatively, our village has a lot of drug users,
  but higher level leaders very rarely come to do
  prevention activities or any kind of work. If
  the government doesnt support or pay attention
  to us, our drug prevention wont work well. They
  just tell our village leaders that they need to
  develop activities, but I think that without
  higher level government support, we cant do
  these things well.
• --48 year old male, control village

27
(No Transcript)
28
Educational programs

• Yes, every one of our schools has a drug and
  HIV/AIDS prevention curriculum. These sort of
  activities have become a regular thing.we just
  took it and made it part of our daily work.
• --38 year old male teacher, intervention village

29
Activity room

• You see, these countryside youth, after they got
  organized, at night after work and after dinner,
  they would go to the cultural activity room to
  rehearse programs, play poker, play chess. They
  didnt have an opportunity to come into to
  contact with drug users.
• --60 year old male, Longchuan County official

30
Village protection teams

• The government also doesnt know who is doing
  drugs where. If the drug users hear of any sort
  of disturbance, they will immediately run and
  hide, so its very difficult for the police. The
  village protection team lives here in the
  village, so they understand the situation
  clearly. They know exactly what time the drug
  users are coming and where they are hiding.
• --42 year old male, intervention village

31
(No Transcript)
32
Fear of the law

• Now there are 7-8 old drug users--there are no
  new drug users. In the past two years, the
  government policy has been stricter. So a few of
  them controlled their own problem and quit
  themselves. Since the policy has ben so strict
  in the past few years, they all got scared, they
  didnt want to go to labor camp, so they quit
  themselves at home.
• --46 year old male, control village

33
Government role

• There were effects, then once in a while the
  government wasnt as strict, and it wasnt as
  effective. The past two years have been good,
  government work groups have come here to help us
  and everything has gotten better. Before our
  farming equipment would all be stolen, in the
  past two years there has been no theft. The
  public safety situation has been better too.
• --46 year old male, control village

34
(No Transcript)
35
Fewer drug users

• So now things are relatively strict. Drug users
  in Longchuan have gone down a lot. One reason is
  because the government is focusing on it, another
  reason is because we have more awareness
  programs. Before it used to be that before the
  old drug users died, there were new drug users,
  it was very difficult to do work. Now with these
  few years of awareness and education programs,
  there are fewer and fewer drug users.
• --60 year old man, Longchuan County official

36
Economic improvement

• sugar cane production has gone up. Before
  since there were a lot of drug users, there was
  less sugar cane planting done on the farmland,
  and there was little output. After these
  programs, sugar cane planting went up.
• --43 year old man, intervention village

37
Increased public security

• There are less drug users now. Before there
  were a lot of drug users, they would do nothing
  at home, all they could do was rely on the help
  of their family. Before there was also a lot of
  theft, if you put anything down at home, it would
  be stolen in a second, you couldnt even relax
  when you were doing things at home. Those drug
  users dont do drugs anymore, theyve also built
  homes and their lives are much better. Their
  quality of life has improved.
• --40 year old woman, intervention village

38
(No Transcript)
39
Phase I - Discussion

• In general, stronger emphasis on enforcement
• Chinese policy toward drug users has changed
  (starting June 2008)
• Social desirability bias
• No interviews with drug users

40
PHASE II - Quantitative studyJune 2008
41
Phase II - Participants

• 774 males 15-39 years old from 18 intervention
  villages and 16 control villages
• 107 females 15-39 years old from two randomly
  selected control villages and two randomly
  selected intervention villages

42
one participant was excluded for being too
intoxicated after partially completing the survey
43
Phase II Data collection

• Village headmen notified all potential
  participants.
• Interviewer administered questionnaire
• door-to-door OR
• at a pre-determined location and time
• Anonymous questionnaire
• Demographics
• HIV knowledge/attitudes
• Tobacco and alcohol use
• Sexual behavior
• Drug behaviors/attitudes

44
Phase II - Analysis

• Descriptive
• ?2 test
• t-test
• Univariate
• PROC GLIMMIX
• Multivariate
• PROC GLIMMIX
• PROC MIXED

45
Phase II Analysis

• Primary outcomes
• Drug initiation after May 1997
• HIV knowledge
• Secondary outcomes
• HIV/AIDS in Longchuan
• Attitudes toward drugs and HIV/AIDS
• Drug prevention programs

46
Drug use
Intervention (n412) Intervention (n412) Control (n362) Control (n362) ?2 p-value
n n
Frequency of drug use
Never used drugs 286 69.8 303 83.7 lt0.0001
Used drugs regularly 52 12.7 47 13.0
Occasionally used drugs 72 17.6 12 3.3
Drug initiation
Never used drugs 286 70.4 303 83.9 lt0.0001
Started drugs before May 1997 53 13.1 21 5.8
Started drugs after May 1997 67 16.5 37 10.3
47
Demographics
Intervention (n412) Control (n362) ?2 p-value
Age
15-19 16.0 21.6 0.1346
20-29 42.0 38.1
30-39 42.0 40.3
Marital Status
Married 58.7 63.2 0.3469
Unmarried 39.8 34.9
Divorced/Other 1.5 1.9
Ethnic group
Han 16.8 24.3 lt0.0001
JingPo 56.7 33.4
Dai 25.8 37.3
Other 0.7 5.0
Education status
Illiterate 5.1 15.2 lt0.0001
Elementary school 38.0 42.8
Middle school 52.3 39.8
gtMiddle school 4.6 2.2
48
Drug initiation and demographics
OR 95 CI p-value
Age
15-19 0.03 (0.01, 0.12) lt0.0001
20-29 0.59 (0.41, 0.85) 0.0046
30-39 1.00 - -
Marital Status
Married 1.00 - -
Unmarried 0.31 (0.20, 0.48) lt0.0001
Divorced/Other 8.32 (2.71, 25.5) 0.0002
Ethnic group
Han 1.00 -
JingPo 3.57 (1.93, 6.61) lt0.0001
Dai 1.50 (0.77, 2.94) 0.2308
Other 3.80 (1.38, 10.48) 0.0099
Education status
Illiterate 3.47 (1.97, 6.11) lt0.0001
Elementary school 1.53 (1.03, 2.26) 0.0345
Middle school 1.00 - -
gtMiddle school 1.22 (0.44, 3.41) 0.6987
49
Risk Behaviors
Intervention (n412) Intervention (n412) Control (n362) Control (n362) ?2 p-value
n n
Has smoked cigarettes in past 30 days
No (lt1 cigarette/day) 105 25.5 114 31.5 0.0642
Yes 307 74.5 248 68.5
Had alcohol in past 30 days
No 92 22.3 101 27.9 0.0480
Yes 320 77.7 261 72.1
Had sex with non-spouse in past year
No 326 79.7 307 86.7 0.0101
Yes 83 20.3 47 13.3
50
Drug initiation and Risk Behaviors
OR 95 CI p-value
Has smoked cigarettes in past 30 days
No (lt1 cigarette/day) 1.00 - -
Yes 8.92 (4.49, 17.70) lt0.001
Had alcohol in past 30 days
No 1.00 - -
Yes 3.44 (2.04, 5.78) lt0.0001
Had sex with non-spouse in past year
No 1.00 - -
Yes 2.12 (1.40, 3.22) 0.0004
51
Multivariate model for drug initiation after May
1997

• Controlling for intervention group, age, marital
  status, ethnicity, education, smoking, alcohol
  use, and pre/extra-marital sex.

OR 95 CI p-value
Intervention group 1.78 (0.96, 3.28) 0.0672
Age 15-19 0.10 (0.02, 0.45) 0.0027
Unmarried 0.35 (0.19, 0.66) 0.0011
JingPo ethnicity 3.04 (1.52, 6.09) 0.0017
Illiterate 1.89 (0.98, 3.63) 0.0578
Smoking 6.90 (2.98, 15.95) lt0.0001
Drinking 1.89 (1.02, 3.51) 0.0431
Extra/pre-marital sex 3.81 (2.11, 6.87) lt0.0001
52
HIV knowledge
Intervention (n412) Control (n362) p-value
Answered all questions correctly on HIV knowledge 43.5 14.4 lt0.0001
Average HIV knowledge score (out of 16) 13.6 10.3 lt0.0001
Chi-squared test t-test
53
Multivariate model for HIV knowledge
Correctly answering all questions on HIV knowledge Correctly answering all questions on HIV knowledge Correctly answering all questions on HIV knowledge HIV score HIV score
OR 95 CI p-value ß p-value
Intervention group 4.66 (2.63, 8.24) lt.0001 2.19 lt.0001
Age 15-19 0.62 (0.36, 1.07) 0.0883 -1.57 0.007
Illiterate 0.18 (0.09, 0.37) lt0.0001 -5.35 lt.0001
Elementary school education 0.39 (0.28, 0.53) lt0.0001 -1.95 lt.0001
Controlling for intervention group, age, marital
status, ethnicity and education.
54
HIV/AIDS data
1989-2007 Intervention (n4404) Control (n4731)
Cumulative cases of HIV/AIDS 352 192
AIDS deaths 179 94
Prevalence of HIV (end of 2007) 3.9 2.1
No significant difference (t-test p-value0.1302)
55
Attitudes
Intervention (n412) Control (n362) ?2 p-value
agree agree
Drug use is a severe problem in my village 46.6 47.8 0.5715
Drug use affects me or my family 82.7 72.9 0.0034
HIV/AIDS is a severe problem in my village 37.8 40.1 0.7087
HIV/AIDS affects me or my family 63.8 51.4 0.0006
Government should be responsible for preventing drug use 94.2 97.5 0.0218
We should work together to help drug users 95.6 96.4 0.5819
Responses were recorded as Agree, Disagree
and Dont know Responses were recorded as
Agree and Disagree
56
Prevention programs
Intervention (n412) Control (n362) ?2 p-value
Is aware that village has Village Protection Team 87.4 83.4 0.2716
Is aware that village has cultural activity room 60.4 62.2 0.0783
57
Phase II Discussion

• Information on drug users scarce
• Response rates relatively low
• Disparity in response rates
• Results suggest that intervention affected
  knowledge without affecting behavior

58
CONCLUSIONS
59
Conclusions

• Improvements in knowledge
• Overall decrease in drug use
• Some programs remain
• Current focus is on law enforcement

60
Conclusions

• Rapid change in China makes it difficult to
  determine long-term effects
• Need for long-term follow-up on prevention
  programs
• Need better strategies to investigate drug use in
  community

61
THANK YOU!
62
(No Transcript)