HIV, Infectious Diseases, and Drug Use in Networks and Communities

1
HIV, Infectious Diseases, and Drug Use in
Networks and Communities

• I would like to acknowledge
• NIDA projects
• R01 DA13128 (Networks, norms HIV risk among
  youth)
• R01 DA13336 (Community Vulnerability and Response
  to IDU-Related HIV),
• P30 DA11041 (Center for Drug Use and HIV
  Research)
• R01 DA10870 (HIV Risk among Women IDUs Who Have
  Sex with Women)
• NIMH project
• R01 MH62280 (Local context, social-control
  action, and HIV risk)
• Hundreds of participants in these studies
• Many collaborators and co-authors

2
Background

3
Cumulative AIDS cases among US adults and
adolescents by Exposure Category, Dec 2002
4
(No Transcript)
5
Partial list of predictors of an IDUs being or
becoming infected with HIV

• Behavioral Sharing syringes, backloading,
  indirect sharing, MSM
• Sociodemographic Black, Puerto Rican, WSW, years
  of injection
• Network High risk injection or sex partners
  member of sociometric microstructure
• Note well Non-injecting heroin users and crack
  smokers are also at increased risk of HIV,
  hepatitis B and C, syphilis, and HSV-2. The most
  likely reason is unprotected sex with IDUs.

6
What characteristics of US metropolitan areas are
associated with the rate of IDUs per capita and
with HIV prevalence rate among IDUs?

• Data from the Community Vulnerability and
  Response to IDU-related HIV study

7
Methods

• Unit of analysis Large Metropolitan Statistical
  Areas
• The 96 MSAs (in USA) with populations of 500,000
  in 1993
• MSAs are defined by
• County boundaries
• Central city population of 50,000 or more
• Based on economic and social integration with
  surrounding areas and on commuting patterns to
  central city
• Ns vary depending on missing data 96 for IDUs
  per capita, 91 for HIV prevalence rate.

8
Dependent Variable 1 IDUs per capita, 1998

• IDUs per capita was estimated by
• Using average of four multipliers to allocate
  total number of IDUs in the USA to the 96 MSAs
• Dividing by the population of the MSA.
• See Friedman et al, J Urban Health, 2004

9
Dependent variable 2 HIV prevalence rate in
IDUs in 1998

• HIV prevalence rates in 91 MSAs were estimated
  by
• In 26 MSAs where research data exists, these data
  were used as the estimates.
• In 65 other MSAs, we used the average of two
  results
• Regressing research results on HIV counseling and
  testing data, and using this equation to predict
  the HIV prevalence rate in the other 65 MSAs
  and
• Using the Lieb techniques to estimate HIV
  prevalence rates as the number of IDUs living
  with HIV (itself estimated as a function of IDUs
  living with AIDS) divided by the number of IDUs
  (Lieb et al, J Urban Health, 2004).

10
(No Transcript)
11
Predictors of IDUs per capita
12
(No Transcript)
13
Final predictors of HIV prevalence rate
14
Limitations

• Causal mechanisms are hard to study at a single
  level of analysis
• Lack of time series data makes causal inference
  difficult
• We plan to conduct time series analyses in the
  near future

15
Summary of this part

• Socioeconomic factors (median income
  unemployment) of MSAs are associated with more
  IDUs/capita and with higher HIV prevalence among
  IDUs.
• Black/white racial residential segregation is
  associated with fewer IDUs per capita (which is
  puzzling).
• Health expenditures are associated with lower HIV
  prevalence.
• Higher rates of arrest for hard drug use are
  associated with more IDUs and more HIV among IDUs.

16
Insights from the New York City HIV epidemic
among IDUs

• This section presents some overviews plus it uses
  new data from the Networks, Norms and HIV Risk
  among Youth study in the Bushwick section of
  Brooklyn

17
I will now show evidence that IDUs, crack
smokers, and other community residents are active
participants in the fight against HIV

• Thus, theories that view IDUs as helpless victims
  of addiction or as uncaring spreaders of HIV and
  other infections seem to be misleading.
• Likewise, theories that see public health
  interventions and drug abuse treatment as the
  actors in HIV prevention are incomplete.

18

19
Intravention and its implications

• Currently, based on decades of experience with
  HIV/AIDS and with street drug use, and perhaps
  with prevention programs, in some neighborhoods
  such as Bushwick residents engage in activities
  to help others to protect themselves

20
Survey findings about other-protective action in
the prior 3 months by hardest drug used in last
3 months
21

• Qualitative interviews confirm that these reports
  refer to recent concrete actions rather than to
  abstract intentions or actions well in the past.
• Other-protective actions directed at drug users
  are more likely by those who are hard drug users
  and thus, perhaps, more likely to interact with
  users
• Drug users frequently act to urge others to
  protect themselves
• These data suggest that
• Urging safer behaviors has been institutionalized
  into the community as a somewhat self-sustaining
  intravention.
• Many IDUs take actions to protect others from
  infection.
• HIV prevention efforts and other programs need to
  take pre-existing intraventions into account.

22
Drug users organizations IDUs can work
collectively and formally against HIV

• Thai drug users network
• Rotterdam junkiebund
• Australian Intravenous League and state
  organizations
• Some US prevention projects are users groups too
• Etc.

23
Users in the community

• IDUs and other users have many social
  relationships with others

24

• We have already discussed drug users
  participation in intraventions
• Now we will look at sexual networks in the
  Bushwick (Brooklyn) community

25
Gender/Sexuality (MSMup triangle, WSWdown
triangle, other femalecircle, other malesquare)
by Hardest Drug Use Ever (from dark red to light
pink IDU, Crack, NI Heroin or Cocaine
blueother)
26

• This preliminary diagram shows that the two
  behavioral groups at highest HIV riskIDUs and
  MSMhave many risk network connections.
• It also shows that women who have sex with women
  may be at risk through injection and sexual
  networks with IDUs and with MSM (including MSM
  who are IDUs)which may help explain other
  studies findings that WSW IDUs are at very high
  HIV risk.
• Finally, it presents many instances of crack
  smokers and other non-injecting cocaine and
  heroin users who have sexual ties to IDUswhich
  may help explain why these groups are at enhanced
  HIV risk.

27
HIV-positive by Gender/Sexuality (MSMup
triangle, WSWdown triangle, other femalecircle,
other malesquare) by Hardest Drug Use Ever (from
dark red to light pink IDU, Crack, NI Heroin or
Cocaine blueother)
28
HCV-positive by Gender/Sexuality (MSMup
triangle, WSWdown triangle, other femalecircle,
other malesquare) by Hardest Drug Use Ever (from
dark red to light pink IDU, Crack, NI Heroin or
Cocaine blueother)
29
HSV2-positive by Gender/Sexuality (MSMup
triangle, WSWdown triangle, other femalecircle,
other malesquare) by Hardest Drug Use Ever (from
dark red to light pink IDU, Crack, NI Heroin or
Cocaine blueother)
30
Health Activism Star by Gender/Sexuality (MSMup
triangle, WSWdown triangle, other femalecircle,
other malesquare) by Hardest Drug Use Ever (from
dark red to light pink IDU, Crack, NI Heroin or
Cocaine blueother)
31
HBV (exposed, immunizedV) by Gender/Sexuality
(MSMup triangle, WSWdown triangle, other
femalecircle, other malesquare) by Hardest Drug
Use Ever (from dark red to light pink IDU,
Crack, NI Heroin or Cocaine blueother)
32
Classification Core, Sex Partners, and Distance

• Core Men who have sex with men (MSM) and
  injection drug users (IDUs) are a core group for
  HIV and HBV infection, and perhaps for other
  infections N 201
• Sex partners (SPs) are defined as sex partners of
  one or more core members N 67
• D2 (distance 2) are sex partners of sex
  partners N 32
• D3 are sex partners of D2 members, or sex
  partners of other D3 sex partners N 19
• Unlinked are non-core subjects who are not
  sexually linked to a core group member by a path
  of any length N 94

33
Blood-borne virus infection (and hep B induced
immunity) by network distance from core

• HIV HCV HBVª HBV
• immune
• Core (MSM /or IDU) 18 60 58 20
• SPs 10 4 24 25
• D2 0 3 22 38
• D3 0 0 8 45
• Unlinked to core 0 1 4 40
• ª HBV exposure among the unvaccinated.
• p chi-sq test for trend) lt .001.

34
Sexually-transmitted infections by network
distance from core

• n HSV2 HSV1 Syph. GC CT
• Core (MSM or IDU) 201 60 58 5 1 4
• SPs 67 56 87 3 2 2
• D2 32 28 77 0 3 9
• D3 19 37 79 0 0 26
• Unlinked to core 94 30 69 0 0 9
• p (chi-sq test for trend) lt .001.

35
Implications

• Treatment centers and other projects
• STI prevention and treatment for users
• Some ethnographic evidence from WSW Project that
  services for drug users assume heterosexuality
  and that this may hurt MSM and WSW.
• HCV education and treatment
• Vaccination for hep B inadequate. Research
  planning needed for HIV, HCV HSV-2 vaccines
• Macro issues matterboth economics and policy.
  They affect SEP, treatment, IDU/capita, HIV
  prevalence among IDUs
• IDUs are part of the community both
  epidemiologically and as part of intraventions
• IDUs, other users, and community residents can be
  ACTORS for public healthand thus allies.