148. Modeling Antiretroviral Therapy as Prevention for HIV: the Impact of Population Mobility and Linkage to Care
Session: Oral Abstract Session: HIV Primary Care and Antiretroviral Therapy, Epidemiology and Testing
Friday, October 21, 2011: 9:45 AM
Room: 157ABC

Background: Recent models suggested that frequent testing for HIV with immediate initiation of antiretroviral therapy (ART) could profoundly curb transmission. The debate about ART as prevention has focused largely on parameter values; structural assumptions regarding linkage to care and population mobility (migration) have received less attention.

Methods: We compared 3 models of ART as prevention to assess the impact of a refined linkage structure as well as population mobility: Model 1) a published model of ART as prevention; Model 2) a revised Model 1 that incorporates a formulation of linkage to care in which individuals who decline testing or treatment re-engage with care in late stage disease; and Model 3) Model 2 with the inclusion of population mobility. We compared incidence and prevalence projections over 30-years. Input data for the models included demographic, clinical, immigration/emigration (10%/4% per annum) and linkage data from a South African township. We examined 2 linkage to care scenarios: high linkage rate (92%) and a community linkage estimate (53%).

Results: By 30 years, incidence was below 0.1% (‘elimination') under both linkage rates in model 1 but only under the higher linkage rate in model 2 (Table 1). 92% linkage in model 2 yielded comparable results to 53% linkage in model 1. Incorporating population mobility (model 3) resulted in higher incidence and prevalence at each time point; elimination did not occur under either linkage scenario.

Conclusion: With a refined linkage to care structure, projections of ART as prevention models were more pessimistic and very sensitive to the rate of linkage to care.  Benefits were also reduced in settings with substantial population mobility, as seen in many Sub-Saharan African communities. Incorporating these features with data from a South African township, we projected the impact of ART as prevention to be substantially less than prior estimates.


Table 1. Projections for the 3 models by linkage rate.

Incidence (%)

Prevalence (%)

10y

20y

30y

10y

20y

30y

Model 1

Linkage 92%

0.23

0.11

0.05

11.3

5.5

2.5

Linkage 53%

0.33

0.18

0.09

12.1

6.3

3.2

Model 2

Linkage 92%

0.33

0.16

0.08

12.1

6.2

3.1

Linkage 53%

0.83

0.52

0.35

14.7

9.7

6.5

Model 3

Linkage 92%

0.84

0.57

0.41

14.5

9.9

7.0

Linkage 53%

0.99

0.93

0.92

16.0

15.0

14.8


Subject Category: H. HIV/AIDS and other retroviruses

Jason Andrews, M.D.1, Robin Wood, MD, PhD2, Linda-Gail Bekker, MBChB, DTMH, DCH, FCP(SA)2 and Rochelle Walensky, MD, MPH, FIDSA1, (1)Division of Infectious Diseases, Massachusetts General Hospital, Boston, MA, (2)Desmond Tutu HIV Centre, Institute of Infectious Diseases and Molecular Medicine, University of Cape Town, Cape Town, South Africa

Disclosures:

J. Andrews, None

R. Wood, None

L. G. Bekker, None

R. Walensky, None

Findings in the abstracts are embargoed until 12:01 a.m. EST Thursday, Oct. 20 with the exception of research findings presented at IDSA press conferences.