Reexamining Assumptions in Compartmental Models of Heterosexual HIV Transmission applied to Eswatini
Bibliographic record
Abstract
Compartmental HIV transmission models help guide global, national, and local epidemic responses via detailed projections, hypothetical scenarios, and mechanistic inference. While such models must make simplifying assumptions due to time and data constraints, prior work has shown that these assumptions can sometimes bias model outputs. In this thesis, I re-examine common assumptions used in compartmental HIV transmission models, and explore their potential influence on key model outputs. I focus on disproportionate risk among female sex workers, within the high-prevalence HIV epidemic of Eswatini. First, I systematically review how differential risk has been captured in prior models exploring HIV treatment scale-up across Sub-Saharan Africa. Among 94 studies, I find that only 2/5 explicitly included sex work, and only 1/4 considered differences in treatment access across risk groups. Next, I design, parameterize, and calibrate a model of heterosexual HIV transmission in Eswatini. The model features 4 partnership types and 8 risk groups, including higher/lower risk female sex workers and their clients. While parameterizing the model, I develop several new adjustments for common sources of bias in sexual behaviour data. I also critically review existing "force of infection equation" approaches, used to model HIV transmission via sexual partnerships within compartmental models, with respect to their implicit assumptions, data needs, strengths, and limitations. Drawing on this review, I develop a new approach - the Effective Partnerships Adjustment - which can overcome some of the key limitations of existing approaches. Comparing the new and existing approaches in the Eswatini model, I show that some existing approaches might be systematically overestimating the impact of prevention in longer partnerships, and underestimating the impact of prevention in shorter partnerships. Finally, I apply the Eswatini model to further explore scenarios with different HIV treatment scale-up across risk groups. I find that the prevention impacts of treatment could be substantially reduced if higher risk groups are "left behind". Taken together, these results suggest that existing compartmental HIV transmission models may underestimate the importance of prioritizing resources to populations at highest risk of HIV acquisition and/or transmission, including female sex workers, even within a high prevalence epidemic such as Eswatini.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.020 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.003 | 0.005 |
| Open science | 0.004 | 0.003 |
| Research integrity | 0.002 | 0.006 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".