Identification of Key Influencers for Secondary Distribution of HIV Self-Testing Kits Among Chinese Men Who Have Sex With Men - Development of an Ensemble Machine Learning Approach

Aug 4, 2026

HIV self-testing has been rapidly expanded, yet secondary distribution success depends on identifying key individuals who can share multiple kits within social networks. This study targets Chinese men who have sex with men and formulates three influencer roles: key distributors, key promoters, and key detectors. An ensemble machine learning framework combines four predictive models and is benchmarked against human identification using leadership self-evaluated scales. Simulation results assess accuracy and the ability to reach first-time and positive testing alters.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

Background

Objective

Methods

Results

Conclusions

Trial Registration

Introduction

JOURNAL OF MEDICAL INTERNET RESEARCH Jing et al Original Paper Identification of Key Influencers for Secondary Distribution of HIV Self-Testing Kits Among Chinese Men Who Have Sex With Men: Development of an Ensemble Machine Learning Approach | Fengshi Jing1,2,3,4*, PhD; Yang Ye4,5*, PhD; Yi Zhou6*, DPH; Yuxin Ni3,7, MPH; Xumeng Yan3,8, MA; Ying Lu3, MA; Jason Ong9,10, PhD; Joseph D Tucker3,9,11, PhD; Dan Wu3,12, PhD; Yuan Xiong3,13, MSW; Chen Xu3, MPH; Xi He 14, BEng; Shanzi Huang6, MD; Xiaofeng Li6, MD; Hongbo Jiang15, PhD; Cheng Wang16, PhD; Wencan Dai6, MD; Liqun Huang6, MD; Wenhua Mei6, MD; Weibin Cheng1,4, PhD; Qingpeng Zhang17*, PhD; Weiming Tang1,3,11*, MD, PhD | | --- | | 1Institute for Healthcare Artificial Intelligence Application, Guangdong Second Provincial General Hospital, Guangzhou, China 2Faculty of Data Science, City University of Macau, Macao Special Administrative Region, China
3University of North Carolina at Chapel Hill Project-China, Guangzhou, China
4School of Data Science, City University of Hong Kong, Hong Kong Special Administrative Region, China
5Center for Infectious Disease Modeling and Analysis, Yale School of Public Health, Yale University, New Haven, CT, United States 6Department of HIV Prevention, Zhuhai Center for Diseases Control and Prevention, Zhuhai, China
7School of Public Health, Boston University, Boston, MA, United States
8Fielding School of Public Health, University of California Los Angeles, Los Angeles, CA, United States 9London School of Hygiene and Tropical Medicine, London, United Kingdom
10Melbourne Sexual Health Centre, Melbourne, Australia
11Division of Infectious Diseases, School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States 12School of Public Health, Nanjing Medical University, Nanjing, China
13School of Social Work, Michigan State University, East Lansing, MI, United States 14Zhuhai Xutong Voluntary Services Center, Zhuhai, China
15Department of Epidemiology and Biostatistics, School of Public Health, Guangdong Pharmaceutical University, Guangzhou, China 16Dermatology Hospital of Southern Medical University, Guangzhou, China
17Institute ofData Science and Department of Pharmacology and Pharmacy, The University of Hong Kong, Hong Kong Special Administrative Region, China
*these authors contributed equally
Corresponding Author:
Weiming Tang, MD, PhD
Institute for Healthcare Artificial Intelligence Application Guangdong Second Provincial General Hospital
466 Xingangzhong Road Guangzhou, 510317 China
Phone: 86 15920567132
[Email: ](Email: weiming_tang@med.unc.edu)[weiming_tang@med.unc.edu](Email: weiming_tang@med.unc.edu)
Abstract | Background: HIV self-testing (HIVST) has been rapidly scaled up and additional strategies further expand testing uptake. Secondary distribution involves people (defined as “indexes”) applying for multiple kits and subsequently sharing them with people (defined as “alters”) in their social networks. However, identifying key influencers is difficult. Objective: This study aimed to develop an innovative ensemble machine learning approach to identify key influencers among Chinese men who have sex with men (MSM) for secondary distribution of HIVST kits. Methods: We defined three types of key influencers: (1) key distributors who can distribute more kits,(2) key promoters who can contribute to finding first-time testing alters, and (3) key detectors who can help to find positive alters. Four machine learning models (logistic regression, support vector machine, decision tree, and random forest) were trained to identify key influencers. [https://www.jmir.org/2023/1/e37719](https://www.jmir.org/2023/1/e37719) XSL• FO RenderX J Med Internet Res 2023 | vol. 25 | e37719 | p. 1 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Jing et al An ensemble learning algorithm was adopted to combine these 4 models. For compa

FAQs

What challenge does the study address in HIV self-testing secondary distribution?

Identifying key influencers who can effectively distribute multiple HIV self-testing kits within social networks is difficult.

How does the study define key influencers?

It defines three types: key distributors (distribute more kits), key promoters (find first-time testing alters), and key detectors (help find positive alters).

What models and evaluation metrics are used to validate performance?

Four machine learning models are trained and combined via ensemble learning, and performance is evaluated using accuracy, precision, recall, and F1-score, compared with a human leadership scale approach through simulation.

Aug 4, 2026

Upload by Lucas Martin