cai-junyang

Last update: August 21, 2025

PhD Student Faculty of Data and Decision Sciences at the Technion Google Scholar: Click here CV: Click here Web: https://cai-junyang.github.io/ Email: junyang.cai@campus.technion.ac.il

• Brief Introduction
• News
• Working Papers
• Published Papers

Brief Introduction

My name is Junyang Cai, and I am currently a doctoral student in the Faculty of Data and Decision Sciences at the Technion. My advisor is Prof Noa Zychlinski. I obtained the Bachelor of Science degree from Harbin University of Science and Technology in 2021 and the Master of Management degree from Shanghai University in 2024. My research direction is operations management in healthcare.

News

• I passed my doctoral qualifying examination on August 20, 2025.

• My paper Expanding Naloxone Accessibility: A Lifesaver or a Risky Setback? won the Excellent Paper Award of the 2025 Doctoral Student Forum of Antai College of Economics and Management, Shanghai Jiao Tong University on July 10, 2025.

• My paper A decision-making framework for supporting an equitable global vaccine distribution under humanitarian perspectives won the Excellent Paper Award of the 2024 Doctoral Student Forum of Antai College of Economics and Management, Shanghai Jiao Tong University on July 11, 2024. The work of this paper corresponds to my master’s thesis Greater disaster triggered by inequity: Equitable vaccine distribution under humanitarian perspectives.

Working Papers

• Expanding Naloxone Accessibility: A Lifesaver or a Risky Setback? (Production and Operations Management, Major Revision)
  • Winner of the 2025 Doctoral Student Forum of Antai College of Economics and Management, Shanghai Jiao Tong University, July 10, 2025
  • Oral presentation, 2025 ORSIS Annual Meeting, Tel-Aviv, May 6, 2025
  • Poster exhibition, Faculty Research Day, Technion, May 14, 2025 | Poster
  • Abstract: The widespread prevalence of opioids has prompted governments to implement targeted interventions aimed at reducing overdose mortality, with naloxone accessibility emerging as one of the most prominent policies. Naloxone, a potent opioid antagonist, is highly effective in reversing overdoses, yet its expanded availability introduces complex trade-offs, particularly in the presence of moral hazard. We develop a dynamic compartmental model that captures transitions between susceptible individuals and those with opioid use disorder (OUD), allowing us to evaluate the impact of naloxone accessibility on overdose mortality and to derive the optimal accessibility policy. We show that full naloxone accessibility is optimal in the absence of moral hazard or when its effect is small. However, when moral hazard is significant—where greater access to naloxone encourages riskier opioid use—expanded accessibility can paradoxically increase overdose deaths. Extending the model to incorporate peer-driven contagion in opioid misuse, we find that the structure of the optimal policy remains robust, preserving the bang-bang nature and the reversal induced by moral hazard. Two additional insights emerge under this interaction-based model. First, in epidemics primarily driven by prescription-induced opioid use, full accessibility remains optimal. In contrast, when opioid use spreads socially—especially as the effectiveness of naloxone declines due to potent synthetic opioids like carfentanil—limited accessibility may become preferable. Second, the relationship between naloxone accessibility and overdose mortality may become non-monotonic, exhibiting an inverted U-shape in which moderate increases in accessibility can initially worsen outcomes. A calibrated case study based on U.S. data suggests that under current epidemic conditions, full accessibility remains optimal—a finding that aligns with existing regulatory policies. However, our results highlight that shifts in epidemic dynamics, such as increased opioid potency or heightened social contagion, may fundamentally alter this conclusion. These findings underscore the need for continuous reevaluation of naloxone distribution policies as the opioid crisis evolves.

Published Papers

’*’ means the corresponding author.

• Cai J*, Zychlinski N. When AI is not enough: Reducing diagnostic errors with radiologist oversight. Service Science, forthcoming.

• Zhou J, Cai J*, Pantelous A, Li Z, Li M. A decision-making framework for supporting an equitable global vaccine distribution under humanitarian perspectives. European Journal of Operational Research, 2025, 327(2):655-672.
  • Winner of the 2024 Doctoral Student Forum of Antai College of Economics and Management, Shanghai Jiao Tong University, July 11, 2024

• Cai J, Xu R, Zhou J*, Gu H. Dynamic-spread assessment and kink detection in the early mpox outbreak. Applied Mathematical Modelling, 2025, 143:115996.

• Cai J, Zhou J*. How many asymptomatic cases were unconfirmed in the US COVID-19 pandemic? The evidence from a serological survey. Chaos, Solitons and Fractals, 2022, 164:112630.

• Li H, Cai J*. Arithmetic operations and expected values of regular interval type-2 fuzzy variables. Symmetry-Basel, 2021, 13(1):2196.