Modeling the effects of routine screening for accidental lab-acquired infections on the risk of potential pandemic pathogen escape from high-biosafety research facilities.

Introduction: Accidental lab-acquired infections (LAIs) risk releasing potential pandemic pathogens from high-biosafety (BSL-3 or BSL-4) research facilities. We constructed a stochastic network infectious disease model to simulate how following an initial LAI, the probability of an outbreak of a pathogen resembling wild-type SARS-COV-2 would be influenced by test-and-isolate interventions over a 100-day horizon. Methods: We varied test frequency (0-7 tests/ week), peak sensitivity (50%-100%), and isolation delay (0-3 days), assuming perfect test specificity to minimize model complexity. For each of 192 parameter combinations, we conducted 1,000 simulations and used logistic regression to quantify how each parameter influenced the likelihood of an outbreak of 50 or more infections. Results: Results indicated that even relatively infrequent routine testing significantly reduced the risk of outbreaks under diverse plausible scenarios, with greater reductions achieved at higher test frequencies. Once-weekly testing reduced outbreak risk by 52% under optimistic assumptions (80% sensitivity, 1-day delay) and by 29% under pessimistic assumptions (50% sensitivity, 2-day delay). Testing two and five times weekly yielded risk reductions of up to 62% and 71%, respectively, under optimistic assumptions, and 43% and 55%, respectively, under pessimistic assumptions. Logistic regression showed each additional weekly test decreased outbreak odds by 20%, each 10-point increase in test sensitivity reduced odds by 10%, and each additional isolation delay day increased odds by 15.5%. Interaction analyses revealed longer isolation delays attenuated the protective effects of greater testing frequency and sensitivity. Discussion: These results show routine lab-worker screening with prompt isolation could substantially mitigate potential pandemic pathogen escape risks. Future research should address operational issues such as false positives, which could influence the feasibility, acceptability, and cost-effectiveness of real-world implementation.