Background: The most recent variant of concern, omicron (B.1.1.529), has caused numerous cases worldwide including the Republic of Korea due to its fast transmission and reduced vaccine effectiveness. Methods: A mathematical model considering age-structure, vaccine, antiviral drugs, and influx of the omicron variant was developed. We estimated transmission rates among age groups using maximum likelihood estimation for the age-structured model. The impact of non-pharmaceutical interventions (NPIs; in community and border), quantified by a parameter μ in the force of infection, and vaccination were examined through a multi-faceted analysis. A theory-based endemic equilibrium study was performed to find the manageable number of cases according to omicron- and healthcare-related factors. Results: By fitting the model to the available data, the estimated values of μ ranged from 0.31 to 0.73, representing the intensity of NPIs such as social distancing level. If μ < 0.55 and 300,000 booster shots were administered daily from February 3, 2022, the number of severe cases was forecasted to exceed the severe bed capacity. Moreover, the number of daily cases is reduced as the timing of screening measures is delayed. If screening measure was intensified as early as November 24, 2021 and the number of overseas entrant cases was contained to 1 case per 10 days, simulations showed that the daily incidence by February 3, 2022 could have been reduced by 87%. Furthermore, we found that the incidence number in mid-December 2021 exceeded the theory-driven manageable number of daily cases. Conclusion NPIs, vaccination, and antiviral drugs influence the spread of omicron and number of severe cases in the Republic of Korea. Intensive and early screening measures during the emergence of a new variant is key in controlling the epidemic size. Using the endemic equilibrium of the model, a formula for the manageable daily cases depending on the severity rate and average length of hospital stay was derived so that the number of severe cases does not surpass the severe bed capacity.

Objectives: Since protecting patients’ privacy is a major concern in clinical research, there has been a growing need for privacy-preserving data analysis platforms. For this purpose, a federated learning (FL) method based on the Observational Medical Outcomes Partnership (OMOP) common data model (CDM) was implemented, and its feasibility was demonstrated. Methods: We implemented an FL platform on FeederNet, which is a distributed clinical data analysis platform based on the OMOP CDM in Korea. We trained it through an artificial neural network (ANN) using data from patients who received steroid prescriptions or injections, with the aim of predicting the occurrence of side effects depending on the prescribed dose. The ANN was trained using the FL platform with the OMOP CDMs of Kyung Hee University Medical Center (KHMC) and Ajou University Hospital (AUH). Results: The area under the receiver operating characteristic curves (AUROCs) for predicting bone fracture, osteonecrosis, and osteoporosis using only data from each hospital were 0.8426, 0.6920, and 0.7727 for KHMC and 0.7891, 0.7049, and 0.7544 for AUH, respectively. In contrast, when using FL, the corresponding AUROCs were 0.8260, 0.7001, and 0.7928 for KHMC and 0.7912, 0.8076, and 0.7441 for AUH, respectively. In particular, FL led to a 14% improvement in performance for osteonecrosis at AUH. Conclusions FL can be performed with the OMOP CDM, and FL often shows better performance than using only a single institution's data. Therefore, research using OMOP CDM has been expanded from statistical analysis to machine learning so that researchers can conduct more diverse research.