Wu, Yao-KuangYao-KuangWuChung, Hsueh-WenHsueh-WenChungChen, Wei-ChihWei-ChihChenYang, Kuang-YaoKuang-YaoYangJao, Lun-YuLun-YuJaoChang, Hou-TaiHou-TaiChangTseng, Chien-HuaChien-HuaTsengChen, Tzu-TaoTzu-TaoChenSHIH-CHI KUTAO-MIN HUANGChiu, Tzu-HsuanTzu-HsuanChiuKao, Kuo-ChinKuo-ChinKaoWang, Chieh-JenChieh-JenWangWang, Chiao-HungChiao-HungWangYang, Tse-BinTse-BinYangSuk, Chi-WonChi-WonSukPeng, Chung-KanChung-KanPengShen, Chih-HaoChih-HaoShenChien, Yu-SanYu-SanChienKuo, Li-KuoLi-KuoKuoLin, Chen-ChunChen-ChunLinLin, Chia-MoChia-MoLinLiu, Wei-LunWei-LunLiuWu, Huang-PinHuang-PinWuChiu, Ming-HuangMing-HuangChiuChan, Ming-ChengMing-ChengChanLin, Chieh-MoChieh-MoLinLee, Chung-ShuChung-ShuLeeCHIA-HAO CHANGChen, Hsing-ChunHsing-ChunChenLI-TA KENGChang, Shih-ChiehShih-ChiehChangLai, Yi-ChunYi-ChunLaiCHEN YI-TINGChen, Chin-MingChin-MingChenFang, Wen-FengWen-FengFangSu, Wen-LinWen-LinSu2025-04-292025-04-292025-04-08https://scholars.lib.ntu.edu.tw/handle/123456789/728478Purpose: To evaluate the impact of fungal colonization and infection phenotypes and other prognostic factors on in-hospital mortality among mechanically ventilated COVID-19 patients (n = 376) admitted to ICUs during the first wave of the pandemic in Taiwan. Materials and methods: A target trial emulation framework was used to minimize immortal time bias. Patients were matched 1:1:2 for age and gender and classified into three groups: 94 in the “Early” group (fungal colonization or infection within 10 days), 94 in the “Late” group (10–30 days), and 188 in the “No” group (no fungal colonization or infection within 30 days). In-hospital mortality and clinical outcomes were compared across groups. Results: Patients in the “Early” group received higher cumulative corticosteroid doses, had lower PaO2/FiO2 ratios, and exhibited higher rates of comorbidities, cytomegalovirus viremia, and lung, heart, and kidney complications. They also had a longer duration of ventilator use, ICU stay, and total hospitalization compared to the “Late” and “No” groups. Time-dependent multivariate Cox regression analysis identified the “Early” phenotype as a strong predictor of in-hospital mortality (adjusted hazard ratio [aHR]= 3.992, 95 % CI: 2.676–5.956, p < 0.001). Additional independent risk factors included Charlson Comorbidity Index (aHR = 1.213, 95 % CI: 1.113–1.323, p < 0.001) and APACHE II score (aHR = 1.028, 95 % CI: 1.011–1.045, p = 0.001). In contrast, higher PaO2/FiO2 ratios (aHR = 0.998, 95 % CI: 0.997–1.000, p = 0.021) and ganciclovir use (aHR = 0.419, 95 % CI: 0.245–0.717, p = 0.002) were associated with reduced mortality. Conclusions: “Early” fungal colonization and infection within 10 days of corticosteroid initiation is an independent risk factor for in-hospital mortality in mechanically ventilated COVID-19 patients. Future research should explore early intervention strategies, including antifungal prophylaxis, optimized corticosteroid dosing, and immune modulation, to improve survival outcomes.enCOVID-19CorticosteroidFungal colonizationMortalityPhenotypePrognostic factorjournal article10.1016/j.jiph.2025.10276740220505