Nineteen years retrospective analysis of epidemiology, antifungal resistance and a nomogram model for 30-day mortality in nosocomial candidemia patients
Dai Zhang, Jingjing Zhang, Minjing Cai, Yunhui Liao, Jingwen Zhang, Naifang Ye, Xinxin Lu, Jiajia Wang, Yun Xiao, Yan Zhang, Yihui Yao, Xianming Liang
Abstract
Background The incidence of nosocomial candidemia has increased in recently years, however, the epidemiological data remain insufficient in China. Methods A total of 234 candidemia patients were included from Xiamen University Zhong Shan hospital between January 2006 and October 2024. Incidence, species proportion, distribution, antifungal drug resistance of candidemia was analyzed. A nomogram model for 30-day morbidity of candidemia was determined using the least absolute shrinkage and logistic regression analysis. Results The incidence of candidemia increased in recent years (2020: 0.025%, 2021: 0.029%, 2023:0.022%). The dominant species of candidemia were Candida albicans (n=99,42.31%), Candida parapsilosis (n=47,20.09%), Candida tropicalis (n=43,18.38%), Candida glabrata (n=31,13.25%). Departments with a higher detection of candidemia included intensive care unit (n=55), emergency department (n=24) and hepatobiliary surgery (n=22). Candida tropicalis performed the highest resistance to azole (fluconazole: 55.81%, voriconazole:55.00% and itraconazole:58.14%). The resistance of Candida albicans to fluconazole, voriconazole and itraconazole were 32.32%, 23.53% and 31.31%. The mortality rate of 30-day discharge for candidemia reached 52.99%. 205 cases of candidemia patients from January 2006 to December 2023 were included as the training set, while 29 cases of candidiasis patients from January to October 2004 were included as the validation set. Five independent factors included Candida albicans , decreased albumin, multiple organ dysfunction syndrome, solid tumor and septic shock were adopted in a nomogram for 30-days mortality of candidemia. In the training set, the area under curve was 0.866 (95%CI: 0.817-0.916), the optimal cutoff value was 0.617, the sensitivity was 80% and the specificity was 80.4%. In the validation set, the area under curve was 0.808 (95%CI:0.737-0.970), the optimal cutoff value was 0.543. The sensitivity was 72.7% and the specificity was 83.3%. Conclusion The incidence of nosocomial candidemia has risen in recent years. Candida albicans remains the primary species, with the highest incidence is intensive care unit. Candida tropicalis exhibits the highest resistance rate to azole drugs. A nomogram predicting 30-day mortality discharge for candidemia patients has been constructed, and the independent risk factors including Candida albicans , multiple organ dysfunction syndrome, septic shock, solid tumors, and decreased albumin.