Abstract
OBJECTIVE: Most patients with Marchiafava-Bignami disease (MBD) had unfavorable prognosis, with disability or death. We aimed to determine the risk factors of early unfavorable prognosis of MBD, and to develop a predictive nomogram for early unfavorable prognosis of MBD. METHODS: MBD patients admitted to our hospital between 1 January 2013 and 31 December 2021 were included. Unfavorable prognosis was defined as mRS score ≥3, the independent risk factors for unfavorable prognosis of MBD with the odds ratio (OR) and 95% confidential interval (CI) acquired by multiple logistic regression were included in development of the predictive nomogram for early unfavorable prognosis of MBD, and the area under curve (AUC) of the receiver operating characteristic curve was calculated. The published case reports of MBD were used as the external validation group to verify the predictive ability of the nomogram. RESULTS: Independent risk factors for early unfavorable prognosis of MBD included Glasgow Coma Scale score (OR = 0.636, 95% CI = 0.506-0.800, p = 0.004) and pneumonia (OR = 2.317, 95% CI = 1.003-5.352, p = 0.049). The AUC of the nomogram was 0.852. Ninety-four case reports, a total of 100 cases of MBD were included as the external validation group, its AUC was 0.840. The online dynamic nomogram for early unfavorable prognosis of MBD was constructed. INTERPRETATION: It is confirmed by external validation that the nomogram has a preferable predictive ability and clinical efficacy, and the dynamic online predictive nomogram is helpful for physicians to quickly assess the prognosis of MBD.