关键词: 结核, 抗药性, 营养状况, 治疗结果, 药物副反应, 模型, 统计学

Abstract:

Objective: To investigate the impact of nutritional status on early efficacy and adverse drug reactions in patients with drug-resistant pulmonary tuberculosis (DR-PTB), and to construct a simplified risk scoring model to assist clinical assessment. Methods: A retrospective analysis was conducted on clinical data of patients with DR-PTB admitted to Anhui Chest Hospital from January 1, 2023 to January 1, 2025. Multivariate logistic regression was used to identify factors influencing sputum culture conversion at the end of the second month of treatment and adverse drug reactions (ADRs). Independent influencing factors were then selected to construct a simplified risk scoring model. Results: A total of 276 patients were enrolled, with a 2-month sputum conversion rate of 75.4% (208/276). Multivariate logistic regression analysis showed that body mass index (BMI)≥18.5 kg/m² (OR=3.043, 95%CI: 1.475-6.277), serum albumin ≥35 g/L (OR=3.610, 95%CI: 1.821-7.155), and hemoglobin ≥90 g/L (OR=4.464, 95%CI: 1.103-18.068) were independently positively associated with 2-month sputum conversion, indicating that patients meeting these criteria were more likely to achieve early conversion. In contrast, a Nutritional Risk Screening 2002 (NRS 2002) score ≥3 (OR=0.376, 95%CI: 0.175-0.808) was independently negatively associated with conversion, suggesting that patients with nutritional risk were less likely to achieve conversion. In terms of ADRs, for liver dysfunction, serum albumin ≥35 g/L was a protective factor (OR=0.286, 95%CI: 0.166-0.494), while an NRS 2002 score of ≥3 was a risk factor (OR=2.871, 95%CI: 1.470-5.608); For bone marrow suppression, an NRS 2002 score of ≥3 was a risk factor (OR=2.422, 95%CI: 1.312-4.472), a CD4⁺ T-cell count ≥410 cells/μl (OR=0.218, 95%CI: 0.081-0.584, P=0.002) and serum albumin ≥35 g/L (OR=0.550, 95%CI: 0.304-0.993) were protective factors. A simplified risk scoring model based on BMI, serum albumin, hemoglobin, and the NRS 2002 score showed that as the risk score increased, the rate of patients failing to achieve sputum culture conversion (${\chi }_{trend}^{2}$=48.733), the incidence of liver dysfunction (${\chi }_{trend}^{2}$=21.396), and the incidence of bone marrow suppression (${\chi }_{trend}^{2}$=11.272) all showed a gradual upward trend (all P-values <0.001). Conclusion: Nutritional status plays an important role in early efficacy and adverse drug reactions in patients with DR-PTB. The simplified risk scoring model based on BMI, serum albumin, hemoglobin, and NRS 2002 score has good clinical applicability and can be used for early risk identification, guiding individualized nutritional interventions and ADRs monitoring.

Key words: Tuberculosis, Drug resistance, Nutritional status, Treatment outcome, Adverse drug reaction, Models, statistical