Acute-on-chronic liver failure (ACLF) is a severe condition characterised by high short-term mortality, requiring rapid diagnosis, prompt treatment of trigger factors and appropriate prognostic evaluation to guide patients towards the best therapeutic options. Although several scores, such as the model for end-stage liver disease (MELD) and the CLIF-C ACLF score, have been proposed over time to assess short-term prognosis, their predictive accuracy remains suboptimal, partly due to heterogeneous diagnostic criteria and the inability to account for underlying comorbidities that significantly affect patient outcomes [1, 2].

Chen et al. [3] aimed to improve prognostic accuracy in patients with hepatitis B (HBV)-related ACLF. They analysed a retrospective cohort of 906 HBV patients with ACLF according to the Asian Pacific criteria and proposed a new score, the age-adjusted Charlson Comorbidity Index (aCCI)-HBV-ACLF score. This score incorporates the aCCI [4] along with key clinical indicators (neutrophil count, INR, serum bilirubin). The novel score accurately predicted short- and mid-term mortality, with areas under the ROC curve of 0.859, 0.869 and 0.868 for 28-day mortality, and 0.822, 0.850 and 0.888 for 90-day mortality in the training, internal validation and external validation cohorts, respectively, outperforming all available scores. Major strengths of this predictive model include the objective assessment of clinical and laboratory values, which reduces inter-clinician variability, and the inclusion of comorbidities. Moreover, although the score demonstrated only a slight increase in prognostic accuracy compared to the CLIF-C ACLF score in predicting 28-day outcomes, the clinical gain was more significant at 90 days, underscoring the potential impact of comorbidities on medium-term outcomes.

While the findings are promising, the score has several limitations. First, it has been tested only in Asian populations, raising questions about its generalizability across other ethnicities. Moreover, ACLF was diagnosed according to the Asian Pacific criteria [5], where liver failure is the primary factor. This explains why the new score identified two of the four factors as liver-specific (bilirubin and INR), whereas other factors that carry significant weight in other prognostic scores were not included.

Regarding the aCCI score, it appears that most points were derived from underlying liver disease across all analysed cohorts. Except for diabetes, none of the other factors included in the aCCI had a prevalence greater than 3%. This may partially undermine the aCCI's effectiveness as a tool for assessing extrahepatic comorbidities in HBV-ACLF patients. Furthermore, it should be noted that the aCCI was originally designed to estimate long-term mortality but has been considered a short-term prognostic tool in the manuscript by Chen et al. Some variables (e.g. AIDS, lymphoma) may now have different prognostic implications than when the score was first developed and validated.

Finally, the score was applied exclusively to patients who had not received liver transplantation, so it should be reserved for this specific cohort. This excludes a significant percentage of patients, given the survival benefit liver transplantation offers to patients with ACLF [6, 7].

In conclusion, the study offers a new score able to accurately predict prognosis in patients with HBVACLF. However, the actual effectiveness of the aCCI in correctly identifying the impact of comorbidities on short-term prognosis in ACLF patients warrants further investigation.

慢加急性肝衰竭 （ACLF） 是一种以短期死亡率高为特征的严重疾病，需要快速诊断、及时治疗触发因素和适当的预后评估，以指导患者寻找最佳治疗方案。尽管随着时间的推移，已经提出了几种评分，例如终末期肝病模型 （MELD） 和 CLIF-C ACLF 评分，用于评估短期预后，但其预测准确性仍然不理想，部分原因是诊断标准异质性，并且无法解释显著影响患者结局的潜在共存疾病 [1， 2]。

Chen 等 [3] 旨在提高乙型肝炎 （HBV） 相关 ACLF 患者的预后准确性。他们根据亚太地区标准分析了 906 名 HBV ACLF 患者的回顾性队列，并提出了一个新的评分，即年龄调整后的查尔森合并症指数 （aCCI）-HBV-ACLF 评分。该评分包括 aCCI [4] 以及关键临床指标（中性粒细胞计数、INR、血清胆红素）。新评分准确预测了短期和中期死亡率，在训练、内部验证和外部验证队列中，28 天死亡率的 ROC 曲线下面积为 0.859、0.869 和 0.868,90 天死亡率的 ROC 曲线下面积分别为 0.822、0.850 和 0.888，优于所有可用分数。该预测模型的主要优势包括对临床和实验室值的客观评估，这减少了临床医生间的变异性，以及纳入合并症。此外，尽管该评分显示，与CLIF-C ACLF评分相比，预测28天结局的预后准确性仅略有提高，但在90天时临床增益更为显著，强调了合并症对中期结局的潜在影响。

虽然研究结果很有希望，但该分数有几个局限性。首先，它仅在亚洲人群中进行了测试，这引发了对其在其他种族中的普遍性质疑。此外，ACLF 是根据亚太标准 [5] 诊断的，其中肝衰竭是主要因素。这就解释了为什么新评分将四个因素中的两个确定为肝脏特异性（胆红素和 INR），而其他在其他预后评分中具有重要权重的因素未包括在内。

关于 aCCI 评分，似乎大多数分数来自所有分析队列中的潜在肝病。除糖尿病外，aCCI 中包含的其他因素的患病率均未超过 3%。这可能会部分破坏 aCCI 作为评估 HBV-ACLF 患者肝外合并症的工具的有效性。此外，应该注意的是，aCCI 最初旨在估计长期死亡率，但 Chen 等人在手稿中被认为是一种短期预后工具。一些变量（例如 AIDS、淋巴瘤）现在可能具有与首次开发和验证评分时不同的预后影响。

最后，该评分仅适用于未接受肝移植的患者，因此应保留给该特定队列。考虑到肝移植为 ACLF 患者提供的生存益处，这排除了很大一部分患者 [6， 7]。

总之，该研究提供了一个能够准确预测 HBVACLF 患者预后的新评分。然而，aCCI 在正确识别合并症对 ACLF 患者短期预后影响方面的实际有效性值得进一步研究。