Predictive approaches in HIV to estimate a patient’s risk to present with relevant health outcomes, such as hospitalizations and AIDS-related death, long before they happen, could be highly useful. We aimed to develop a risk classification instrument for virological failure through a scoring system that identifies patients with a low, medium, and high risk after six months of ART treatment. A case–control design was implemented through 355 HIV-positive Colombian adults who were assessed using the designed instrument. The variables with independent predictive values were selected using logistic regression analysis, and the diagnostic performance of the prediction score was evaluated using the area under the curve. The prediction score included relevant psychosocial and biological risk factors, some of them modifiable variables like substance use and low health literacy. The area under the curve value for the total prediction score was 0.85 (CI 0.80–0.90). Therefore, this instrument could be a valuable tool to identify at-risk patients of virological failure. In low and middle-income countries, the associated risk factors of virological failure are little known. Assessing such risk would lead to make individualized decisions regarding the patient’s management and minimize the chance of non-desirable outcomes.

艾滋病毒的预测方法可以很有效地用于估计患者在相关健康结果（如住院和与艾滋病相关的死亡）发生之前的风险，这些方法很早就可以发挥作用。我们的目标是通过评分系统开发一种用于病毒学失败的风险分类工具，该系统可以识别出接受ART治疗六个月后的低，中和高风险患者。病例对照设计是通过使用设计工具进行评估的355名艾滋病毒呈阳性的哥伦比亚成年人实施的。使用逻辑回归分析选择具有独立预测值的变量，并使用曲线下的面积评估预测得分的诊断性能。预测分数包括相关的社会心理和生物学危险因素，其中一些可修改的变量，例如物质使用和低健康素养。总预测得分的曲线值下的面积为0.85（CI 0.80–0.90）。因此，该仪器可能是鉴定病毒学衰竭高危患者的有价值的工具。在中低收入国家，病毒学失败的相关危险因素鲜为人知。评估此类风险将导致就患者的治疗做出个性化的决策，并最大程度地降低不良后果的机会。