基于潜变量增长模型的心衰患者报告临床结局动态轨迹研究

Dynamic Trajectory of Patient Reported Outcomes with Heart Failure Based on Latent Growth Model

目的分析慢性心力衰竭纵向随访数据,揭示患者动态变化轨迹规律,为高危患者在疾病不同阶段采取个性化干预以及提高重症患者生存质量提供理论依据。方法收集心衰患者4个时点生存质量随访数据,分别拟合线性潜变量增长模型、二次函数型潜变量增长模型、非定义类型的潜变量增长模型。通过比较模型拟合效果,揭示患者报告临床结局评分的动态变化轨迹。结果通过在患者报告临床结局评分的领域水平拟合3种潜在增长模型,结果均表明二次函数型潜在增长模型较其余两种模型为较优模型,其拟合指标更接近可接受限值。在领域水平,个体间的生存质量起始值不同,而个体内的变化趋势差异并不显著。进一步在生理领域的3个维度水平拟合潜在增长模型,在躯体症状、食欲睡眠、独立性维度拟合结果表明最优模型分别为非定义类型、线性、二次函数型潜在增长模型。总体来说患者在出院后的报告临床结局评分呈先升高后平稳的变化趋势。结论通过探究院后患者报告临床结局随时间的变化轨迹规律,了解个体生存质量变化趋势并进一步解释个体间存在的差异,对患者不同阶段生命历程给予个性化干预措施。

Objective The longitudinal follow-up data of chronic heart failure were analyzed to reveal the dynamic changes of patients,and to provide theoretical basis for individualized interventions in different stages of the disease for high-risk patients and for improving the quality of life of severe patients.Methods The follow-up data of quality of life of patients with heart failure at 4 time points were collected,and the linear latent variable growth model,quadratic latent variable growth model and undefined latent variable growth model were fitted respectively.By comparing the fitting effect of the model,the dynamic change trajectory of the Patient Reported Outcomes was revealed.Results Three potential growth models were fitted at the domain level of patient report clinical outcome score.The results showed that the quadratic potential growth model was better than the other two models,and its fitting index was closer to the acceptable limit.At the domain level,the initial value of QOL is different among individuals,but the variation trend within individuals is not significant.Furthermore,potential growth models were fitted horizontally in three dimensions of physiology.The fitting results of somatic symptoms,appetite sleep and independence dimensions showed that the optimal models were non-defined,linear and quadratic potential growth models.Overall,the clinical outcome score of patients after discharge showed a trend of elevation first and then stable change.Conclusion By investigating the track of Patient Reported Outcomes after hospitalization over time,we can understand the trend of individual quality of life and further explain the differences among individuals,and give individualized interventions to patients at different stages of life.