临床预测模型的可解释性及应用进展

尽管机器学习已在临床预测研究中得到广泛使用，但黑盒机制严重限制了其在临床实践中的推广。近年来，越来越多的研究人员采用模型可解释方法来揭示预测模型的决策机制；同时，可解释的临床预测模型也在实践中得到更多应用。为此，本文对临床预测模型可解释性方法进行了梳理和总结，并讨论了可解释的临床预测模型在疾病三级预防体系中的应用。本文将有助于读者了解模型可解释方法和应用，为推动机器学习走向临床实践提供参考。     

危重病预后评分系统

文章回顾了危重病预后评分系统的建立和发展过程，简述了疾病预后评分系统的原理和用途，介绍了死亡概率预测模型（ＭＰＭ），简化的急性生理评分（ＳＡＰＳ）和急性生理、年龄、慢性病评分（ＡＰＡＣＨＥ）。认为危重病预后评分系统作为对疾病严重程度评价已广泛应用，它既有利于临床诊断、病情的确定和治疗护理措施的制定和改进，也有利于提高科研水平。     

COX回归联合决策树建立宫颈癌预后的预测模型

目的 运用COX回归与决策树探讨宫颈癌患者预后的影响因素并建立预测模型。方法 收集1 075例自2013—2019年入院治疗的宫颈癌患者的临床资料和随访数据。采用 检验、COX回归模型探讨预后的影响因素，并借助决策树建立预测模型。结果 多因素COX回归模型显示更晚的FIGO分期、非鳞癌或腺癌的病理类型、深层肌层浸润、赘生物直径≥ 4 cm均是宫颈癌预后的独立影响因素，宫旁阳性、阴道穹隆受累是肿瘤进展的独立影响因素，而脉管侵犯、尖锐湿疣是死亡的独立影响因素（P＜0.05）。决策树结果显示肌层浸润和赘生物直径的影响最显著。进展及死亡模型AUC分别为0.698，0.745，正确分类预测百分比为89.9%、93.6%。结论 肌层浸润、赘生物直径、宫旁阳性、阴道穹窿受累、病理类型、脉管侵犯、尖锐湿疣是宫颈癌患者预后的独立影响因素。COX回归联合决策树建立预测模型的方法可以联合两种模型优势，结果可视化，为临床评估提供参考。     

早产儿支气管肺发育不良相关预测模型研究

支气管肺发育不良(BPD)是一种好发于早产儿的、以肺泡简单化及肺血管发育迟缓为主要特征，同时伴有肺纤维化、囊性变及肺不张的慢性肺疾病，至今尚无有效治疗办法，一经发生不可逆转，严重影响患儿的生存率和近远期预后。因此，早发现、早预防、早治疗对评估及改善患儿预后极其重要。目前，国内外针对BPD发生的高危因素进行了大量研究，并建立了多种BPD发生的早期预测模型。本文将有关早产儿BPD预测模型的最新研究进展作一综述，旨在为临床早期干预、预防BPD的发生和发展提供理论依据。     

急性心力衰竭再入院及死亡风险预测模型研究进展

心力衰竭是各种心脏疾病的严重和终末期表现,具有高住院率、高病死率等特点,已成为重要的公共卫生问题。急性心力衰竭患者出院后再入院率及死亡率是评价心力衰竭医疗质量的重要指标,基于此开展急性心力衰竭患者疾病预后风险预测研究,对量化疾病风险、落实分层管理、优化临床决策、提高生存质量、改善患者预后、全面提升我国急性心力衰竭医疗质量至关重要。近20年来,国外学者已开发出数十个急性心力衰竭再入院及死亡风险预测模型,我国学者也开发出了近十个基于中国人群的预测模型,但目前国内指南中尚无推荐使用的急性心力衰竭预后风险预测模型。本文旨在通过介绍国内外主要急性心力衰竭再入院和死亡风险预测模型,重点概述现有模型局限性及今后发展方向,包括整合多源数据、挖掘新兴生物标志物、构建多基因风险评分、优化机器学习方法、推进模型适用性调整及拓宽应用渠道等,以期为国内急性心力衰竭再入院和死亡风险预测模型相关研究提供思路。     

道路交通事故致颅脑损伤预后影响因素及预测模型建立的探讨

本研究旨在通过对交通事故致颅脑损伤的回顾性调查分析,筛查对临床预后有重要影响的因素,同时对建立预后预测模型进行探讨。对象与方法1.病例选择:1999年1月至2004年12月间武汉某两家大型综合医院收治的因交通伤害第一次住院的病人中,出院主要诊断为颅脑损伤的病人共计906例。颅脑损伤分级采用1997年中华医学会制定的分类标准。2.研究方法:数据收集采用回顾性调查方法,资料来源于住院病历档案。多变量统计模型主要用lo-gistic回归模型,模拟效果的检查采用了似然比检验(likelihood ratio test)、Hosmer-Lemeshow拟合优度检验(Hosmer and Le…     

基于logistic回归的二分类结局临床预测模型的Stata实现方法

目的 结合早期胃癌患者淋巴结转移临床预测模型的实例,探讨通过Stata软件建立、评价和验证二分类结局的临床预测模型的实现方法。方法 选取2010年至2018年收集的早期胃癌患者淋巴结转移数据为实例数据集,并以2017年7月1日作为患者入组分界点,将数据分为建模集和验证集,通过实例介绍利用Stata/SE 15.0建立、评价和验证二分类结局临床预测模型的方法。结果 实例中建模集746例胃癌患者中144例(19.3%)发生淋巴结转移。预测模型最终纳入T分期、肿瘤最大径、分化程度和脉管浸润4个变量。模型的区分度评价指标C指数为0.864,模型校准度Hosmer-Lemeshow检验P=0.983,临床决策曲线显示临床适用度较好。在验证集中,模型的C指数为0.911,校准度Hosmer-Lemeshow检验的P值为0.631。结论 利用Stata软件可以简单、快捷地实现临床预测模型的建立、评价和验证过程,尤其在列线图的绘制方面存在优势。     

医学生心肺复苏考试成绩预测模型的建立

目的 利用临床医学系学生的实习数据建立心肺复苏考试成绩预测模型.方法 回顾性收集2016年1月至2018年12月在广州医科大学第二临床学院实习的临床医学系实习生数据.学生数据包括学生实习期间各科考试成绩、性别及客观结构化临床考试的心肺复苏考站成绩,通过逻辑回归分析,建立心肺复苏考试成绩预测模型.结果 对382例医学生实...     

基于大数据平台的糖尿病视网膜病变预测模型的建立及验证

目的建立简易的预测模型并验证其有效性,以预测糖尿病视网膜病变的发生风险,从而减少不良结局的发生。方法在医疗大数据平台筛选导出2010年1月1日—2016年12月31日出院诊断为2型糖尿病的病例,根据是否患视网膜病变分为DR组和NDR组。采用SPSS 22.0软件进行组间比较、影响因素分析并建立预测模型,采用Medcalc软件绘制受试者工作曲线。结果 DR组和NDR组患者的性别、年龄、吸烟、饮酒、既往疾病史、既往手术史、糖尿病家族史、糖尿病病程和高血压史差异有统计学意义(P0.05);饮酒、既往疾病史、糖尿病病程和高血压史是DR发生的主要影响因素。据此建立预测模型的ROC曲线下面积为0.837,临界值为0.210 7,敏感度为88.80%,特异度为61.82%。结论预测模型具有中等程度的预测价值,对糖尿病视网膜病变早期诊断具有一定的预测价值。加强糖尿病患者血压监控,限制饮酒甚至戒酒,制定科学的干预措施,有助于减轻患者视力损伤,提高患者的生活质量。     

心血管病预测模型研究进展

20世纪六七十年代,大多数工业化国家冠心病的死亡率急剧上升,并达到高峰.为预测个人冠心病发病风险,有针对性地采取干预措施,西方国家开发出以弗明汉模型(Framingham model)为代表的多种心脑血管疾病预测模型,这些模型的应用方便了临床诊断和防治,提高了公众对疾病危险因素的认识,且有利于卫生资源合理分配.为此,国内一些学者参照相关研究开发出了适合国人的心脑血管疾病预测模型,也取得了较好的效果.本文就国内外主要的心血管病的预测模型研究综述如下.     

基于奇异谱分析的ARIMA模型在山西省流感预测中的应用

目的 分析基于奇异谱分析（singular spectrum analysis, SSA）的自回归移动平均模型（Autoregressive integrated moving average, ARIMA）模型预测流感样病例 (influenza like illness, ILI) 发病趋势的可行性，为流感防控工作提供合理的预测方法。 方法 利用山西省2010年第14周-2017年第13周的流感监测资料以不同长度配比的训练集、测试集构建SSA-ARIMA模型，并与ARIMA、BP神经网络（Back propagation neural network, BPNN）、广义回归神经网络（General Regression Neural Network, GRNN）模型进行比较。采用平均绝对误差（Mean Absolute Error，MAE）、均方误差（Mean Squared Error，MSE）、均方根误差（Root Mean Squared Error，RMSE）比较各模型预测效果。 结果 模型拟合方面，SSA-ARIMA模型在预测未来一个月发病趋势时的MAE、MSE、RMSE分别为0.163、0.061、0.248；预测六个月时分别为0.161、0.061、0.248；预测一年时分别为0.168、0.066、0.256；均低于ARIMA、BPNN、GRNN。模型预测方面，在预测未来一个月发病趋势时的MAE、MSE、RMSE分别为0.056、0.005、0.068；预测六个月时分别为0.189、0.081、0.285；预测一年时分别为0.210、0.075、0.273；也均低于ARIMA、BPNN、GRNN。 结论 SSA-ARIMA模型对山西省ILI的预测效果优于ARIMA、BPNN、GRNN，可为流感预测提供科学依据。     

基因表达对膀胱癌生存结局的预测:稀疏与混合Cox模型的实证比较研究

目的 研究稀疏Cox(coxlasso)与混合Cox模型(coxlmm)在全基因表达数据中对膀胱癌预后的预测表现.方法 通过计算一致性指数(C-index)评价两种模型在膀胱癌全基因表达数据中(TCGA,GSE31684和GSE13507)的预测精度,同时在混合Cox模型中将膀胱癌的生存方差划分为临床(PCE)和转录组...     

Construction and validation of a prognostic model across several studies, with an application in superficial bladder cancer

Many models for clinical prediction (prognosis or diagnosis) are published in the medical literature every year but few such models find their way into clinical practice. The reason may be that since in most cases models have not been validated in independent data, they lack generality and/or credibility. In this paper we consider the situation in which several compatible, independent data sets relating to a given disease with a time-to-event endpoint are available for analysis. The aim is to construct and evaluate a single prognostic model. Building a multivariable model from the available prognostic factors is accomplished within the Cox proportional hazards framework, stratifying by study. Non-linear relationships with continuous predictors are modelled by using fractional polynomials. To assess the discrimination or separation of a survival model, we use the D statistic of Royston and Sauerbrei. D may be interpreted as the separation (log hazard ratio) between the survival distributions for two independent prognostic groups. To evaluate the generality of a prognostic model across the data sets, we propose 'internal-external cross-validation' on D: each study is omitted in turn, the model parameters are estimated from the remaining studies and D is evaluated in the omitted study. Because the linear predictor of a survival model tells only part of the story, we also suggest a method for investigating heterogeneity in the baseline distribution function across studies which involves fitting completely specified, flexible parametric survival models (Royston and Parmar). Our final models combine the prognostic index (obtained with stratification by study) with the pooled baseline survival distribution (estimated parametrically). By applying this methodology, we construct two prognostic scores in superficial bladder cancer. The simpler of the two scores is more suited to clinical application. We show that a three-group prognostic classification scheme based on either score produces well-separated survival curves for each of the data sets, despite identifiable heterogeneity among the baseline distribution functions and to a lesser extent among the prognostic indexes for the individual studies.     

上海市流感样病例就诊百分比时间序列分析和预测模型研究

【目的】利用季节性差分移动自回归平均模型(SARIMA)预测上海市流感样病例就诊百分比(ILI%)的发病趋势，为及时采取针对性防控措施提供重要的参考依据。【方法】对2015年第15周至2019年第52周上海市疾病预防控制中心ILI%监测数据进行时间序列分析并建立预测模型，使用前212周数据建立SARIMA模型，后36周数据评估模型预测效果。【结果】2015年第15周—2019年第52周上海市ILI%平均值为1.494%，有较明显的流行高峰出现。最终建模SARIMA(1,0,0)(2,0,0)₅₂，模型残差为白噪声序列，真实值均在预测值95%置信区间内。【结论】SARIMA(1,0,0)(2,0,0)₅₂可用于上海市ILI%的中期预测，并为全市流感流行和暴发起到预警作用。     

Practical experiences on the necessity of external validation

The validity of prognostic models is an important prerequisite for their applicability in practical clinical settings. Here, we report on a specific prognostic study on stroke patients and describe how we explored the prediction performance of our model. We considered two practically highly relevant generalization aspects, namely, the model's performance in patients recruited at a later time point (temporal transportability) and in medical centers different from those used for model building (geographic transportability). To estimate the accuracy of the model, we investigated classical internal validation techniques and leave-one-center-out cross validation (CV). Prognostic models predicting functional independence of stroke patients were developed in a training set using logistic regression, support vector machines, and random forests (RFs). Tenfold CV and leave-one-center-out CV were employed to estimate temporal and geographic transportability of the models. For temporal and external validation, the resulting models were used to classify patients from a later time point and from different clinics. When applying the regression model or the RFs, accuracy in the temporal validation data was well predicted from classical internal validation. However, when predicting geographic transportability all approaches had difficulties. We observed that the leave-one-center-out CV yielded better estimates than classical CV. On the basis of our results, we conclude that external validation in patients from different clinics is required before a prognostic model can be applied in practice. Even validating the model in patients recruited merely at a later time point does not suffice to predict how it may fare with regard to another clinic.     

可变剪接事件结合剪接因子预测乳头状肾细胞癌生存模型的建立

目的 探讨可变剪接事件结合剪接因子对乳头状肾细胞癌患者预后的预测作用。方法 从TCGA数据库下载乳头状肾细胞癌患者的临床相关资料（性别、年龄、分期等）及转录组数据，并从TCGASpliceSeq数据库获得可变剪接事件信息。通过R语言软件及程序包，使用LASSO、Cox回归模型筛选相关可变剪接事件及相关特异剪接因子，建立乳头状肾细胞癌预后模型，进行独立预后分析，并绘制剪接事件相关的调控网络。结果 通过单因素Cox比例风险回归模型确定1 405个可变剪接事件与乳头状肾细胞癌相关，包括可变受体位点（AA）事件98个、可变供体位点（AD）事件101个、可变启动子（AP）事件333个、可变终止子（AT）事件346个、外显子跳跃（ES）事件448个、外显子互斥（ME）事件3个和内含子保留（RI）事件76个。经过LASSO、多因素Cox比例风险回归分析，并结合临床数据，显示可变剪接事件（C16orf13|32924|ES、TSFM|22759|ES、MACF1|1881|ES、ATP5C1|10726|ES、UNG|24277|AP、UNKL|33077|AP）可能是乳头状肾细胞癌的独立预后风险因素（HR=1.028，95%CI：1.018～1.039，P<0.01）。通过相关性分析寻找乳头状肾细胞癌中可变剪接事件与剪接因子之间的潜在调控关系，构建调控网络，共116个剪接因子及1 059个可变剪接事件纳入分析，其中正、负相关调控事件分别为553、506个。结论 可变剪接事件结合相关临床数据对预测乳头状肾细胞癌患者预后具有一定意义，对探索乳头状肾细胞癌的研究具有重要作用。     

A framework for quantifying net benefits of alternative prognostic models

New prognostic models are traditionally evaluated using measures of discrimination and risk reclassification, but these do not take full account of the clinical and health economic context. We propose a framework for comparing prognostic models by quantifying the public health impact (net benefit) of the treatment decisions they support, assuming a set of predetermined clinical treatment guidelines. The change in net benefit is more clinically interpretable than changes in traditional measures and can be used in full health economic evaluations of prognostic models used for screening and allocating risk reduction interventions. We extend previous work in this area by quantifying net benefits in life years, thus linking prognostic performance to health economic measures; by taking full account of the occurrence of events over time; and by considering estimation and cross-validation in a multiple-study setting. The method is illustrated in the context of cardiovascular disease risk prediction using an individual participant data meta-analysis. We estimate the number of cardiovascular-disease-free life years gained when statin treatment is allocated based on a risk prediction model with five established risk factors instead of a model with just age, gender and region. We explore methodological issues associated with the multistudy design and show that cost-effectiveness comparisons based on the proposed methodology are robust against a range of modelling assumptions, including adjusting for competing risks.     

Reclassification of predictions for uncovering subgroup specific improvement

Risk prediction models play an important role in prevention and treatment of several diseases. Models that are in clinical use are often refined and improved. In many instances, the most efficient way to improve a successful model is to identify subgroups for which there is a specific biological rationale for improvement and tailor the improved model to individuals in these subgroups, an approach especially in line with personalized medicine. At present, we lack statistical tools to evaluate improvements targeted to specific subgroups. Here, we propose simple tools to fill this gap. First, we extend a recently proposed measure, the Integrated Discrimination Improvement, using a linear model with covariates representing the subgroups. Next, we develop graphical and numerical tools that compare reclassification of two models, focusing only on those subjects for whom the two models reclassify differently. We apply these approaches to BRCAPRO, a genetic risk prediction model for breast and ovarian cancer, using data from MD Anderson Cancer Center. We also conduct a simulation study to investigate properties of the new reclassification measure and compare it with currently used measures. Our results show that the proposed tools can successfully uncover subgroup specific model improvements. Copyright © 2013 John Wiley & Sons, Ltd.     

Decision curve analysis: a novel method for evaluating prediction models.

BACKGROUND: Diagnostic and prognostic models are typically evaluated with measures of accuracy that do not address clinical consequences. Decision-analytic techniques allow assessment of clinical outcomes but often require collection of additional information and may be cumbersome to apply to models that yield a continuous result. The authors sought a method for evaluating and comparing prediction models that incorporates clinical consequences,requires only the data set on which the models are tested,and can be applied to models that have either continuous or dichotomous results. METHOD: The authors describe decision curve analysis, a simple, novel method of evaluating predictive models. They start by assuming that the threshold probability of a disease or event at which a patient would opt for treatment is informative of how the patient weighs the relative harms of a false-positive and a false-negative prediction.This theoretical relationship is then used to derive the net benefit of the model across different threshold probabilities.Plotting net benefit against threshold probability yields the "decision curve." The authors apply the method to models for the prediction of seminal vesicle invasion in prostate cancer patients. Decision curve analysis identified the range of threshold probabilities in which a model was of value, the magnitude of benefit, and which of several models was optimal. CONCLUSION: Decision curve analysis is a suitable method for evaluating alternative diagnostic and prognostic strategies that has advantages over other commonly used measures and techniques.     

老年人群心血管疾病风险预测模型研究进展

风险预测模型（模型）对于老年人群心血管疾病（CVD）的一级预防具有重要意义。国内外针对老年人群构建的CVD模型共检索到15篇文献。模型的结局定义差异较大;10个模型报告时缺少方法、结果的重要信息;10个模型存在高偏倚风险;13个模型在内部验证时仅表现出中等区分度;仅有4个模型经过外部验证。老年人群CVD模型在模型算法、预测因子与结局的关联强度方面与一般人群模型存在差异,且老年人群模型的预测能力有所下降。未来仍需补充高质量的外部验证研究证据,并探索增加新的预测因子、采用竞争风险模型算法、机器学习算法、联合模型算法、改变预测时间范围等途径对模型进行优化。