Inference following multiple imputation for generalized additive models: an investigation of the median p-value rule with applications to the Pulmonary Hypertension Association Registry and Colorado COVID-19 hospitalization data.
Inference following multiple imputation for generalized additive models: an investigation of the median p-value rule with applications to the Pulmonary Hypertension Association Registry and Colorado COVID-19 hospitalization data.
复制标题
DOI：
10.1186/s12874-022-01613-w
复制
发表时间：
2022-05-21
影响因子：
4
通讯作者：
Peterson RA
中科院分区：
文献类型：
--
作者：
Bolt MA;MaWhinney S;Pattee JW;Erlandson KM;Badesch DB;Peterson RA
数据丢失给数据分析带来麻烦；最好的情况是它们会降低研究的统计功效，最坏的情况是它们会导致参数估计出现偏差。通过链式方程进行多重插补是处理缺失数据的流行技术。然而，多重插补后组合和汇集拟合广义加性模型 (GAM) 结果的技术尚未得到很好的探索。我们在 MCAR、MAR 和 MNAR 框架下模拟缺失数据，并利用随机森林和预测均值匹配插补来研究在多重插补后将 GAM 与二元和正态分布结果组合的各种规则。我们比较了多种合并程序，包括“D2”方法、柯西组合检验和中值 p 值 (MPV) 规则。 MPV 规则涉及简单地计算和报告所有插补的中值 p 值。研究了其他临时方法，例如平均 p 值规则和单一插补方法。还检查了这些方法在汇集 B 样条结果时的可行性，以获得正常结果。然后在两个案例研究中应用这些不同的汇集技术，其中一个研究了海拔对肺动脉高压患者六分钟步行距离（正常结果）的影响，另一个研究了住院的 COVID-19 患者插管的危险因素（二分结果）。与适用于完整数据集的广义相加模型的结果相比，中值 p 值规则的表现与其他检查方法一样好（如果不是更好的话）。在备择假设为真的情况下，柯西组合检验显得过于有力，而备择方法显得有力不足，而中值 p 值规则产生的结果与完整数据分析的结果相似。对于拟合 GAM 来乘以插补数据集后的合并结果，中值 p 值是一种简单但有用的方法，它平衡了检测重要关联的能力和 I 类错误的控制能力。在线版本包含可在 10.1186/s12874-022-01613-w 获取的补充材料。
Missing data prove troublesome in data analysis; at best they reduce a study’s statistical power and at worst they induce bias in parameter estimates. Multiple imputation via chained equations is a popular technique for dealing with missing data. However, techniques for combining and pooling results from fitted generalized additive models (GAMs) after multiple imputation have not been well explored. We simulated missing data under MCAR, MAR, and MNAR frameworks and utilized random forest and predictive mean matching imputation to investigate a variety of rules for combining GAMs after multiple imputation with binary and normally distributed outcomes. We compared multiple pooling procedures including the “D2” method, the Cauchy combination test, and the median p-value (MPV) rule. The MPV rule involves simply computing and reporting the median p-value across all imputations. Other ad hoc methods such as a mean p-value rule and a single imputation method are investigated. The viability of these methods in pooling results from B-splines is also examined for normal outcomes. An application of these various pooling techniques is then performed on two case studies, one which examines the effect of elevation on a six-minute walk distance (a normal outcome) for patients with pulmonary arterial hypertension, and the other which examines risk factors for intubation in hospitalized COVID-19 patients (a dichotomous outcome). In comparison to the results from generalized additive models fit on full datasets, the median p-value rule performs as well as if not better than the other methods examined. In situations where the alternative hypothesis is true, the Cauchy combination test appears overpowered and alternative methods appear underpowered, while the median p-value rule yields results similar to those from analyses of complete data. For pooling results after fitting GAMs to multiply imputed datasets, the median p-value is a simple yet useful approach which balances both power to detect important associations and control of Type I errors. The online version contains supplementary material available at 10.1186/s12874-022-01613-w.
登录
查看更多内容
影响因子：
3.7
作者：
Liu Y;Xie J
通讯作者：
Xie J
影响因子：
4.5
作者：
FRIEDMAN, JH
通讯作者：
FRIEDMAN, JH
DOI：
10.1080/00949655.2018.1491577
发表时间：
2018-01-01
影响因子：
1.2
作者：
Schouten, Rianne Margaretha;Lugtig, Peter;Vink, Gerko
通讯作者：
Vink, Gerko
DOI：
10.2478/jccm-2021-0035
发表时间：
2022-01
期刊：
Journal of critical care medicine (Universitatea de Medicina si Farmacie din Targu-Mures)
影响因子：
--
作者：
Windham S;Hirsch K;Peterson R;Douin D;Chauhan L;Heery L;Fling C;Vukovic N;Holguin F;Zimmer S;Erlandson K
通讯作者：
Erlandson K
影响因子：
3.1
作者：
Azur, Melissa J.;Stuart, Elizabeth A.;Frangakis, Constantine;Leaf, Philip J.
通讯作者：
Leaf, Philip J.