Chen Ying-Ying, Huang Hai-Xing, Xiao Min
Department of Obstetrics and Gynecology, Hangzhou Linping District Maternal & Child Health Care Hospital, Hangzhou, Zhejiang, China.
Am J Reprod Immunol. 2025 May;93(5):e70086. doi: 10.1111/aji.70086.
Premature rupture of membranes (PROM) can lead to adverse maternal and neonatal outcomes. Identifying predictors of length of hospital stay (LOS) in PROM patients can aid in improving management and prognosis.
The Medical Information Mart for Intensive Care IV database was searched to acquire data on PROM individuals. A univariable analysis, LASSO regression, and multivariable generalized linear model (GLM) were adopted to identify factors closely related to LOS. Next, the change trends in the key variable over multiple time points were determined using a latent class trajectory model (LCTM). The association of these trajectories with LOS was analyzed using GLMs.
The retrospective study finally enrolled 868 samples. The significant factors included initial hemoglobin, chorioamnionitis, premature birth, and rupture to delivery interval <24 h. Notably, higher initial hemoglobin level was independently associated with shorter LOS (β = -0.599; 95% CI = [-0.846, -0.352]; p < 0.05). The stratified analysis confirmed this association in all subgroups except those with chorioamnionitis. The LCTM identified three hemoglobin trajectories. Compared to other trajectory groups, the "lowest, declining" group had longer LOS (p < 0.05).
PROM patients with lower initial hemoglobin levels and the "lowest, declining" trajectory group were linked to longer LOS. Monitoring and managing hemoglobin levels is important in improving patient outcomes.
胎膜早破（PROM）可导致不良的母婴结局。识别胎膜早破患者住院时间（LOS）的预测因素有助于改善管理和预后。
检索重症监护医学信息集市IV数据库以获取胎膜早破患者的数据。采用单变量分析、套索回归和多变量广义线性模型（GLM）来识别与住院时间密切相关的因素。接下来，使用潜在类别轨迹模型（LCTM）确定关键变量在多个时间点的变化趋势。使用广义线性模型分析这些轨迹与住院时间的关联。
这项回顾性研究最终纳入了868个样本。显著因素包括初始血红蛋白、绒毛膜羊膜炎、早产以及破膜至分娩间隔<24小时。值得注意的是，较高的初始血红蛋白水平与较短的住院时间独立相关（β = -0.599；95%置信区间 = [-0.846, -0.352]；p < 0.05）。分层分析在除绒毛膜羊膜炎患者之外的所有亚组中均证实了这种关联。潜在类别轨迹模型识别出三种血红蛋白轨迹。与其他轨迹组相比，“最低、下降”组的住院时间更长（p < 0.05）。
初始血红蛋白水平较低的胎膜早破患者以及“最低、下降”轨迹组与较长的住院时间相关。监测和管理血红蛋白水平对改善患者结局很重要。
