Leijser Lara M, Scott James N, Roychoudhury Smita, Zein Hussein, Murthy Prashanth, Thomas Sumesh P, Mohammad Khorshid
Section of Neonatology, Department of Pediatrics, University of Calgary, Calgary, AB, Canada.
Department of Diagnostic Imaging, University of Calgary, Calgary, AB, Canada.
Pediatr Res. 2021 Aug;90(2):403-410. doi: 10.1038/s41390-020-01245-0. Epub 2020 Nov 12.
Post-hemorrhagic ventricular dilatation (PHVD) in preterm infants can be assessed with ventricular size indices from cranial ultrasound. We explored inter-observer reliability of these indices for prediction of severe PHVD.
For all 139 infants with IVH, serial neonatal ultrasound at 3 time points (days 4-7, day 14, 36 weeks PMA) were assessed independently by 3 observers with differing levels of training/experience. Ventricular index (VI), anterior horn width (AHW), and fronto-temporal horn ratio (FTHR) were measured and used to diagnose PHVD. For all, inter-observer reliability and predictive values for receipt of surgical intervention were calculated.
Inter-observer reliability for all observers varied from poor to excellent, with higher reliability for VI/AHW (ICC 0.49-0.84/0.51-0.81) than FTHR (0.41-0.82), particularly from the second week. Good-excellent inter-expertise reliability was found between observers with ample experience/training (0.65-0.99), particularly for VI and AHW, while poor-moderate when comparing with an inexperienced observer (0.28-0.88). Slightly higher predictive value for PHVD intervention (n = 12) was found for AHW (AUC 0.86-0.96) than for VI and FTHR (0.80-0.96/0.80-0.95).
AHW and VI are highly reproducible in experienced hands compared to FTHR, with AHW from the second week onwards being the strongest predictor for receiving surgical intervention for severe PHVD. AHW may aid in early PHVD diagnosis and decision-making on intervention.
While ventricular size indices from serial cUS are superior to clinical signs of increased intracranial pressure to assess PHVD, questions remained on their inter-observer reproducibility and reliability to predict severity of PHVD. AHW and VI are highly reproducible when performed by experienced clinicians. AHW from the second week of birth is the strongest predictor of PHVD onset and severity. AHW, combined with VI, may aid in early PHVD diagnosis and decision-making on need for surgical intervention. Consistent use of these indices has the potential to improve PHVD management and therewith the long-term outcomes in preterm infants.
可通过头颅超声测量的脑室大小指数评估早产儿出血后脑室扩张（PHVD）。我们探讨了这些指数在预测重度PHVD方面的观察者间可靠性。
对139例颅内出血婴儿，由3名训练/经验水平不同的观察者独立评估3个时间点（出生后4 - 7天、14天、孕龄36周）的系列新生儿超声。测量脑室指数（VI）、前角宽度（AHW）和额颞角比值（FTHR），并用于诊断PHVD。计算所有观察者间的可靠性以及接受手术干预的预测值。
所有观察者间的可靠性从差到优不等，VI/AHW的可靠性（组内相关系数ICC 0.49 - 0.84/0.51 - 0.81）高于FTHR（0.41 - 0.82），尤其是从第二周开始。经验丰富/训练充分的观察者之间具有良好到优秀的专业间可靠性（0.65 - 0.99），特别是对于VI和AHW，而与经验不足的观察者比较时可靠性较差到中等（0.28 - 0.88）。AHW对PHVD干预（n = 12）的预测值略高于VI和FTHR（AUC 0.86 - 0.96比0.80 - 0.96/0.80 - 0.95）。
与FTHR相比，AHW和VI在经验丰富者手中具有更高的可重复性，从第二周起AHW是重度PHVD接受手术干预的最强预测指标。AHW可能有助于早期PHVD诊断和干预决策。
虽然系列头颅超声的脑室大小指数在评估PHVD方面优于颅内压升高的临床体征，但关于其观察者间的可重复性和预测PHVD严重程度的可靠性仍存在疑问。经验丰富的临床医生进行测量时，AHW和VI具有高度可重复性。出生后第二周起的AHW是PHVD发生和严重程度的最强预测指标。AHW与VI相结合，可能有助于早期PHVD诊断和手术干预需求的决策。持续使用这些指数有可能改善PHVD的管理，从而改善早产儿的长期预后。
