Replication Data for: Reliability and validity of the World Health Organization reading standards for paediatric chest radiographs used in the field in an impact study of Pneumococcal Conjugate Vaccine in Kilifi, Kenya

<p>Radiologically-confirmed pneumonia (RCP) is a specific end-point used in trials of Pneumococcal Conjugate Vaccine (PCV) to estimate vaccine efficacy. </p>

<p>The dataset is for a study which was set up to ascertain repeatability and reliability of paediatric chest radiographs(CXR) interpretation using percent agreement and Cohen’s kappa in
a study of the impact of PCV on pneumonia. The study was conducted in Kilifi County Hospital(KCH) between April 2006 and March 2014. All children aged ≥2 months to <12 years who were admitted to KCH and residents of Kilifi Health and Demographic Surveillance System (KHDSS) were eligible for the study.
Radiographs were obtained from 2,716 children admitted with clinically defined severe or very-severe pneumonia. The images were randomly grouped into batches of approximately 100 images each. To assess repeatability, a 5% random sample was presented twice. All the images were read once by two local primary readers.
Discordant readings and 13% of concordant readings were arbitrated by a panel of three expert radiologists. The images were read between January 2013 and October 2014.
</p>

<p>The findings were that Intra- and inter-observer agreements on interpretations of RCP are moderately good to high. Good sensitivity and high specificity make RCP, determined in the field, a suitable endpoint to evaluate vaccine effectiveness. </p>

<p>放射学确诊肺炎（Radiologically-confirmed pneumonia，RCP）是肺炎球菌结合疫苗（Pneumococcal Conjugate Vaccine，PCV）临床试验中用于评估疫苗效力的特定终点指标。</p>

<p>本数据集用于一项研究，该研究旨在通过百分比一致性和科恩kappa系数，确定在评估PCV对肺炎影响的研究中儿童胸部X光片（chest radiographs，CXR）解读的可重复性和可靠性。研究于2006年4月至2014年3月在基利菲县医院（Kilifi County Hospital，KCH）开展。所有年龄≥2个月且<12岁、入住KCH并为基利菲健康与人口监测系统（Kilifi Health and Demographic Surveillance System，KHDSS）居民的儿童均符合研究条件。研究获取了2716名因临床确诊的重度或极重度肺炎入院儿童的X光片。这些图像被随机分组为每批约100张图像。为评估可重复性，随机抽取5%的样本进行两次呈现。所有图像由两名本地初级阅片者各读取一次。不一致的读数以及13%的一致读数由三名专家放射科医生组成的小组进行仲裁。图像读取工作于2013年1月至2014年10月期间完成。</p>

<p>研究结果表明，RCP解读的观察者内和观察者间一致性中等至高度良好。良好的敏感性和高特异性使现场确定的RCP成为评估疫苗有效性的合适终点指标。</p>