EPI-Eval/nhsn-hrd

CDC NHSN医院呼吸道数据（HRD）是一个由美国疾病控制与预防中心（CDC）国家医疗安全网络（NHSN）收集的公开数据集，专注于医院呼吸道疾病监测。数据集覆盖时间从2020年8月8日至2026年4月18日，以每周为频率更新，包含美国国家级、次国家级区域和州级（共67个唯一位置ID）的数据。数据涉及三种主要呼吸道病原体：流感、SARS-CoV-2（COVID-19）和RSV。关键数据列包括新实验室确认的流感、COVID-19和RSV住院人数（每周）、床位容量和占用情况（如总配备住院床位、ICU床位占用）、ICU患者数以及按人口标准化的住院率。此外，数据集还提供季节性累计住院人数。从2024-25季节开始，该数据集作为FluSight预测中心的评分目标。数据收集在2024年11月从自愿报告转为强制报告，因此长期建模时需注意2024年11月前后数据收集制度的变化。数据集还包括解释性注意事项，例如床位计数仅包括配备人员的床位，RSV数据在2024年第46周前为自愿报告可能不完整。数据可用于呼吸道疾病趋势分析、预测模型评估和公共卫生监测。

The CDC NHSN Hospital Respiratory Data (HRD) is an open dataset collected by the U.S. Centers for Disease Control and Prevention (CDC) National Healthcare Safety Network (NHSN), focusing on hospital respiratory disease surveillance. The dataset covers the period from August 8, 2020, to April 18, 2026, with weekly updates, and includes data at national, subnational-region, and subnational-state levels (67 unique location IDs) in the United States. It involves three major respiratory pathogens: influenza, SARS-CoV-2 (COVID-19), and RSV. Key data columns include new laboratory-confirmed hospital admissions for influenza, COVID-19, and RSV (weekly), bed capacity and occupancy (e.g., total staffed inpatient beds, ICU bed occupancy), ICU patient counts, and admission rates per 100,000 population. Additionally, the dataset provides cumulative seasonal admissions. Starting from the 2024-25 season, it serves as the scoring target for the FluSight forecast hub. Data collection transitioned from voluntary to mandatory reporting in November 2024, so long-term modeling should treat pre- and post-November 2024 as different observational regimes. The dataset includes interpretation caveats, such as bed counts only including staffed beds and RSV data being sparse before epidemiological week 2024-46. It is useful for respiratory disease trend analysis, forecast model evaluation, and public health surveillance.