Zeng G, Ding Y P, Cheng Y K
Chinese Academy of Preventive Medicine, Division of Epidemiology, Beijing.
Zhonghua Liu Xing Bing Xue Za Zhi. 1997 Oct;18(5):270-4.
A systematic analysis, including retrospective test by means of correlation analysis and extrapolation test was carried out to demonstrate a new hypothesis named Z-D phenomenon which states that the skew of a seasonal case distribution curve of an infectious disease may appear to be the omen of its epidemic trend in the following year. A total number of 353 data sets belong to 17 notifiable infectious diseases in 29 provinces and the whole nation were chosen. Data from retrospective test showed that 91.2% (322/353) of the correlation coefficients of between the skew of the seasonal distribution curve of the previous year and the morbidity variation of its following year were negative. The extrapolation test showed that 68.1% (203/295) of the data sets accorded with the actually epidemic situation at the level of r < -0.5. The results proved the existence of Z-D phenomenon may serve as an indicator for the purpose of forecasting. Using this systematic analysis, we found that Z-D phenomenon was more expressive in viral hepatitis, pertussis, epidemic cerebrospinal meningitis and scarlet fever than in other diseases.
进行了一项系统分析，包括通过相关性分析进行回顾性检验和外推检验，以证明一种名为Z-D现象的新假说，该假说指出，传染病季节性病例分布曲线的偏斜可能预示着其下一年的流行趋势。选取了全国29个省份17种法定传染病的353个数据集。回顾性检验数据显示，上一年季节性分布曲线的偏斜与其下一年发病率变化之间的相关系数有91.2%（322/353）为负。外推检验显示，在r < -0.5的水平上，68.1%（203/295）的数据集符合实际流行情况。结果证明Z-D现象的存在可作为预测的一个指标。通过这种系统分析，我们发现Z-D现象在病毒性肝炎、百日咳、流行性脑脊髓膜炎和猩红热中比在其他疾病中表现得更明显。
