住院儿童传染性疾病就诊月份分布的对应分析

Correspondence Analysis on the Monthly Distribution of Infectious Diseases in Hospitalized Children

目的:应用对应分析探讨住院儿童传染性疾病与就诊月份的相关性。方法:从病案系统中导出2013年至2015年住院患儿(0~18岁)国际疾病分类(international classification of diseases,ICD)编码为A和B的病案首页信息,对顺位在前十位的病种(ICD-10)与就诊月份的关系进行对应分析。结果:对应分析结果显示,第一维惯量0.138,特征根为0.372,占总惯量构成比的61.1%;第二维惯量0.055,特征根为0.234,占总惯量构成比的24.3%;前二维惯量累积贡献率为85.4%,说明二维因子能解释资料的大部分信息,对应分析较好地揭示了月份和传染性疾病病种之间的联系(χ2=5 215.891,P<0.001)。对应分析因子负荷图提示,流感嗜血杆菌感染、呼吸道合胞体病毒感染、衣原体感染与2月、3月、4月关系密切;轮状病毒肠炎与11月、12月、1月关系密切;细菌性肠道感染与10月、11月关系密切;肺炎支原体感染、巨细胞病毒感染、猝发疹(第六病)与7月、8月关系密切;疱疹性咽峡炎、手足口病与5月、6月关系密切。结论:儿童传染性疾病呈现不同的月份分布特点。通过对应分析图,可以帮助临床医生更好的掌握疾病的高发时间,做好临床诊疗。

Objective： To explore the monthly distribution of infectious diseases in hospitalized children. Methods： The information about the children（ 0 to 18 years old） hospitalized on the first page of the medical records on the first page of the medical records from 2013 to 2015,which was encoded as A and B according to international classification of diseases（ ICD） was exported from the medical record system,with correspondence analysis made about the relationship between diseases（ ICD-10） and the admission month. Results： Correspondence analysis results showed that the first dimension of inertia was 0. 138 and its characteristic root was0. 372,accounting for 61. 1 % of the total,while the second dimension of inertia was 0. 055 and its characteristic root was 0. 234,accounting for 24. 3 %. The cumulative contribution from the first two dimensions was 85. 4 %,which could explain the most information of the data. The correspondence analysis effectively found the exact relationship between diseases and the hospitalized month（ χ2= 5 215. 891,P〈0. 001）. The factor loading diagram showed that haemophilus influenzal infection,respiratory syncytial virus infections and chlamydia infection were closely related with February,March and April; rotavirus enteritis was closely related with November,December and January; Bacterial intestinal infection was closely related with October and November; mycoplasma pneumoniae infection,cytomegalovirus infections,exanthema subitum（ Six Disease） was closely related with July and August; herpangina and hand,foot and mouth disease was closely related with May and June. Conclusion： Most of the hospitalized children diseases present different distribution characteristics in different months. To master the occurrence time of diseases can provide guidelines for the allocation of resources for clinical diagnosis and treatment.