Josephson A, Karanfil L, Alonso H, Watson A, Blight J
Epidemiology Department, SUNY-Health Science Center, Brooklyn 11203.
Am J Med. 1991 Sep 16;91(3B):131S-137S. doi: 10.1016/0002-9343(91)90358-5.
Because nosocomial infection rates vary by hospital area and service, most infection control programs calculate area-specific rates to augment the reporting of their hospital-wide data. Rate development is often limited by the availability of appropriate specific denominator data to support important comparisons. Our university hospital reports a 20 month experience in which numerator data was collected as per the National Nosocomial Infections Surveillance System criteria for hospital-wide, high-risk nursery and ICU surveillance. These data were then combined with data in our hospital's patient-specific denominator file. This has enabled the development of risk-specific infection rates based on the analytic control of important variables available in both the numerator and denominator files. We found rate differences that were length of stay cohort specific, hospital day specific, age specific, birthweight specific, and survival cohort specific when examining our data by both the cumulative incidence and incidence density methods.
由于医院感染率因医院区域和服务而异，大多数感染控制项目会计算特定区域的感染率，以补充其全院数据的报告。感染率的计算往往受到能否获得适当的特定分母数据以支持重要比较的限制。我们的大学医院报告了20个月的经验，在此期间，按照国家医院感染监测系统的标准收集了全院、高危新生儿病房和重症监护病房监测的分子数据。然后，这些数据与我院患者特定分母文件中的数据相结合。这使得能够基于分子和分母文件中可用的重要变量的分析控制，制定特定风险的感染率。当我们通过累积发病率和发病密度方法检查数据时，发现感染率差异存在于住院时间队列特定、住院日特定、年龄特定、出生体重特定和生存队列特定等方面。
