Chinn S
Department of Public Health Medicine, UMDS St Thomas's Campus, London, UK.
Int J Epidemiol. 1995;24 Suppl 1:S75-8. doi: 10.1093/ije/24.supplement_1.s75.
The National Study of Health and Growth (NSHG) was set up in 1972 to monitor the growth of primary school children. Areas were selected in England and Scotland by stratified random sampling. Schools within these areas were visited annually until 1982, biennially thereafter, resulting in a mixed longitudinal design. The reasons for the original design and the study as it has operated are reviewed, with advantages and disadvantages compared to the monitoring system now to be implemented by the Department of Health.
Description of the statistical and interpretive problems of monitoring rates of growth and a comparison of the response rates achieved in the two types of monitoring.
Although the design of the NSHG was selected in order to monitor rates of growth this presents statistical problems. The usable information is contained in trends in attained height rather than in rates of growth. This study has achieved an average response rate of over 95%; less than 78% can be expected from the proposed survey of households.
The small loss of efficiency of the mixed longitudinal design compared with repeated cross-sectional studies is more than compensated for by its high response rate and the comparability of data over time, neither of which can be guaranteed by the proposed survey.
全国健康与成长研究（NSHG）于1972年设立，旨在监测小学生的成长情况。通过分层随机抽样在英格兰和苏格兰选取了一些地区。1982年之前每年对这些地区内的学校进行走访，之后每两年走访一次，形成了混合纵向设计。回顾了最初设计的原因以及该研究的实施情况，并将其与卫生部即将实施的监测系统的优缺点进行了比较。
描述监测生长率的统计和解释问题，并比较两种监测方式所达到的回应率。
尽管NSHG的设计目的是监测生长率，但这带来了统计问题。可用信息包含在达到的身高趋势中，而非生长率中。本研究的平均回应率超过了95%；预计拟议的家庭调查的回应率不到78%。
与重复横断面研究相比，混合纵向设计效率上的小损失被其高回应率和数据随时间的可比性所弥补，而这两点拟议的调查都无法保证。
