ABSTRACT
Methods are presented for dealing with screening or diagnostic variables which exhibit trends over time. The number and spacing of observations is assumed to vary among patients. Criteria for establishing 'normal' trends, and for detecting extremes and pattern deviations, are developed. The methods are applied to a pregnancy study in which longitudinal data on alpha-fetoprotein in material were obtained in 1501 patients.
Subject(s)
Fetal Monitoring/methods , alpha-Fetoproteins/analysis , Birth Weight , Female , Humans , Infant, Newborn , Pregnancy , Radioimmunoassay/methods , Statistics as TopicABSTRACT
Point estimates can be generated by simple immunoassays when combined with a sequential statistical procedure. The method is called an extended simple immunoassay. A simple immunoassay for alpha-fetoprotein (counterimmunoelectrophoresis) was used to test the hypothesis that simple immunoassays are not restricted to qualitative "yet" or "no" results, but can provide quantitative results similar to radioimmunoassay. Potential benefits of this finding are discussed.