A comparison of different methods for calculating overall risk scores from risk factors ascertained in a computerized obstetric information system

Abstract

Current obstetric risk-scoring systems depend upon simple addition of weighted or unweighted risk factors. In this study a population of 994 pregnant women (470 primiparae; 524 multiparae) has been examined in order to compare the use of weighted and unweighted risk scores, together with Bayes theorem, in estimating an overall risk score from the presence of individual risk factors. The findings have been related to fetal outcome in these pregnancies and expressed as receiver-operating characteristic (ROC) curves for different cutoff levels between normal and abnormal. In primiparae, the use of weighted risk factors, with or without Bayes theorem, was clearly superior to the use of unweighted factors. In multiparae there was only a marginal difference between the three approaches. The increasing use of computerised information system in antenatal care should make calculation of risk scores for every pregnant woman a practical option.

Evaluation of a Fetal Risk-Scoring System

“…Fetal outcome was unsatisfactory in 138 of the primiparas and 134 of the multiparas. For the purposes of the present study, this population was divided into two parts: the 991 women who were the topic of the previous publication 6 (referred to here as the first population); and the 1038 women immediately following this group (the second population). The frequency of the 40 risk factors ascertained at the time of presentation (Table 1) was compared in women with a favorable fetal outcome and those with an unfavorable outcome.…”

“…The advantages of using Bayes theorem have been outlined in a previous study. 6 For the purposes of the present study, the assumption was made that the factors were independent. Although this is probably not the case, it seems likely that the results would not be grossly affected by a certain measure of dependence.…”

“…Obstetric risk scoring presents a number of problems. 2 - 6 First, it is often difficult to make quantitative estimates of the numerical risk associated with a given factor; published information of this type is usually available only for the most obviously dangerous factors (such as previous stillbirth). Second, risk factors must be summed to generate an overall probability.…”