Antepartum evaluation of monochorionic diamniotic (MD) twins; MD-twin score: a new scoring method for perinatal...

Our purpose was to establish a new scoring method to survey monochorionic diamniotic (MD) twins during antepartum periods.

A retrospective study was performed regarding MD twins delivered between January 1992 and July 1996. Maternal and neonatal records were assessed for the following 5 perinatal variables; birth-weight discordance, amniotic-fluid discordance, hydrops fetalis, umbilical-cord insertion, and fetal-heart-rate monitoring. Each variable was coded as normal or abnormal and then assigned an arbitrary weight of 0 if normal and 1 if abnormal, yielding a range of scores from 0 (all normal) to 5 (all abnormal). The relationships between individual variables and their combinations and the outcome of pregnancy was determined. A poor pregnancy outcome consisted of intrauterine death, neonatal death, or neurological sequelae of at least one twin. The 5-variable combination was termed as the MD-twin score. A chi-square test and logistic regression analysis were used to determine statistical significance.

There were 59 MD pregnancies, of which 13 pregnancies resulted in a poor outcome. The single variable that most likely contributed to a poor outcome was amniotic-fluid discordance. All 35 pregnancies with an MD-twin score of < or = 2 had a good outcome. There were 14 pregnancies with a score of 3, and 21% of them had a poor outcome. All of the pregnancies with a score of > or = 4 had a poor outcome. When we chose the MD-twin score of 3 as the critical point for a poor outcome, the likelihood ratio statistics became the highest of any single variable or any combination of variables.

The MD-twin score predicted poor outcomes better than did any single variable or combination of variables.