In a case-control study, we compared the ability of two indices of ant
enatal care use, the new Adequacy of Prenatal Care Utilization (APNCU)
index and the Kessner index, to predict low birthweight. In crude ana
lyses, both indices showed a linear trend with low birthweight. After
controlling for confounding, however, the APNCU index was unrelated to
low birthweight. To learn whether the Kessner index added explanatory
information to the APNCU index (or vice versa), we regressed the APNC
U index on the Kessner index (and vice versa) and computed residuals f
or both indices. In logistic regression analyses, the residuals of the
Kessner index added meaningful information to the APNCU index, wherea
s the opposite did not occur.