This study addresses the effect of maternal substance abuse on the cost of neonatal care in a sample of all singleton live births in Maryland in 1991. Most other cost studies have analyzed data from only one hospital; we analyzed data from 54 hospitals and therefore can control for individual hospital effects and correlation of observations within hospitals. We find that maternal drug abuse has a significant positive effect on total hospital charges, length of stay, and average daily charges, with the increase in length of stay being proportionally greater than the increase in average daily charges. Maternal alcohol abuse also has a positive effect on hospital charges and length of stay, but the effects are not statistically significant. About half the effect of drug abuse on total charges works indirectly through premature birth and other comorbidities.
