Tocolytic treatment for the management of preterm labor: a review of the evidence.

Preterm labor is often a prelude to early births and the significant attendant burden of infant morbidity and mortality. Treatment consists of bedrest, hydration, pharmacologic interventions, and combinations of these. We systematically reviewed the effectiveness of tocolytics to stop uterine contractions (first-line therapy) or maintain quiescence (maintenance therapy). Our objective was to evaluate the evidence on the benefits and harms of five classes of tocolytic therapy for treating uterine contractions related to preterm labor--beta-mimetics, calcium channel blockers, magnesium, nonsteroidal anti-inflammatory agents, and ethanol. Reports of randomized controlled trials and other study designs in English, French, and German identified from searches of MEDLINE, EMBASE, specialized databases, bibliographies of review articles, unpublished literature, and discussions with investigators in the field were identified. Studies on women with preterm labor between 1966 and February 1999 that met our inclusion criteria were included. Through dual review, we abstracted the following information: study design and masking; definitions of preterm labor and successful tocolysis; patient inclusion/exclusion characteristics; patient demographic characteristics; drug and cointerventions; and numerous birth, maternal, and neonatal outcome measures. Of the 256 articles evaluated, we abstracted data from 60 first-line and 15 maintenance studies. Of these, 16 first-line and 8 maintenance studies met more stringent requirements for meta-analyses. Studies of first-line tocolysis (grade Fair) reveal a mixed outcome pattern with small improvement in pregnancy prolongation and birth at term relative to placebo. Data were insufficient to show directly a beneficial effect on neonatal morbidity or mortality. Ethanol was less beneficial than, and beta-mimetics were not superior to, other tocolytic options. Maintenance tocolytics (grade Poor) showed no improvements in birth or infant outcomes relative to placebo; these results were confirmed through meta-analysis. In contrast to other tocolytic treatments, maternal harms from beta-mimetics were rated High; all tocolytics were rated as Low risk for short-term neonatal harms. Management of uterine contractions with first-line tocolytic therapy can prolong gestation. Among the tocolytics, however, beta-mimetics appear not to be better than other drugs and pose significant potential harms for mothers; ethanol remains an inappropriate therapy. Continued maintenance tocolytic therapy has little or no value.