A METAANALYSIS OF HOME UTERINE ACTIVITY MONITORING

OBJECTIVE: Our purpose was to assess by metaanalysis the evidence from randomized clinical trials regarding home uterine activity monitoring . STUDY DESIGN: Six randomized controlled trials of home uterine activ ity monitoring, the same six trials reviewed by the U.S. Preventive Se rvices Task Force on home uterine activity monitoring, were studied. D ata were extracted from published reports of the six trials. In additi on, unpublished data were obtained by personal communication from the trials' principal investigators. insofar as possible, the principle of intention-to-treat was maintained. Data regarding twins were handled by use of numbers of pregnancies rather than numbers of infants as sam ple sizes. Stratified metaanalyses were conducted according to whether the trial did or did not control in study design for the nursing cont act factor that accompanies home uterine activity monitoring. In addit ion, stratified metaanalyses were conducted for singleton and twin pre gnancies. The four outcomes investigated were incidence of preterm bir th, incidence of preterm labor combined with cervical dilatation >2 cm , infant referral to the intensive care unit, and mean birth weight. R ESULTS: Overall, for all pregnancies home uterine activity monitoring was associated with a statistically significant reduction of 52% in ri sk of preterm labor combined with cervical dilatation >2 cm (relative risk = 0.48, p = 0.001) and a statistically significant increase of 86 gm in mean birth weight (p = 0.038). When stratified by singleton or twin pregnancy, the pooled results generally differed by strata. Among singleton pregnancies, home uterine activity monitoring was associate d with a statistically significant reduction of 24% in risk of preterm birth (relative risk 0.76, p = 0.037) and a statistically significant increase of 126 gm in mean birth weight (p = 0.009). Among twin pregn ancies, there was a statistically significant effect of home uterine a ctivity monitoring with a reduction of 56% in risk of preterm labor co mbined with cervical dilatation >2 cm (relative risk 0.44, p = 0.005). There were no statistically significant effects found overall and in any stratum with regard to infant referral to the intensive care unit. Metaanalyses of studies that controlled in design for the nursing con tact factor that accompanies home uterine activity monitoring showed e ither no difference or stronger pooled effects compared with metaanaly ses of those studies that did not control for nursing contact. This su ggests that the potential bias attributed to the nursing contact featu re that accompanies home uterine activity monitoring is not as appreci able as home uterine activity monitoring critics have suggested. CONCL USIONS: Metaanalysis of existing clinical trial evidence regarding hom e uterine activity monitoring reveals statistically significant benefi ts of home uterine activity monitoring. Of the outcomes investigated, home uterine activity monitoring is associated with reductions in risk s of preterm birth (in singleton pregnancies only) and preterm labor c ombined with cervical dilatation >2 cm, as well as with increased mean birth weight (in singleton pregnancies only).