Methodologic standards for controlled clinical trials of early contact and maternal-infant behavior.

To provide an objective evaluation of published studies on the effect of early contact on subsequent maternal-infant behavior, a set of 11 methodologic standards generally applicable to controlled clinical trials of perinatal care was developed. Sixteen reports of early contact trials were assessed and seven of the 11 standards were found to be satisfactorily fulfilled. The four "problem" standards were: adequate definition of subjects, randomization, subject bias, and treatment contamination (care giver) bias. Of the five best trials fulfilling eight or more of the standards, three reported a beneficial effect of early contact, while two demonstrated no effect. The evidence that early contact improves subsequent maternal-infant behavior thus remains inconclusive. It is urged that for future research in this domain more attention be given to adequate subject definition, strict randomization procedures, and safeguards against bias by the subjects or their care givers.