A critical review of studies of newborn discharge timing.

The duration of hospitalization for newborns has declined dramatically, driven by efforts to control health-care costs as well as by efforts to demedicalize childbirth. In order to determine the clinical basis for this practice, the quality of the published literature on discharge timing was analyzed. Thirteen experimental or quasi-experimental studies were retrieved through a computer search. Seven characteristics that influenced the quality of these studies were reviewed: research design; measures of effect; sample descriptions; statistical methods; reliability measures; sample size; and the definition of early discharge, including the use of any related interventions. Although all 13 studies suggest that there are no differences between infants discharged early and their compeers, these studies have three limitations. First, with one exception, these reports are from hospitals where well-defined assessment and follow-up protocols have been established, potentially limiting their wide applicability. Second, these studies lack statistical power to assess the likelihood of rare events such as readmission. Third, few studies report outcomes other than readmission and medical conditions diagnosed within 1 to 6 weeks. Early discharge as the standard of care for well newborns has not been well established by empirical studies. Pediatricians and local public health officials have a responsibility to assure that the health objectives of hospitalization are met whether this occurs in the hospital or through other mechanisms, such as routine home visiting.