Thirteen studies (n=4,708) were included in the review. Sample sizes ranges from 30 to 1,970 (mean 362, median 147). Most of the included studies involved some randomisation of the participants to the treatment and control groups.
Three studies reported sample size calculations. All studies used convenience samples. Less than half of the studies reported the frequency and length of home visits, or explained the rationale for implementation. The reliability and validity of the outcome measures was inconsistently reported. Drop-out rates (reported in 11 studies) ranged from 2 to 35% and were higher in the 2 longer term interventions. Training to deliver the intervention deliverers was reported in 8 studies. None of the studies achieved the highest rating for the full application of a theoretical framework to explain success factors. Three studies scored 1, 2 studies scored 2 and 8 studies scored 0.
Positive intervention effects were reported for improvement in the physical health of mother and child (4 studies, n=1,805) and, specifically, in pregnant women (2 studies, n=1,539), improved maternal mental health, improved parent-child interactions and home environment (1 study, n=181), perceptions of infant behaviour (1 study, n=30) and fewer incidents of child abuse (2 studies, n=1,539). In terms of child health, one study (n=114) found fewer low birth weight babies (less than 2,500 g) and another (n=145) reported fewer respiratory problems. Two long-term studies (over 2 years) showed improved child development and intelligence quotient (n=437). Three studies (n=1,713) showed reductions in the use of health care services, including emergency room contacts. More nursing contact was associated with less ongoing drug abuse and improved compliance with primary care in one study (n=70). The long-term sustainability of the outcomes was not addressed in the majority of studies. However, 2 studies over 3 and 15 years (n=1,139 and n=400, respectively) demonstrated positive effects upon the spacing between births of subsequent children, use of welfare services, child and substance abuse, and the number of criminal offences.
Despite several positive effects (reported above), an absence of effect was noted in 3 studies and conflicting outcomes were reported amongst others. For example, one of 3 studies reported improvements in the use of health services (the other 2 studies reported no improvements); two of 4 studies reported improvements in child development and cognition (the other 2 studies reported no improvement); and one of 4 studies reported improvements in low birthweight and pre-term birth (the other 3 studies reported no improvement).
Full details of the individual studies and their results were given.