Study 1
In this study, Olds and colleagues (2004) used a randomized controlled trial design with three groups to study the long-term effects of the Nurse-Family Partnership (NFP). The participants were 735 low-income, pregnant women with no prior live births, who were recruited from 1994 through 1995 from public- and private-care settings in Denver, Colorado. Participants were randomized to control (n = 255), paraprofessional (n = 245), or nurse condition (n = 245) groups. Study participants were 85 percent unmarried and 47 percent Mexican American, 35 percent white non-Mexican American, 15 percent Black, and 3 percent American Indian/Asian.
The control group participants were provided developmental screening and referral services for their children. Treatment groups received developmental screening and referral services, plus home visits by nurses (nurse condition) or paraprofessionals (paraprofessional condition). Treatment goals for the home-visitation program consisted of improved maternal and fetal health during pregnancy, improved maternal care-giving, and improved maternal personal development by promoting planning of future pregnancies and helping women continue their education and find work. Home visits were provided from pregnancy until the child was two years old.
Two years after the intervention ended, when the children were approximately four years old, data was collected through interviews with the mothers, home observations, and child assessments conducted in the home. Outcomes were collected on approximately 85 percent of the original randomized sample. The outcomes of interest included maternal reports of subsequent pregnancies, participation in education and work, use of welfare, marriage, cohabitation, experience of domestic violence, mental health, substance use, sense of mastery, and mother-child interaction. For children, the outcomes of interest were children’s language and executive functioning and externalizing behavior problems. Aside from self-reports from the mothers, several different instruments were used to collect information on the outcome measurements. The Shipley Intelligence Scale Adult measured intelligence, the Mental Health Inventory measured adult mental health, the Mastery Scale measured mastery, the Child Behavior Checklist measured externalizing behavior, the Preschool Language Scale measured children’s language skills, and the Conflict Tactics Scale measured domestic violence.
Continuous variables were analyzed with the general linear model and dichotomous variables were analyzed with the logistic–linear model. For dichotomous correlated outcomes, generalized estimating equations, with a logit link function and assumption of an exchangeable (compound symmetry) correlation structure, were used. No subgroup analyses were conducted.
Study 2
This 2010 study by Kitzman and colleagues is a follow-up study of the effects of the NFP on children through age 12. The original sample consisted of 1,139 young, low-income, and primarily African American women who were less than 29 weeks pregnant and were recruited from an obstetric clinic in Memphis, Tennessee, from 1990 through 1991. Women were randomly assigned to receive nurse home visits (n = 228) or comparison (control group) services (n = 515).
At the 12-year follow-up assessment, the control group included 422 study participants and the nurse–home visitation group included 191 participants. The control group was mostly African American (5.7 percent of the control group were another race) and the vast majority were unmarried (1.4 percent were married). The nurse visitation group was almost entirely African American (8.4 percent of the group were another race) and unmarried (1.0 percent were married). The two groups were not significantly different, except that at the original intake the nurse-visited women lived in households with less discretionary income, higher person-per-room density, and higher scores on the household poverty index.
Women in the control group were provided free transportation for scheduled prenatal care plus developmental screening and referral for the child at 6, 12, and 24 months of age. Women in the nurse-visited treatment group were provided the same services, plus prenatal, infancy, and child home visits through the child’s second birthday.
The primary child outcomes were derived from tests of children’s academic achievement, interviews with children and parents, reviews of children’s school records, and teachers’ ratings of children. Standardized tests and end-of-year grade-point average were used for assessing reading and math achievement. The tests included the Peabody Individual Achievement Test and the Tennessee Comprehensive Assessment Program for grades 1 through 6. Information on children's use of cigarettes, alcohol, and marijuana was obtained from the child interview. Information on externalizing and internalizing problems and total behavior problems was obtained from teachers’, parents’, and children’s reports using forms from the Achenbach System of Empirically Based Assessment. Arrests by age 12 came from the parent and child reports.
The study also assessed four secondary outcomes: 1) special education placement, 2) grade retention, and 3) grade retention and conduct grades obtained from school records, and 4) sustained attention obtained from the Letter R Sustained Attention Test.
Data was analyzed using an intent-to-treat analysis. Continuous dependent variables were analyzed in the general linear model and dichotomous variables in the logistic linear model. For low-frequency count outcomes, the data was analyzed in generalized linear models with negative binomial error assumptions. No subgroup analyses were conducted.
Study 3
There have been several reports that examined the outcome results of a randomized trial conducted in Elmira, New York, looking at the effects of the NFP intervention on a study sample of 400 socially disadvantaged pregnant women with no previous births. Eckenrode and colleagues (2000) conducted a long-term follow-up of the women and their children from the original sample who received home visitation between April 1978 and September 1980. The original study sample was randomly assigned to receive one of three intervention conditions: 1) routine perinatal care (the control group), 2) routine care plus nurse home visit during pregnancy only (first intervention), or 3) routine care plus nurse home visits during pregnancy and through the child’s second birthday (second intervention). The original sample included 47 percent of mothers who were under age 19, 62 percent who were unmarried, and 61 percent who came from households classified as low socioeconomic status. There were no significant differences in the age, education, or marital status of the women in the study. The 15-year follow-up included 324 mothers and their children (81 percent of mothers who were originally randomized). The control group included 184 participants, the group that received home visitations during pregnancy only included 100 participants, and the group that received home visitations during pregnancy through the child’s second birthday included 116 participants.
At the 15-year follow-up, mothers were interviewed using a life-history calendar designed to help them recall major life events, such as births of subsequent children, marriages/partnerships, education, employment, moves, and housing arrangements. Mothers also reported their exposure to domestic violence using the violence subscale of the Conflict Tactics Scale. The measure used for analysis consisted of the total number of times the mother reported having experienced any form of partner-perpetrated violence since the birthday of the study child. Variables were also constructed reflecting frequency of major and minor violence. Minor violence included throwing items, pushing, and slapping. Major violence included kicking, biting, hitting with a hand or an object, beating, choking, threatening with a knife or gun, or using a knife or gun. New York State Child Protective Services (CPS) records were also reviewed. Reports involving either the mother as the person who perpetrated the crime or the study child as the subject were coded. Substantiated reports were abstracted to ascertain key features of the maltreatment incident.
The primary outcome measure of interest for the analysis was the number of substantiated reports over the 15-year follow-up period involving the study child, regardless of the identity of the person who perpetrated the crime, or involving the mother as the person who perpetrated the crime, regardless of the identity of the child. Maltreatment type was distinguished between neglect only and abuse only. The analyses included a 3x2x2 factorial structure: treatment (control group versus the first intervention group versus the second intervention group), maternal marital status (married versus unmarried at study registration) and social class (Hollingshead levels III or IV versus I or II at registration). The abuse and neglect outcome results were reported as incidence and log incidence. The distribution of outcomes used a Poisson log-linear model. No subgroup analyses were conducted.
