Manuscript Detail

View Revisions

Olds, D. L., Henderson, C. R., Cole, R., Eckenrode, J., Kitzman, H., Luckey, D., et al. (1998). Long-term effects of nurse home visitation on children’s criminal and antisocial behavior: 15-year follow-up of a randomized controlled trial. JAMA: The Journal of the American Medical Association, 280(14), 1238–1244.

Model(s) Reviewed: Nurse-Family Partnership (NFP)®
Additional sources:

WWHV013051

Olds, D. L. (1999). "Long-term effects of nurse home visitation on children's criminal and antisocial behavior: 15-year follow-up of a randomized controlled trial": Reply. JAMA: The Journal of the American Medical Association, 281(15), 1377.
Manuscript screening details
Screening decision Screening conclusion
Passes screens Eligible for review
Study design details
Rating Design Attrition Baseline equivalence Reassignment Confounding factors
High Randomized controlled trial Low Established on race and SES. None None
Notes:

In 2020, HomVEE updated this review in several ways. First, HomVEE removed two findings entitled  "Ever stopped by police" from the Reductions in Juvenile Delinquency, Family Violence, and Crime domain because ACF determined that this outcome is ineligible for review by HomVEE. Second, HomVEE moved eight findings about the child running away, number of minor antisocial acts, acting out problems, and externalizing problems  from the Reductions in Juvenile Delinquency, Family Violence, and Crime domain to the Child Development and School Readiness domain because ACF determined that running away and other behavior and behavioral health outcomes belong in HomVEE's Child Development and School Readiness domain along with other behavior and behavioral health outcomes. Finally, HomVEE moved 15 findings related to measures of sexual behavior, pregnancy, alcohol, and drugs from the Child Health domain to the Child Development and School Readiness because ACF determined that all measures of child behavioral health belong in HomVEE's Child Development and School Readiness domain. 

footnote30

Submitted by user on Fri, 03/15/2019 - 14:29

The Elmira sample included two deviations from the randomization procedure. First, six housemates of women already randomly assigned and enrolled in the study were assigned to the same treatment as the women already enrolled. Second, the probability of being assigned to one of the treatment groups was increased in the last 6 months of the 30 month enrollment period. The first issue suggests a mismatch between the unit of assignment (adult in the household) and the unit of analysis, which may lead to overstating the precision of the standard errors. The second issue should lead to a weighting strategy in the analysis, so that those who were enrolled later receive less weight in the analysis. Weighting, however, was not used in these studies.

Study characteristics
Study participants The sample included pregnant, first-time mothers who were less than 30 weeks pregnant. The study actively recruited and included pregnant, first-time mothers who were less than 25 weeks pregnant, were less than 19 years old, were single parents, or had low socioeconomic status. Between April 1978 and September 1980, 500 women were interviewed and 400 were randomly assigned. This study is a longitudinal follow-up at 15 years for the women originally enrolled in the study and their children. At this follow-up, the study included data on 330 of the original mothers (177 in the program group and 152 in the comparison group) and 315 children of the original mothers (171 in the program group and 144 in the comparison group). At enrollment, on average, the women included in this study were about 19 years old and had approximately 11 years of education. Roughly 40 percent of the sample was married.
Setting The study was conducted in Elmira, a metropolitan area within a semi-rural county in the Appalachian region of New York that has approximately 100,000 residents (information obtained from other studies using the Elmira sample).
Intervention services The study included two treatment groups, which were combined for the analyses. The first treatment group received home visits from a nurse during pregnancy. The nurse visited the family every other week and made nine visits, on average, which lasted one hour and 15 minutes. The treatment group also received the screening and transportation services described below for the comparison groups. The second treatment group received the same services as the first treatment group, but the home visiting continued until the child was 2 years old. Home visits were once a week for the first month after delivery, decreasing over time to once every 6 weeks when the child was 18-24 months. Home visits focused on parent education, enhancing the women’s support systems, and linkages to community services.
Comparison conditions The study included two comparison groups, which were combined for the analyses. The first comparison group did not receive any services during pregnancy. When the children were 12 and 24 months old, they were screened for sensory and developmental problems and referred to other specialists, as appropriate. The second treatment group received free transportation (through a contract with a local taxi company) for prenatal and well-child care at local clinics and doctors’ offices. The second comparison group also received the 12- and 24-month developmental screening.
Staff characteristics and training All home visitors were nurses. No other information on training is provided.
Funding sources Senior Research Scientist Award 1-K05-MH01382-01 (Dr. Olds); the Prevention Research and Behavioral Medicine Branch of the National Institute of Mental Health (R01-MH49381) and the Assistant Secretary for Planning and Evaluation, Department of Health and Human Services (grant 96ASPE278A).
Author affiliation David L. Olds, a study author, is a developer of this model.

Findings details

Child development and school readiness
Rating Outcome measure Effect Sample Timing of follow-up Sample size Intervention group Comparison group Group difference Effect size Statistical significance Notes
High Alcohol and drug impairment (mother report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15 year follow-up 245 adolescents Adjusted mean = 0.28 Adjusted mean = 0.18 Mean difference = -0.10 Not available Not statistically significant,
p = 0.68
High Alcohol impairment (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 0.47 Adjusted mean = 0.52 Mean difference = -0.05 Not available Not statistically significant,
p = 0.35
High Ever pregnant or made someone pregnant
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents % (adjusted) = 4.00 Adjusted mean % = 3.00 Mean difference = 1.00 HomeVEE calculated = 0.18 Not statistically significant,
p = 1.00

footnote2

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the odds ratio.

High Incidence of sex partners (number)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 1.16 Adjusted mean = 1.56 Mean difference = -0.40 Not available Not statistically significant,
p = 0.90

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–cigarettes smoked per day
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 1.28 Adjusted mean = 1.30 Mean difference = -0.02 Not available Not statistically significant,
p = 0.76

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–days drank alcohol
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 1.87 Adjusted mean = 1.57 Mean difference = -0.30 Not available Not statistically significant,
p = 0.96

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–days used drugs
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 2.04 Adjusted mean = 2.28 Mean difference = -0.24 Not available Not statistically significant,
p = 0.54

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–times ran away (self report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 0.34 Adjusted mean = 0.29 Mean difference = 0.05 Not available Not statistically significant,
p = 0.07

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Number of acting out problems (teacher report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 9.47 Adjusted mean = 9.61 Mean difference = -0.14 Not available Not statistically significant,
p = 0.85
High Number of externalizing problems (self-report and mother report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 13.88 Adjusted mean = 13.73 Mean difference = 0.15 Not available Not statistically significant,
p = 0.89
High Number of internalizing problems
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 11.66 Adjusted mean = 10.58 Mean difference = 1.08 Not available Not statistically significant,
p = 0.19
High Number of minor antisocial acts (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 2.88 Adjusted mean = 2.99 Mean difference = -0.11 Not available Not statistically significant,
p = 0.86
High Alcohol and drug impairment (mother report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 277 adolescents Adjusted mean = 0.20 Adjusted mean = 0.18 Mean difference = 0.02 Not available Not statistically significant,
p = 0.96
High Alcohol impairment (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 0.50 Adjusted mean = 0.52 Mean difference = -0.02 Not available Not statistically significant,
p = 0.95
High Ever had sexual intercourse
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents % (adjusted) = 35.00 Adjusted mean % = 35.00 Mean difference = 0.00 HomeVEE calculated = 0.00 Not statistically significant,
p = 1.00

footnote2

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the odds ratio.

High Ever pregnant or made someone pregnant
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents % (adjusted) = 2.00 Adjusted mean % = 3.00 Mean difference = -1.00 HomeVEE calculated = -0.25 Not statistically significant,
p = 0.97

footnote2

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the odds ratio.

High Incidence of sex partners (number)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 1.10 Adjusted mean = 1.56 Mean difference = -0.46 Not available Not statistically significant,
p = 0.48

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–cigarettes smoked per day
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 0.91 Adjusted mean = 1.30 Mean difference = -0.39 Not available Not statistically significant,
p = 0.49

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–days drank alcohol
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 1.81 Adjusted mean = 1.57 Mean difference = 0.24 Not available Not statistically significant,
p = 0.97

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–days used drugs
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 3.55 Adjusted mean = 2.28 Mean difference = 1.27 Not available Not statistically significant,
p = 0.49

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–times ran away (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 0.23 Adjusted mean = 0.29 Mean difference = -0.06 Not available Not statistically significant,
p = 0.83

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Number of acting out problems (teacher report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 8.97 Adjusted mean = 9.61 Mean difference = -0.64 Not available Not statistically significant,
p = 0.41
High Number of externalizing problems (self-report and mother report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 13.65 Adjusted mean = 13.73 Mean difference = -0.08 Not available Not statistically significant,
p = 0.95
High Number of internalizing problems
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 11.19 Adjusted mean = 10.58 Mean difference = 0.61 Not available Not statistically significant,
p = 0.46
High Number of minor antisocial acts (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 2.54 Adjusted mean = 2.99 Mean difference = -0.45 Not available Not statistically significant,
p = 0.50
Reductions in juvenile delinquency, family violence, and crime
Rating Outcome measure Effect Sample Timing of follow-up Sample size Intervention group Comparison group Group difference Effect size Statistical significance Notes
High Ever was person inneed of supervision (PINS; self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up Unknown % (adjusted) = 8.00 Adjusted mean % = 13.00 Mean difference = -5.00 HomeVEE calculated = -0.33 Not statistically significant,
p = 0.33

footnote2

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the odds ratio.

High Incidence of arrests (mother report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 0.08 Adjusted mean = 0.12 Mean difference = -0.04 Not available Not statistically significant,
p = 0.37

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–convictions and probation violations (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 0.10 Adjusted mean = 0.27 Mean difference = -0.17 Not available Statistically significant,
p < 0.01

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–long-term school suspensions
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 0.01 Adjusted mean = 0.04 Mean difference = -0.03 Not available Not statistically significant,
p = 1.00

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–sent to youth corrections (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 0.04 Adjusted mean = 0.05 Mean difference = -0.01 Not available Not statistically significant,
p = 0.98

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–short-term school suspensions
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 0.27 Adjusted mean = 0.28 Mean difference = -0.01 Not available Not statistically significant,
p = 0.97

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Number of major delinquent acts (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy + infancy and comparison (Elmira) 15-year follow-up 245 adolescents Adjusted mean = 3.57 Adjusted mean = 3.02 Mean difference = 0.55 Not available Not statistically significant,
p = 0.48
High Ever was person inneed of supervision (PINS; self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents % (adjusted) = 11.00 Adjusted mean % = 13.00 Mean difference = -2.00 HomeVEE calculated = -0.12 Not statistically significant,
p = 0.75

footnote2

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the odds ratio.

High Incidence of arrests (mother report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 277 adolescents Adjusted mean = 0.11 Adjusted mean = 0.12 Mean difference = -0.01 Not available Not statistically significant,
p = 0.73

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–arrests (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 0.16 Adjusted mean = 0.36 Mean difference = -0.20 Not available Statistically significant,
p < 0.01

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–convictions and probation violations (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 0.06 Adjusted mean = 0.27 Mean difference = -0.21 Not available Statistically significant,
p < 0.01

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Incidence–sent to youth corrections (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 0.05 Adjusted mean = 0.05 Mean difference = 0.00 Not available Not statistically significant,
p = 0.98

footnote1

Submitted by user on Fri, 03/15/2019 - 14:29

Statistical significance reported on authors’ calculations of the log-incidence difference.

High Number of major delinquent acts (self-report)
FavorableUnfavorable or ambiguousNo Effect
Nurse visited during pregnancy and comparison (Elmira) 15-year follow-up 227 adolescents Adjusted mean = 2.79 Adjusted mean = 3.02 Mean difference = -0.23 Not available Not statistically significant,
p = 0.93

Outcome measure summary

Child development and school readiness
Outcome measure Description of measure Data collection method Properties of measure

Alcohol and drug impairment

Responses to questions regarding children’s behavioral problems attributable to alcohol and drug use were summarized with a scale ranging from 0 to 10.

Parent/caregiver report

Not reported by author

Alcohol impairment

Responses to questions regarding the effect of alcohol on five domains, (trouble with parents, trouble at school, problems with friends, problems with someone they were dating, trouble with police) were summarized in an alcohol-use behavioral problem scale with a range of 0 to 5.

Adolescent report

Not reported by author

Ever had sexual intercourse

Percentage of youth reporting that they had experienced sexual intercourse

Youth report

Not applicable

Ever pregnant or made someone pregnant

Percentage of youth reporting that they had ever been pregnant or made a sexual partner pregnant

Youth report

Not applicable

Incidence of sex partners

Number of sexual partners the youth had during the youth’s lifetime

Youth report

Not applicable

Incidence—cigarettes smoked per day

A variable constructed to characterize the total number of cigarettes the youth currently smoked per day

Youth report

Not applicable

Incidence—days drank alcohol

A variable constructed to characterize the number of days the youth drank alcohol during the previous 6 months

Youth report

Not applicable

Incidence—days used drugs

A variable constructed to characterize the number of days the youth used illegal drugs during the previous 6 months

Youth report

Not applicable

Incidence—long-term school suspensions

Number of long-term school suspensions the youth had received

Review of school records

Not applicable

Incidence—short-term school suspensions

Number of short-term school suspensions the youth had received

Review of school records

Not applicable

Incidence–times ran away

Counts of the number of times the youth had run away from home up to their 15th year

Adolescent report

Not applicable

Number of acting out problems

A scale that rated youth’s disruptive behavior in the classroom

Current Mathematics and English teacher reports

Not reported by author

Number of externalizing problems

An average of parent and child reports of externalizing behavioral problems Parent/caregiver and child report

Not reported by author

Number of externalizing problems

An externalizing problems scale representing delinquency and aggression behavioral problems was drawn from the Achenbach Youth Self Report of Problem Behaviors.

Adolescent report

Not reported by author

Number of internalizing problems

An average of parent and child reports of internalizing behavioral problems Parent/caregiver and child report

Not reported by author

Number of minor antisocial acts

Reports of antisocial and delinquent acts were factor analyzed and produced two factors, one of which was classified as minor antisocial acts

Adolescent report

Cronbach’s α = .68

Reductions in juvenile delinquency, family violence, and crime
Outcome measure Description of measure Data collection method Properties of measure

Ever was person in need of supervision (PINS)

Percentage of youth adjudicated as PINS Adolescent report

Not applicable

Incidence of arrests

Counts of the number of times the youth had been arrested up to their 15th year Parent/caregiver report

Not applicable

Incidence–arrests

Counts of the number of times the youth had been arrested up to their 15th year Adolescent report

Not applicable

Incidence–convictions and probation violations

Counts of the number of times the youth had been convicted of an original crime or a parole violation up to their 15th year Adolescent report

Not applicable

Incidence–sent to youth corrections

Counts of the number of times the youth was sent to a correctional facility up to their 15th year Adolescent report

Not applicable

Number of major delinquent acts

Reports of antisocial and delinquent acts were factor analyzed and produced two factors, one of which was classified as major antisocial acts Adolescent report

Cronbach’s α = .82