Princeton University

Universal Reach at BirthAuthor(s): Kenneth A. Dodge and W. Benjamin Goodman

Source: The Future of Children , Vol. 29, No. 1, Universal Approaches to Promoting
Healthy Development (SPRING 2019), pp. 41-60
Published by: Princeton University

Stable URL: https://www.jstor.org/stable/10.2307/26639555

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Princeton University is collaborating with JSTOR to digitize, preserve and extend access to The
Future of Children

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
 Universal Reach at Birth: Family Connects

Universal Reach at Birth: Family Connects

Kenneth A. Dodge and W. Benjamin Goodman

Summary
How do we screen all families in a population at a single time point, identify family-specific
risks, and connect each family with evidence-based community resources that can help them
overcome those risks—an approach known as targeted universalism? In this article, Kenneth
A. Dodge and W. Benjamin Goodman describe Family Connects, a program designed to do
exactly that.
Developed and tested in Durham, NC, Family Connects—now in place at 16 sites in
the United States—aims to reach every family giving birth in a given community. The
program rests on three pillars. The first is home visiting: trained nurses (or other program
representatives) welcome new babies into the community, typically at the birthing hospital,
then work with the parents to set up one or more home visits when the baby is about three
weeks old so they can identify needs and connect the family with community resources. The
second pillar, community alignment, is an assembly of all community resources available to
families at birth, including child care agencies, mental health providers, government social
services, and long-term programs for subgroups of families with identified needs, such as
Healthy Families and Early Head Start. The third pillar, data and monitoring, is an electronic
data system that acts as a family-specific psychosocial and educational record (much like an
electronic health record) to document nurses’ assessments of mother and infant, as well as
connections with community agencies.
In randomized clinical trials, Family Connects has shown promising results. Compared to
control group families, families randomly assigned to the program made more connections to
community resources. They also reported more positive parenting behaviors and fewer serious
injuries or illnesses among their infants, among other desirable outcomes. And in the first
five years of life, Family Connects children were significantly less likely to be subject to Child
Protective Services investigations than were children in a control group.

www.futureofchildren.org
Kenneth A. Dodge is the Pritzker Professor of Public Policy and a professor of psychology and neuroscience at Duke University. W.
Benjamin Goodman is a research scientist at the Duke Center for Child and Family Policy, a senior fellow at the Center for Child and
Family Health, and the director of research and evaluation at Family Connects International. The authors acknowledge generous
support from The Duke Endowment, the National Institutes of Health, and the Pew Center on the States.

VOL. 29 / NO. 1 / S PR ING 2019 41

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
S
Kenneth A. Dodge and W. Benjamin Goodman

ince University of Colorado community resources. We then describe

pediatrician Henry Kempe first findings from three independent evaluations
identified battered child syndrome of Family Connects. These findings suggest
in 1962, most of the nation’s that the program can be implemented
efforts in this area have been with broad reach, high quality, and positive
directed toward protecting and treating impacts for infants and families. We conclude
children after the fact of maltreatment. by discussing possibilities and challenges for
This is a never-win situation because the disseminating and sustaining the model, as
maltreated children keep on coming. More well as future opportunities for innovation.
recently, researchers, service providers, and
policy makers have been shifting toward Challenges to Scaling Up Targeted
prevention of maltreatment. But most of Interventions
these efforts have been underfunded or
Several major early intervention programs,
have targeted a small number of children
such as Nurse Family Partnership and
in a community. Frustrated by the modest
Early Head Start, are long-term, intensive,
(at best) overall impact of child protection
and expensive.2 To save money, they limit
programs that serve only a small number of
eligibility to a subpopulation based on
families in a community, policy makers and
demographic characteristics. The logic
scholars have called for new approaches that
behind child abuse prevention programs that
seek population impact—that is, lowering
target a demographic subgroup is efficiency:
the maltreatment rate for all children and
families in a particular area, such as a state, if we can identify a subgroup that is known
county, city, or school system.1 to be at high risk for maltreatment, then
restricting intervention to that subgroup
Achieving population impact requires more should save money because the intervention
than simply scaling up a proven intervention won’t be “wasted” on a low-risk group that
that had been delivered to only a small would most likely have a healthy outcome
number of children. It requires an approach even without costly intervention. This is a
that involves the entire community of service sensible strategy when three criteria are met:
providers, policy makers, and population of
families from the outset. To illustrate this 1. The targeted subgroup includes a high
point, in this article we identify challenges proportion of the problem outcome
that occur when attempting population cases;
impact by developing and scaling up
2. Intervention can encompass a high
programs targeted only to small subgroups.
proportion of this targeted group; and
Next, we introduce the theoretical model
for Family Connects, a program that 3. Intervention effectively lowers the rate
seeks to overcome these challenges and of problem outcomes in this group.
to improve population indicators of infant
health, wellbeing, and child maltreatment These criteria are not easily achieved in
through collaboration with the community of child maltreatment prevention programs.
intervention providers, brief postnatal home The first problem with targeting subgroups
visits to all birthing families, and family- of participants in child maltreatment
specific connections between families and prevention is that risk exists across all

42 T H E F U T UR E OF C HI L DRE N

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
 Universal Reach at Birth: Family Connects

demographic groups.3 Even though relative of all families in the targeted group who
risk may be higher in some subgroups actually receive the intervention). Although
than in the rest of the population, risk still randomized controlled trials (RCTs), in
accrues in supposedly lower-risk groups. which a group of families who receive the
And because these lower-risk groups intervention are compared to a control
are larger, they account for most child group of families who don’t, often report
maltreatment cases. This is an example a high participation rate, families in such
of the “prevention paradox” first noted by trials represent a unique subgroup of the
epidemiologist Geoffrey Rose in 1981, in targeted population—they are the ones who
which the majority of cases of a negative have provided prior written consent to get
outcome occur in populations at low or into the study. Nonconsenting families never
moderate risk for that outcome, because get into the study, but they are still part of
those at highest risk represent only a small the targeted population. When targeted
portion of the population.4 In the case of interventions are rolled out and scaled up in a
child maltreatment, imagine a targeted community, the participation rate is typically
subgroup (for example, low-income, first- lower than during the trials, a loss described
time mothers) that constitutes perhaps 20 as the “scale-up penalty” by Northeastern
percent of the full population and has a two- University criminologist Brandon Welsh and
fold higher risk for child maltreatment than Nurse-Family Partnership developer David
the rest of the population. In this scenario, Olds.5 One reason for the low penetration
two-thirds of all maltreatment cases will rate is that stigma makes some families
occur in the nontargeted 80 percent that hesitant to participate. If the targeting factor
will never receive intervention. Even if an is high risk for abuse, some families might not
intervention with the targeted subgroup is want to be identified as being in that group.
highly effective (say, cutting maltreatment
in half), the full effect would be to reduce Another challenge in reaching all families
population-level maltreatment by only in the targeted group is that funding
17 percent. Many problem cases will be rarely allows for saturation of the targeted
missed if the intervention is restricted to one population, partly because the cost would
subgroup and the impact on the population be prohibitively high.6 Penetration rates will
drop even further because targeted programs
as a whole will be small.
rarely have the funds to saturate the eligible
population. Even more worrisome, limited
funding opens the door to cherry-picking
Because lower-risk groups participants so that the families that enroll in
are larger than higher-risk the program are at relatively low risk because
they are high in compliance or motivation,
groups, they account for most and the highest-risk and most needy families
child maltreatment cases. are left out.

Another problem with interventions

The second problem is that intervention that target a small subgroup is that the
with targeted subgroups rarely has a intervention must have a large impact to
high penetration rate (the proportion achieve population-level outcomes. Although

VOL. 29 / NO. 1 / S PR ING 2019 43

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman

numerous targeted programs achieve a Family Partnership are the neurosurgery

statistically significant impact compared with of child abuse prevention, and universal
a control group, the size of the intervention approaches are the well-baby pediatric care.
effect may not be large enough to affect the We need both to achieve a comprehensive
full population. system of developmental psychosocial-
educational care.
Attractiveness and Challenges of
Universality A universal approach to early intervention
has several advantages.7 Programs restricted
The problems with targeted interventions we to the poor are, rightly or wrongly, popularly
describe don’t mean that they fail to help a considered “poor programs.” Unless
subgroup of families promote their infants’ participation brings outright cash payments,
healthy development. A targeted strategy has potential participants might may not join
advantages, of course. Risk among the target at high rates. Popular and political support
group is indeed higher, by definition. Also, is more likely if a program is offered to
the content and focus of the intervention everyone in the community giving birth,
can be tailored to a relatively homogeneous rather than a select few, and such support
group of families that may have common increases the likelihood of funding. Over
needs and may respond in a similar way. time, universal reach means that popular
Further, not all families in a community support is more likely to continue and
are likely to need the long-term, intensive funding is more likely to be sustained.
services that such programs provide. Some Because of economies of scale, universal
families may respond favorably to short-term reach means that the per-family cost can
programs or alternate interventions that be reduced and recruitment can be less
meet their particular needs. The challenge complicated, less ambiguous, and less
is to reach all families in a population and awkward and stigmatizing.
then quickly triage and match interventions
to a family’s clinically assessed needs. Rather A universal goal leads to a very important
than force communities to choose between shift in aspiration and focus. If an
universal and targeted intervention, the intervention is directed at an individual
best strategy may be to embed targeted family or only a small subset of families, it
interventions in a universal strategy necessarily prioritizes helping the family
that reaches the entire population while respond to a fixed system of community
offering intensive interventions for targeted resources: such an intervention teaches a
subgroups. A useful analogy is our health parent how to be first in line for the best
care system for young children: pediatricians child care facility, cash payments from
see children universally for well-baby visits, government, and health services. The
during which they screen and triage children intervention makes no effort to improve the
to identify subgroups that could benefit from overall quality of childcare in the community
specialized services, such as neurosurgery, or the efficiency with which families in
speech therapy, or ear tubes. Candidates general can connect to community resources.
for neurosurgery are not selected based If an evaluation of the intervention’s
on demographics but on clinically assessed effectiveness compares a small number of
need. Targeted interventions like the Nurse- participating families to families in a control

44 T H E F U T UR E OF C HI L DRE N

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
 Universal Reach at Birth: Family Connects

group, it is plausible that the intervention approaches that are both cost-beneficial
could be shown to improve the lot of these and that minimize total cost to make the
families. What happens when this program approach palatable in challenging financial
is brought to scale and every family tries to and political times.
be first in line? Only when universality and
population impact become the goals will In addition, universal reach doesn’t
program developers find it cost-effective mean that every family receives the same
to direct intervention not only at individual intervention program and the same dollar
families but also—to encourage better- expenditure. Returning to the analogy of
coordinated efforts to support children and pediatric care, universal reach is achieved
families—at community agencies. by matching every child with a primary care
physician at birth. If families don’t voluntarily
Whether child policy should be targeted identify a pediatrician, the birthing hospital
or universal was the central question when typically matches them with a provider,
public schooling was first considered two even if that means a clinic. Then the family
centuries ago. Some advocates argued is encouraged to attend a series of age-
that middle-class families would find ways related well-baby visits that include physical
to get tutoring and other schooling for examination and assessment; brief, universal,
their children even without government developmentally appropriate interventions
support, and that tax dollars should be (for example, “Have your baby sleep on her
conserved. Other advocates argued that back”); brief, family-specific interventions
universal public education would be of (for example, discussion of breast-feeding
higher quality. Universal public education and its challenges); and referral to specialists
won out, of course. It’s difficult to imagine when problems are identified (for example,
a well-functioning public education system ear tubes for otitis media or an oncology
that doesn’t include children from all referral for leukemia). Of course, not every
backgrounds, even recognizing that many child gets neurosurgery, which is reserved for
affluent families send their children to the few whose clinical diagnoses show it is
private schools. needed.

Universal reach doesn’t necessarily mean What Families Need at Birth

a single program with a single funding
source and mandatory attendance by all. It would be fortuitous if developmental
Pediatric care again provides an analogy science could identify a single environmental
and an example of near-universal reach or family factor at birth that predicts
in a voluntary system in which cost and the majority of variance in important
funding are challenges. In the United child outcomes. But empirical evidence
States, pediatric care is supported with indicates that many factors, ranging from
multiple funding sources, both public (like family financial instability to parents’
Medicaid and CHIP) and private (like health mental health problems, provide unique
insurance). For universal programs that aim incremental predictions. One factor, such as
to prevent child maltreatment and promote a mother’s substance abuse, might account
healthy development, funding will be an for a maltreatment outcome in one family,
important issue, pushing clinicians to develop whereas a different factor, such as financial

VOL. 29 / NO. 1 / S PR ING 2019 45

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman

instability, might account for maltreatment in specialist, such as a urologist or surgeon

another.8 who can deliver targeted intervention when
needed. This model forms the basis for
Family Connects.
The task becomes one of
engineering: How do we The Family Connects Model

screen all families at a single Family Connects aims to reach every family
giving birth in a community so that it can
time point, such as birth,
identify family-specific risks and needs
identify family-specific risks and then connect each family with the
and needs for intervention, community resources to meet those needs,
strengthen and enhance the parent-child
and connect each family with relationship, and improve parent and child
evidence-based community wellbeing.10 For communities to afford
resources to address that implementing the model universally, the
per-infant cost must be modest. Costs are
family’s risk? contained by guidelines that restrict the
number of intervention contacts (in-person
visits, phone calls, etc.) to a maximum
These findings suggest that no single
of seven and by limiting the time period
intervention can successfully resolve risk in
all families.9 Instead, different interventions to the first 12 weeks (except in unusual
will be necessary for different at-risk families. circumstances, such as a long-term stay
Some prevention programs rely only on in a neonatal intensive care unit). The
demographic characteristics to identify a program isn’t a continuous intervention
subgroup for targeted intervention (for or a case management system. Instead, it
example, low-income first-time mothers), consists of time- and cost-limited outreach
but the evidence indicates that clinical in the spirit of a public health model. In
characteristics such as maternal depression or implementations thus far, the total cost
parental substance abuse provide a stronger has averaged between $500 and $700
basis for targeted interventions. per infant birth, an amount we believe
is affordable in communities where the
The task becomes one of engineering: cost of public education totals more than
How do we screen all families at a single $8,000 per older child annually. Of course,
time point, such as birth, identify family- funding for public education comes from
specific risks and needs for intervention, and a combination of local, state, and federal
connect each family with evidence-based sources; we believe funding for public
community resources to address that family’s health approaches like Family Connects
risk? This sort of targeted universalism has can also come from combined local,
been pioneered in other areas. For example, state, and federal sources if community
it describes the state of optimal health leaders can figure out how to braid these
care delivery in which a general pediatric resources. Because the program’s fixed-cost
practitioner sees every family, assesses child- infrastructure expenses are relatively large,
specific risk, and then refers the child to a it can’t be implemented at modest cost

46 T H E F U T UR E OF C HI L DRE N

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
 Universal Reach at Birth: Family Connects

with only a small subgroup of families in a help with their family-specific needs. The
community. visits are flexibly timed to avoid disrupting
community standards of care, such as well-
The program rests on three pillars. The first baby visit schedules. The nurse also aims to
pillar, home visiting, is a system to reach accommodate family situations, for example,
all families giving birth in a community, by delaying a visit because of extended stays
typically at the birthing hospital. A trained by relatives or by speeding up a visit to attend
public health nurse (or other program to urgent needs. The nurse promises to bring
representative) welcomes the baby into the “goodies,” such as free diapers, to make the
community. The nurses invite themselves visits more attractive.
to one or more home visits (when the baby
is about three weeks old) so that they can Ideally, both the mother and her partner
identify needs and connect the family with (usually the father) are present during the
community resources. The second pillar, visit, although the nurse takes the mother’s
community alignment, is an assembly of all lead in deciding whether the partner’s
community resources available to families at presence is appropriate. During one pilot
birth, including child care agencies, mental of Family Connects, the nurses assumed
health providers, government social services, that partners would participate and went to
and programs for subgroups of families, such great lengths to invite them. But this strategy
as Healthy Families, Parents as Teachers, sometimes led the mother to withdraw,
and Early Head Start. The third pillar, data presumably because of conflict between
and monitoring, is an electronic data system the parents or difficulty in scheduling a
that acts as a family-specific psychosocial and convenient time. Because of this, the protocol
educational record (much like an electronic was changed: now the nurse listens to the
health record) to document nurses’ mothers’ advice on whether to include their
assessments of mother and infant, as well as partners, who participate about half the time.
connections with community agencies.
The initial home visit typically lasts between
Nurse Home Visits 90 and 120 minutes. The nurse is trained
to conduct a structured clinical interview
Nurses or program representatives greet that includes several hundred scored items
the mother at the birthing hospital to and covers a diverse set of topics in a
congratulate the family and to welcome conversational tone. The topics aren’t covered
the baby into the community. They deliver in a preset order; typically, the nurse follows
the message that the community wants the mother’s interests. The oral interview
to partner with the parents to support responses are supplemented by standardized
their child’s long-term success in health, screening for particularly sensitive or high-
education, and wellbeing. They also tell the risk circumstances, such as substance abuse.
parents that research shows every parent The nurse assesses risk in 12 key domains
can be successful but that at the same time, (see table 1) that predict adverse outcomes
every parent can benefit from support. In among children. Consistent with an ecological
that spirit, the nurse would like to visit the approach to health and wellbeing, these
parents in their home when the infant is domains encompass not only the needs of
about three weeks old to understand and individual family members, but also the

VOL. 29 / NO. 1 / S PR ING 2019 47

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman

family’s needs within its environment.11 The and together they develop a course of action
domains include child characteristics, such based on the needs identified. The plan
as temperament or health risk, which may may include follow-up home visits, phone
make an infant more challenging to care calls, or contact with external agencies.
for; family characteristics, such as parents’ This approach encourages parent buy-in
substance use, parent-child relationship and protects them from stigma in several
quality, or household safety; and community important ways. First, the course of action
characteristics, such as neighborhood is grounded in the needs identified by the
violence or access to resources. parents and nurse during the home visit,
rather than presumed needs based on family
Risk in each domain is scored quantitatively demographic characteristics. The experience
on a simple four-point scale: 1 indicates no is similar to, say, having high blood pressure
risk; 2 indicates risk that can be resolved identified during a routine physical and
through a brief intervention by a nurse (for working with a doctor on a course of action
example, if the parent has no knowledge to treat the problem. Second, because the
about how to select out-of-home childcare, plan is collaborative rather than directive,
the nurse educates her and helps her identify the parents are active participants in
childcare plans); 3 indicates significant determining what’s best for their family.
ongoing risk that requires a connection with
a community resource to resolve, such as a After developing a course of action and
mother’s substance abuse problem requiring gaining parental consent, the nurse
professional treatment; and 4 (used in communicates in writing with the infant’s
less than 1 percent of cases) indicates an pediatrician and the mother’s primary care
emergency requiring crisis intervention (for provider to create a bridge to ongoing care
example, imminent risk to the infant’s health after the nurse’s work ends. Four weeks
or infant maltreatment). after closing the case, the program makes a
follow-up phone call to check on the family’s
At the end of the interview, the nurse progress and to determine whether referrals
summarizes the findings with the parents, to community agencies were successful.

Table 1. 12 Risk Factors Assessed and Scored in Family Connects Nurse Home Visits

Domain Risk Factor

Parenting and child care Child care plans
Parent-infant relationship
Management of infant crying
Family violence and safety Material supports
Family violence
Mother’s past experience of being maltreated
Parent mental health and wellbeing Depression and anxiety
Substance abuse
Social and emotional support from others
Health care Parent health
Infant health
Health care plan, medical home, and insurance

48 T H E F U T UR E OF C HI L DRE N

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
 Universal Reach at Birth: Family Connects

When families report an unsuccessful leaders and agency directors who guide
community connection, Family Connects program direction, and a second board
staff members either make another attempt consists of frontline agency workers who
to get the family and the agency to connect work to solve problems, such as long wait lists
(if desired) or to help the family access and misunderstandings, as they arise.
alternative community resources. If an
agency has high rates of missed connections, Integrated Data System
program staff also work with it directly to
Family Connects staff document all their
reduce systemic barriers.
work in a family case record created from
Alignment with Community Agencies birth reports. The record begins with
attempts to schedule a home visit; it also
To make efficient referrals, nurses need to includes the nurse’s clinical interview,
understand the array of community agencies screening instrument responses, scoring of
that serve families with young children. The risk, referrals made to community agencies,
second pillar of the program is alignment interventions received at the agencies, and
with these agencies. A Family Connects staff parents’ “consumer satisfaction” responses to
member reaches out to as many agencies follow-up calls.
as possible to recruit their participation; to
document agency goals, service capacity, The case record serves three important
eligibility criteria, fees, wait-list time, and functions. First, it’s a key clinical tool that
evidence of effectiveness; and to assemble integrates information from many sources
agencies into an annotated electronic to guide intervention decision-making. The
directory for the nurses to use. In Durham, family-specific case record helps the nurse
NC, for example, the program’s directory match family needs with known evidence-
includes over 400 agencies, ranging from based interventions in the community.
childcare facilities and volunteer parent It incorporates information from the
groups to professional intervention services. agency, such as whether the family entered
Staff prioritizes identifying evidence- intervention and made progress. Much of this
based programs. And although in many information is communicated to the infant’s
communities the array of agencies includes a pediatrician at the end of the program.
wide variety of government and professional However, all information sharing requires
services, as well as nonprofits, identifying the parents’ consent, so that they remain in
informal resources, such as faith-based giving control.
closets or food pantries, is equally important,
especially in underfunded and rural Second, the case records are scrubbed of
communities. In communities with fewer identifying information and aggregated to
formal resources, or communities in which provide a summary about each community
formal services lack high quality, informal agency. If the case records include almost
resources may be a critical source of support all births in a community, these agency
for families with young children. records will be fairly comprehensive. They
can be used to document agency service
Coordination with community agencies to families, the quality of that service, and
is helped along by one (or more) advisory parents’ satisfaction with the agency. Family
boards. One board consists of community Connects uses the aggregated information

VOL. 29 / NO. 1 / S PR ING 2019 49

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman

to communicate with agencies about their practice home visits that are observed by
performance and to solve problems like a supervisor. They are tested on how well
chronic long waiting lists or excessive they adhere to the way questions are posed
dropouts. to parents and how reliably they score risk
in each of the 12 domains. Finally, they
Third, the family case records are aggregated receive certification that they’re ready
to create population-level indicators of to implement the program. Nurses from
family needs, and the agency records are remote areas travel to the national site, in
aggregated to map community assets and Durham, NC, for several days of training.
resources. The collective family-needs Back in their home communities, an onsite
information is then matched with the nurse supervisor works with a national site
aggregated community resources to identify overseer.
gaps in a community’s ability to serve all its
families. For example, in one community Nurse supervisors conduct quarterly fidelity
the case records indicated that 11 percent of checks with all home visiting nurses at their
all birthing mothers had a substance abuse site, in which the supervisor accompanies
problem meriting external intervention. nurses to parents’ homes and evaluates
But the asset map indicated that the adherence to 62 components of the visit.
community had resources to serve only 3 These items include properly administering
percent of mothers. The program passed this health assessments for the mother and baby;
information on to county commissioners, completing all assessment queries across
who were able to fund additional substance the 12 domains of family risks and needs;
abuse treatment for mothers who needed to teaching about infant care and safety; and
be home with a newborn. developing of a course of action with the
family. Supervisors also evaluate family risk
The aggregated information also provides in each of the 12 domains and compare their
population-level indicators of family ratings with those of the home visiting nurse
functioning at birth and can be used to to ensure consistency. After initial scoring,
track community progress across cohorts all records are stored electronically at the
of parents and children. In this way, the national office in Durham so that rates of
integrated data system not only serves the adherence to the model can be tracked over
needs of individual families and agencies but time both within and across program sites.
also advances public health.

Training, Supervision, and Fidelity Nurse supervisors conduct

The Family Connects program requires quarterly checks with
training of each staff member, certification of all home visiting nurses,
nurses in the protocol, ongoing supervision,
and documentation that implementation
accompanying the nurses
adheres to the program model (known to parents’ homes and
as fidelity). After qualified nurses are evaluating adherence to 62
hired, their training begins with reading
detailed manuals. Then they observe expert
components of the visit.
nurses during home visits and conduct

50 T H E F U T UR E OF C HI L DRE N

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
 Universal Reach at Birth: Family Connects

Family Connects Evaluation We were also interested in impact, which

was determined by comparing outcomes
To be eligible to receive funds from the for intervention families to those for control
Maternal, Infant, and Early Childhood families in the domains of connectedness
Home Visiting (MIECHV) program, or to the community, parenting quality, parent
to become known as “evidence-based,” a mental health, infant maltreatment, and use
program must undergo a rigorous RCT to of emergency health care.
demonstrate its impact when implemented
in a research setting, and we report findings First RCT
from two such community-level RCTs. Less
commonly completed but just as important The first RCT included all 4,777 resident
is an evaluation of how well a program is children born at the two hospitals in Durham
implemented and how much of an impact it County, NC, between July 1, 2009, and
has when it’s scaled up and brought to new December 31, 2010.12 About 40 percent
communities. We offer such a field evaluation were European American, 37 percent were
of Family Connects. African American, and 23 percent belonged
to other groups or were multiracial. Twenty-
In these three trials, we were interested six percent reported Hispanic ethnicity, 62
first in whether the program could be percent received Medicaid or had no health
implemented at scale with high quality. insurance, and 56 percent were not married.
Indicators of quality included:
Families of babies born on even dates were
• penetration rates, that is, the proportion assigned to receive the intervention and
of all families giving birth that were those born on odd dates were assigned to be
recruited into and successfully controls. The evaluation was based on intent-
completed the program; to-treat status, meaning that nonparticipants
were included and that recruitment into the
• fidelity of implementation, defined
intervention was part of the intervention
by the portion of families for whom
itself. The intervention began at birth and
quantitative scores were completed for
almost always ended by the time the child
nurses’ adherence to the protocol (as
was 12 weeks old. For a subset of analyses,
a percentage of all possible items) and
interviews and in-home observations of
the reliability of nurses’ scoring of risk
factors (computed as chance-corrected parenting and the home were completed
agreement with an independent when the infant was about six months old, at
supervisor’s scoring of the same least three months after the Family Connects
protocol), and; program had been completed. To contain
costs, a randomly selected, representative
• connection rates, defined as the portion sample of 664 intervention and control
of all families for whom need was families were recruited for data collection;
identified, the portion of all families for 549, or 80 percent, agreed to participate.
which an external agency connection In addition, administrative records were
was proposed, and then the portion retrieved from emergency departments
for whom a successful connection was for counts of emergency room visits and
established. overnight stays in a hospital, as well as

VOL. 29 / NO. 1 / S PR ING 2019 51

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

Table 2. Implementation and Impact Findings across Three Trials

RCT I RCT II Field
Population All resident county All resident county All resident county
births at two births at one Durham births in four low-
Durham County, NC hospitals County, NC hospital income, rural
(July 2009– (Jan. 2014– counties in eastern
Dec. 2010) June 2014) NC
(Feb. 2014–
Dec. 2015)

FC Implementation sample 2,327 456 994

% minority 70% 72% 67%
% Medicaid or no health insurance 61% 68% 69%
Penetration rates
Initial participation rate 80% 76% 84%
Of those initiating, completion rate 86% 84% 77%
Program completion rate 69% 64% 64%
Fidelity
Adherence to protocol 84% 90% 87%
Agreement in assessing risk factors k=.69 k=.75 k=.78
Connection Rates
% of families with at least one referral 45% 42% 54%
% of referred families for initiating service 79% 83% 94%
Impact
Connectedness
Number of community connections I < C** I > C*** I>C
Parenting and parent mental health
Number of positive parenting behaviors I>C
Father-infant relationship quality I = C**** I=C I>C
Maternal anxiety/depression rate I<C I<C I=C
Infant health and wellbeing
Number of emergency care episodes
Birth to 6 months, records I<C I=C Not assessed
Birth to 6 months, mother’s reports I<C I=C I<C
6–12 months, records I<C I=C Not assessed
Birth to 24 months, records I<C I>C Not assessed
Child protective services investigations I<C Not assessed Not assessed

*Cohen’s kappa coefficient is a statistic that measures agreement between raters after accounting for the possibility of
agreement by chance. Kappa values greater that 0.60 are considered to be substantial.
** I > C: Intervention group (I) average is greater than the control group (C) average. There is a greater than 95% likelihood
that this difference is not due to random chance or error.
*** I < C: Intervention group (I) average is less than the control group (C) average. There is a greater than 95% likelihood
that this difference is not due to random chance or error.
**** I = C: Intervention group (I) average is equivalent to the control group (C) average.

from Child Protective Services for counts date, 1,863 (80 percent) were successfully
of child maltreatment investigations and recruited into the program, which required
substantiations. that they listen to the goals and framing
of the program, interact with a Family
Evaluation of Implementation
Connects staff member (usually at the
Penetration rates. As table 2 shows, of the birthing hospital), and schedule an initial
2,327 families that gave birth on an even home visit. Of the recruited families, 86

52 T H E F U T UR E OF C HI L DRE N VOL. 29 / NO. 1 / S PR ING 2019 44

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

percent successfully completed the program, consultation or education about how to find
including receiving referrals to community high-quality child care); and 44 percent had
resources when appropriate, yielding a serious needs requiring connection with
population-wide full completion rate of 69 a community resource, such as substance
percent. abuse treatment, depression treatment,
or social services. One month after the
Fidelity of implementation. Adherence nurse terminated her involvement with the
to the protocol manual is essential for family, Family Connects staff telephoned
program quality. In this trial, an independent each family to find out whether they had
quality-control expert accompanied nurses successfully made a community resource
on 116 randomly selected home visits connection. Of the families referred to a
and independently documented whether community resource, 79 percent reported
a nurse correctly completed each of 62 they had indeed followed through to initiate
model elements. Overall, nurses adhered the connection.
to 84 percent of all elements, which is
considered high. The quality-control nurse Satisfaction with Family Connects. During
also independently scored each of the 12 the phone call, family members were asked
risk factors to ensure consistent family risk whether they would recommend Family
assessment. Because each of the 12 factors Connects to another new mother, and 99
was rated on a scale with only four possible percent said yes.
entries (1 through 4), the expert and the
Evaluation of Impact
home visiting nurse might have had a high
rate of agreement on factors score by chance, To evaluate the impact of Family Connects,
as they would be expected to agree some of interviewers visited a representative subset
the time simply by guessing. To overcome of intervention and control families when
this problem, the nurses’ reliability in the infant was six months old. To avoid
assessing risk was evaluated using Cohen’s potential bias, interviewers weren’t told
kappa coefficient, a statistic that measures which families had been eligible to receive
agreement between raters after accounting Family Connects, and participating families
for the possibility of agreement by chance.13 weren’t told that the primary study goal was
Kappa values greater than 0.60 are to evaluate Family Connects. Consent was
considered to be substantial. For the current obtained to access administrative records,
trial, chance-corrected agreement across the and hospital and Child Protective Services
116 observed home visits was 0.69. records were accessed then and later.
Findings are summarized in table 2.
Risk and connection rates. Ninety-four
percent of all families were scored as having Connectedness. Six months after the birth,
at least one need that merited intervention. intervention mothers reported 16 percent
In 1 percent of families, the need was a crisis more community connections than did
requiring immediate emergency intervention; control mothers.
49 percent had modest to moderate needs
that the nurse could resolve through Parenting and parent mental health.
additional home visits and brief counseling Intervention mothers reported more
(for example, through breastfeeding positive parenting behaviors than did control

45 T H E F UT UR E OF CHI LDRE N VOL. 29 / NO. 1 / S PR ING 2019 53

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

mothers, and a higher-quality father-infant might have had no impact on some groups
relationship. Screening indicated that of families. Instead, the additional analyses,
intervention mothers were 28 percent less which focused on emergency care episodes
likely than control mothers to exhibit signs of through age 12 months, found positive
clinical anxiety. impact for every group studied. Though
both groups benefited, infants with one or
Infant health and wellbeing. Compared
more birth risks had stronger intervention
to control mothers, intervention mothers
impact, defined as the difference between
reported 35 percent fewer serious injuries
intervention and control infants, than did
or illnesses among their infants that
required emergency department care or infants with no birth risks. Similarly, infants
hospitalizations. Administrative records from with Medicaid or no insurance experienced
the two community hospitals indicated that stronger intervention impact than infants
from birth to age six months, intervention with private insurance. The program’s impact
infants had 59 percent fewer emergency was stronger for majority than for minority
medical episodes than did control infants. families, and stronger for boys than for girls,
Between six and 12 months, intervention but in both cases significantly positive for
infants had 31 percent fewer such episodes. each group.

Follow-up analyses examined whether Recent analyses have explored the program’s
Family Connects had a positive impact for impact on children’s involvement in the
various subgroups. Even if it had a favorable Child Protective Services system over the
impact on the population as a whole, it children’s first five years.14 After accounting

Figure1.2.Child
Figure ChildProtective
ProtectiveServices
Services Investigations
Investigations through
through Age
Age 60
60 Months:
Months:RCT
RCT I.I.

0.5
0.5

0.45
0.45

0.4
0.4
Investigations
Mean Investigations

0.35
0.35

0.3
0.3

0.25
0.25
Cumulative Mean

Control
Control Mean
Mean
0.2
0.2
Family
Family Connects
ConnectsMean
Mean
Cumulative

0.15
0.15

0.1
0.1

0.05
0.05

0
0
0 6 12 18
18 24 30
30 36 42
42 48
48 54
54 60
60

Child
Child Age
Age in
in Months
Months

54 T H E F U T UR E OF C HI L DRE N VOL. 29 / NO. 1 / S PR ING 2019 46

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

for demographic risk factors, including they show that the program reached a large
birth complications, Medicaid status, percentage of families across trials and time.
minority status, and single parent status,
results indicate a 39 percent reduction in Fidelity of implementation. Nurses adhered
the rate of total child protective service to 90 percent of all elements, considered very
investigations for suspected child abuse or high. And when the quality control nurse
neglect (see figure 1). scored each of the 12 risk factors, chance-
corrected agreement as assessed by Cohen’s
Second RCT kappa was found to be high at 0.75.

Often, early-years intervention programs Risk and connection rates. Ninety-five

initially show positive impacts that aren’t percent of all families had at least one family
replicated when a second independent trial need meriting intervention, according
is conducted. We initiated a second RCT in to nurses’ scores. One percent required
Durham in 2014.15 Because one of the two emergency intervention, 52 percent had
original hospitals changed its policies and modest to moderate needs that could be
now prohibits patient contact with non- resolved by the nurse, and 42 percent had
hospital staff, births at this hospital were serious needs requiring connection with a
excluded from the new trial. As a result, community resource. In follow-up phone
the trial was conducted with all resident calls, 83 percent of the families referred to a
births at one hospital from January to June community resource reported that they had
of 2014. This time, the 456 families giving initiated the connection.
birth on an odd date were assigned to the
Family Connects intervention, and the Evaluation of Impact
479 families giving birth on an even date
were assigned as controls. Interviewers Connectedness. As in the first trial,
completed outcome assessments with intervention mothers reported more
a subsample when the children were community connections than did control
four to eight months old. Again, to avoid mothers—in this case, 17 percent more.
potential bias, interviewers weren’t told
Parenting and parent mental health. Unlike
which families were eligible to receive
in the first trial, in which intervention
Family Connects, and participating families
mothers fared better, in the second trial
weren’t told that the primary study goal was
intervention and control mothers had similar
to evaluate Family Connects.
levels of positive parenting behaviors and
Evaluation of Implementation father-infant relationship quality. As in the
first trial, intervention mothers were less
Penetration rates. As table 2 shows, of the likely than control mothers to exhibit signs of
456 families who gave birth on an odd clinical anxiety—in this case, 20 percent less
date, 349 (77 percent) were successfully likely.
recruited into the program; 84 percent of
these successfully completed the program, Infant health and wellbeing. Unlike the first
yielding a population-wide completion rate trial, in which intervention infants fared
of 64 percent. These figures are slightly better, in the second trial intervention and
lower than those for the first RCT, but control infants had similar rates of serious,

47 T H E F UT UR E OF CHI LDRE N VOL. 29 / NO. 1 / S PR ING 2019 55

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

emergency medical care episodes between from September 1, 2014, to December 31,
birth and six months. In the first trial, 2015. To reduce potential participation and
the intervention group had a mean of 1.5 response bias, all intervention group families
episodes per family by 24 months of age, participating in the impact evaluation were
and the control group had 2.4. In the second recruited without regard for their participation
trial, the intervention group’s mean was 1.1, status. The evaluation was also double-blind:
which was lower than the mean in the first families didn’t know that the survey’s primary
trial. Yet the control group mean was lower goal was to examine how Family Connects
still, at 0.9 episodes. We have no explanation affected child and family wellbeing, and
for the precipitous drop in these episodes interviewers didn’t know which families
among the control group. Involvement with actually had completed the Family Connects
child protective services hasn’t yet been program. Because we were comparing families
evaluated for the second RCT. of infants born in different time periods,
we had to consider the possibility that time-
Field Trial
related factors, such as the state of the global
The first two RCTs were conducted in economy, could account for any differences
Durham, NC, where the program was that would otherwise be attributed to Family
developed. It’s plausible that implementation Connects.
quality could be higher at this site than in
other places, and that its impact elsewhere The four rural counties in the trial had
could be lower.16 We sought to complete relatively few institutional community
a rigorous evaluation of the program’s resources upon which nurses could draw.
implementation and impact when it was However, the community alignment
brought to new sites and implemented by organization phase of the implementation
local staff members.17 revealed many informal resources, such as
an intervention program administered by the
After winning an Early Learning Challenge Rotary Club, social groups at the Veterans of
Grant from the US Department of Foreign Wars, and church-related support.
Education’s Race to the Top program, North Because nurses had to drive great distances to
Carolina allocated funds to Family Connects. some families’ homes, we anticipated that the
Beginning in 2014, Family Connects was program would have a lower penetration rate.
introduced in four low-income, rural Thus, the field trial presented an important
counties in northeast North Carolina opportunity to advance public health, public
(Beaufort, Bertie, Chowan, and Hyde). policy, and early childhood home visiting by
Conducting an RCT didn’t meet the grant’s evaluating the dissemination of a low-cost,
goal of delivering services to all children and universal home visiting program to rural
families in the four counties, so we evaluated communities characterized by very low
program impact through what’s called a resources and chronic poverty.
natural comparison design. Specifically, we
compared outcomes for families of infants Evaluation of Implementation
born from February 1, 2014, through July
31, 2014—before Family Connects came to Penetration rates. As table 2 shows, of the
their county—with outcomes for families 994 families with a child born during the
of infants born during implementation, implementation period, 770, or 77 percent,

56 T H E F U T UR E OF C HI L DRE N VOL. 29 / NO. 1 / S PR ING 2019 48

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

were successfully recruited into the program. intervention families. No significant

Of these families, 83 percent successfully differences were observed for possible
completed the program through the step clinical anxiety among mothers, although
of making connections with community intervention mothers reported 18 percent
resources, yielding a population-wide full lower levels of possible clinical depression.
completion rate of 64 percent. These figures
Infant health and wellbeing. The Family
are similar to those for the first two RCTs,
Connects intervention had consistent,
again showing a high penetration rate across
positive impacts on whether infants received
trials and time.
emergency medical care. Specifically,
Fidelity of implementation. Nurses adhered intervention mothers reported that between
to 87 percent of all elements, and chance- birth and infant age six months, they had
corrected agreement in scoring risk using sought emergency medical care for their
Cohen’s kappa was high at 0.78. infants 25 percent less often.

Risk and connection rates. Nurses scored Dissemination, Innovation, and

99.5 percent of all families as having at least Policy Engagement
one need meriting intervention. Less than
Family Connects recently established a
one percent of families required emergency
national office to support three missions:
intervention, 45 percent had modest to
broad dissemination, research and
moderate needs that could be resolved by
innovation, and policy engagement.
the nurse, and 54 percent had serious needs
requiring connection with a community Dissemination Challenges
resource. In follow-up calls, 83 percent of the
families referred to a community resource The Family Connects model is now
reported that they had initiated a connection. being disseminated in over two dozen
communities across the United States. These
Evaluation of Impact communities range from midsize cities to
small rural communities. Almost all of them
The comparison group of 343 infants born
requested that Family Connects come to
from February 1, 2014, to June 30, 2014,
their community, rather than responding to
was close in demographic and community
marketing. Recently, though, the program
characteristics to the Family Connects group
has adopted a new strategy of reaching
born from September 1, 2014, to December
31, 2015 out to selected communities to help them
consolidate community support for Family
Connectedness. As in the first two RCTs, Connects and find sustainable funding.
intervention mothers reported greater
community connectedness than did control
mothers. A major challenge of most of
Parenting and parent mental health.
these communities is finding
Intervention and control-group mothers a way to sustain funding over
reported similar levels of positive parenting
behaviors, but father-infant relationship
time.
quality was significantly higher for

49 T H E F UT UR E OF CHI LDRE N VOL. 29 / NO. 1 / S PR ING 2019 57

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

Communities have found financing for the Continued evaluation faces challenges,
Family Connects program in many ways. however. First, communities often lack
Sources include local public funding, state funding to conduct rigorous evaluation.
grants, private philanthropy, for-profit health Second, their willingness to do so may also
care organizations, Medicaid reimbursement, suffer due to difficulties with partners or
and federal funding awarded to states funders, or the fear that negative findings
through MIECHV. Most communities use could result in loss of funds for services that
funding from multiple sources. they believe are helping families in their
community. Ultimately, we need the political
A major challenge for most of these will to increase funding to support continued
communities is finding a way to sustain implementation of additional program
funding over time. Philanthropic funding, replication trials, to require evaluation
especially, is typically awarded in the short as a condition of receiving funds for new
term or year to year, making long-term programs, and to establish evaluation as
planning difficult and slowing the process a tool for continued learning and quality
of getting community agencies to buy in improvement.
to the communal effort. One strategy is to
take advantage of numerous federal funding Some innovation plans derive from findings
mechanisms, such as the new Family First from the three trials already conducted.
Prevention Services Act. The long-term For example, positive impact on fathers was
answer, though, may be policy change at found in two trials but not the third; thus one
the federal and state levels to make funding innovation will be to focus more on fathers
for early childhood programs as much of a and evaluate what happens rigorously.
priority as it is for later childhood and elder
care. At several sites, innovation is related to
topics that a community wants to prioritize.
Research and Innovation For example, one community received
funding as part of a broader effort to
The Family Connect program’s second improve trauma-informed services, that is,
mission is research and development. services for children and families that have
All dissemination sites are required to experienced various forms of trauma; in this
evaluate the implementation as part of community, a module of enhanced training
initial certification and ongoing monitoring, of staff members, assessment of trauma, and
and plans are under way to aggregate data intervention is being planned. This additional
across sites to understand natural patterns in component will be layered on top of Family
implementation quality over time and across Connects so that implementation isn’t
sites. Coupled with this is a new study of compromised. Other topical modules may
implementation cost that could help explain target nutrition and early literacy.
variation in quality. Each new site presents an
important opportunity for impact evaluation Another innovation—conducted
and continued learning. Rigorously designed collaboratively with other nationally known
impact evaluations won’t be possible at all home-visiting models—is an effort to apply a
sites, but alternate designs can be used, and similar rationale for universal screening in the
several sites could offer opportunities for new prenatal and postnatal periods. And one more
RCTs. goal is to understand how Family Connects

58 T H E F U T UR E OF C HI L DRE N VOL. 29 / NO. 1 / S PR ING 2019 50

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

can collaborate more closely with pediatric to help communities make such programs
care. financially sustainable.

Policy Engagement Conclusions

The concept of universal reach at a The story of Family Connects offers promise
critical point in the lifespan, coupled with for population-level impact on the prevention
assessment and individualized connection of child abuse and neglect. But the history of
with community resources to support a young innovation and rigorous evaluation in this field
child’s healthy behavioral development and to shows a mixed pattern: some implementations
prevent child maltreatment, has a long way to are positive, and some are disappointing;
go before it becomes routine for every family some evaluations show positive impacts, and
in the United States. To push this movement some show none. To make true headway,
forward, program developers and researchers evaluators and program developers need to
will need to work with policy makers at the acknowledge the full range of findings about
local, state, and federal levels. They need to a program’s implementation and impact,
understand what drives policy so that they and they must embrace continued research
can frame their case in a way that reaches rather than cutting it off after favorable
attentive ears. They will need to conduct findings emerge. At the same time, funders
benefit-cost analyses and analyses of financing of research and implementation in the public,
models, and they will need to keep providing philanthropic, and private sectors should
rigorous empirical evidence about program exercise patience, thus easing any pressure to
implementation and impact. Finally, they will produce favorable findings. With a committed
need to understand how state and federal partnership between programs and funders,
funding streams, such as Medicaid, MIECHV, we are optimistic that we’re not far from
and the Title V maternal and child health achieving population impact in reducing child
services block grant program, may be used abuse and neglect.

51 T H E F UT UR E OF CHI LDRE N VOL. 29 / NO. 1 / S PR ING 2019 59

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms
Kenneth A. Dodge and W. Benjamin Goodman Universal Reach at Birth: Family Connects

Endnotes
1. Ajay Chaudry et al., Cradle to Kindergarten: A New Plan to Combat Inequality (New York: Russell
Sage, 2017); Alan E. Kazdin and Stacy L. Blase, “Rebooting Psychotherapy Research and Practice to
Reduce the Burden of Mental Illness,” Perspectives on Psychological Science 6 (2011): 21–37, https://doi.
org/10.1177/1745691610393527.
2. David L. Olds et al., “Long-Term Effects of Nurse Home Visitation on Children’s Criminal and Antisocial
Behavior: 15-Year Follow-Up of a Randomized Trial,” Journal of the American Medical Association 280
(1998): 1238–44, https://doi.org/10.1001/jama.280.14.1238; John M. Love et al., “The Effectiveness
of Early Head Start for 3-Year-Old Children and Their Parents: Lessons for Policy and Programs,”
Developmental Psychology 41 (2005): 885–901, https://doi.org/10.1037/0012-1649.41.6.88.
3. Karen Appleyard et al., “Preventing Early Child Maltreatment: Implications from a Longitudinal Study of
Maternal Abuse History, Substance Use Problems, and Offspring Victimization,” Prevention Science 12
(2011): 139–49, https://doi.org/10.1007/s11121-010-0193-2.
4. Geoffrey Rose, “Strategy of Prevention: Lessons from Cardiovascular Disease,” British Medical Journal
282 (1981): 1847–51.
5. Brandon C. Welsh, Christopher J. Sullivan, and David L. Olds, “When Early Crime Prevention Goes
to Scale: A New Look at the Evidence,” Prevention Science 11 (2010): 115–25, https://doi.org/10.1007/
s11121-009-0159-4.
6. Chaudry et al., Cradle to Kindergarten.
7. Deborah Daro and Kenneth A. Dodge, “Creating Community Responsibility for Child Protection:
Expanding Partnerships, Changing Context,” Future of Children 19, no. 2 (2009): 67–93.
8. Appleyard et al., “Preventing Early Child Maltreatment.”
9. Daro and Dodge, “Creating Community Responsibility.”
10. Kenneth A. Dodge et al., “Toward Population Impact from Home Visiting,” Zero to Three 33 (2013):
17–23.
11. Urie Bronfenbrenner and Pamela A. Morris, “The Ecology of Developmental Process,” in Handbook of
Child Psychology, vol. 1, Theoretical Models of Human Development, 5th ed., ed. Richard M. Lerner
(Hoboken, NJ: Wiley, 1998): 993–1028.
12. Kenneth A. Dodge et al., “Randomized Controlled Trial of Universal Postnatal Nurse Home Visiting:
Impact on Emergency Care,” Pediatrics 132 (2013): S140–6, https://doi.org/10.1542/peds.2013-1021M;
Kenneth A. Dodge et al., “Implementation and Randomized Controlled Trial Evaluation of Universal
Postnatal Nurse Home Visiting,” American Journal of Public Health 104 (2014): S136–43, https://doi.
org/10.2105/AJPH.2013.301361.
13. Jacob Cohen, Statistical Power Analysis for the Behavioral Sciences, 2nd ed. (Hillsdale, NJ: Lawrence
Erlbaum Associates, 1988).
14. W. Benjamin Goodman et al., “Evaluation of Impact of Family Connects on Child Protective Services
Involvement through Age 5 Years,” unpublished manuscript, Duke University, 2018.
15. Kenneth A. Dodge et al., “Evaluation of Implementation and Impact of Family Connects: A Replication
Randomized Controlled Trial,” unpublished manuscript, Duke University, 2018.
16. Welsh, Sullivan, and Olds, “A New Look.”
17. W. Benjamin Goodman et al., “Evaluation of Family Connects Implementation and Impact in Rural
Communities,” unpublished manuscript, Duke University, 2018.

60 T H E F U T UR E OF C HI L DRE N VOL. 29 / NO. 1 / S PR ING 2019 52

This content downloaded from

137.189.166.12 on Tue, 23 Aug 2022 00:54:35 UTC
All use subject to https://about.jstor.org/terms