Selected Findings from the Cross-Site Evaluation of the Federal
Healthy Start Program
Vonna Lou Caleb Drayton • Deborah Klein Walker •
Sarah W. Ball • Sara M. A. Donahue •
Rebecca V. Fink
Published online: 28 November 2014
� Springer Science+Business Media New York 2014
Abstract Initiated in 1991, the Federal Healthy Start
Program includes 105 community-based projects in 39
states, the District of Columbia and Puerto Rico. Healthy
Start projects work collaboratively with stakeholders to
ensure participants’ continuity of care during pregnancy
through 2 years postpartum. This evaluation of Healthy
Start projects examined relationships between implemen-
tation of nine core service and system program components
and improvements in birth and project outcomes. Program
components and outcomes were examined using data from
a 2010 Healthy Start project director (PD) survey
(N = 104 projects) and 2009 performance measure data
from the Maternal and Child Health Bureau Discretionary
Grant Information System (N = 98 projects). We explored
bivariate relationships between the nine core program
components and (a) intermediate and long-term project
outcomes and (b) birth outcomes. We assessed independent
associations of implementation of all core program com-
ponents with birth outcomes, adjusting for project charac-
teristics and activities. In 2010, 57 projects implemented all
nine core program components: 104 implemented all five
core service components and 69 implemented all four core
systems components. Implementation of all core program
components was significantly associated with several PD-
reported intermediate and long-term project outcomes, but
was not associated with singleton low birth weight or infant
mortality among participants’ infants. This evaluation
revealed a mixed set of relationships between Healthy Start
projects’ implementation of the core program components
and achievement of project outcomes. Although the find-
ings demonstrated a positive impact of Healthy Start pro-
jects on birth outcomes, only a few associations were
statistically significant.
Keywords Maternal and child health � Healthy Start
Program � Cross-site evaluation � Program evaluation
Introduction
The Federal Healthy Start Program began in 1991 as a
response to high infant mortality rates (IMR) in the United
States as well as the large gap in these rates between white
and non-white infants. The first Healthy Start projects were
funded as demonstration sites in 15 communities with IMR
1.5–2.5 times the national average. By 2012, the program
had expanded in size and mission to include 105 projects in
39 states, the District of Columbia and Puerto Rico,
including projects in both urban and rural areas. As spec-
ified by Health Resources and Services Administration
(HRSA) guidance documents [1–3] the core Program goals
include: (1) a reduction of racial and ethnic disparities in
access to and utilization of health serv.
A Critique of the Proposed National Education Policy Reform
Selected Findings from the Cross-Site Evaluation of the Federa.docx
1. Selected Findings from the Cross-Site Evaluation of the Federal
Healthy Start Program
Vonna Lou Caleb Drayton • Deborah Klein Walker •
Sarah W. Ball • Sara M. A. Donahue •
Rebecca V. Fink
Published online: 28 November 2014
� Springer Science+Business Media New York 2014
Abstract Initiated in 1991, the Federal Healthy Start
Program includes 105 community-based projects in 39
states, the District of Columbia and Puerto Rico. Healthy
Start projects work collaboratively with stakeholders to
ensure participants’ continuity of care during pregnancy
through 2 years postpartum. This evaluation of Healthy
Start projects examined relationships between implemen-
tation of nine core service and system program components
and improvements in birth and project outcomes. Program
components and outcomes were examined using data from
2. a 2010 Healthy Start project director (PD) survey
(N = 104 projects) and 2009 performance measure data
from the Maternal and Child Health Bureau Discretionary
Grant Information System (N = 98 projects). We explored
bivariate relationships between the nine core program
components and (a) intermediate and long-term project
outcomes and (b) birth outcomes. We assessed independent
associations of implementation of all core program com-
ponents with birth outcomes, adjusting for project charac-
teristics and activities. In 2010, 57 projects implemented all
nine core program components: 104 implemented all five
core service components and 69 implemented all four core
systems components. Implementation of all core program
components was significantly associated with several PD-
reported intermediate and long-term project outcomes, but
was not associated with singleton low birth weight or infant
mortality among participants’ infants. This evaluation
revealed a mixed set of relationships between Healthy Start
3. projects’ implementation of the core program components
and achievement of project outcomes. Although the find-
ings demonstrated a positive impact of Healthy Start pro-
jects on birth outcomes, only a few associations were
statistically significant.
Keywords Maternal and child health � Healthy Start
Program � Cross-site evaluation � Program evaluation
Introduction
The Federal Healthy Start Program began in 1991 as a
response to high infant mortality rates (IMR) in the United
States as well as the large gap in these rates between white
and non-white infants. The first Healthy Start projects were
funded as demonstration sites in 15 communities with IMR
1.5–2.5 times the national average. By 2012, the program
had expanded in size and mission to include 105 projects in
39 states, the District of Columbia and Puerto Rico,
including projects in both urban and rural areas. As spec-
ified by Health Resources and Services Administration
4. (HRSA) guidance documents [1–3] the core Program goals
include: (1) a reduction of racial and ethnic disparities in
access to and utilization of health services, (2) an improved
local health care system, and (3) an increased consumer or
community voice in health care decisions.
V. L. C. Drayton (&)
Booz Allen Hamilton, One Preserve Parkway, Rockville,
MD 20852, USA
e-mail: [email protected]
D. K. Walker � S. W. Ball � S. M. A. Donahue � R. V. Fink
Abt Associates, 55 Wheeler Street, Cambridge, MA 02138-
1168,
USA
e-mail: [email protected]
S. W. Ball
e-mail: [email protected]
S. M. A. Donahue
e-mail: [email protected]
R. V. Fink
e-mail: [email protected]
123
Matern Child Health J (2015) 19:1292–1305
5. DOI 10.1007/s10995-014-1635-4
http://crossmark.crossref.org/dialog/?doi=10.1007/s10995-014-
1635-4&domain=pdf
http://crossmark.crossref.org/dialog/?doi=10.1007/s10995-014-
1635-4&domain=pdf
The Federal Healthy Start Program focuses on improv-
ing the health and well-being of women, infants, children
and their families through the implementation of evidence-
based practices and innovative community interventions. In
2010, Healthy Start projects served almost 30,000 pregnant
women, many of whom were black or African American,
34 years and younger, with incomes below 100 percent of
the federal poverty level [4].
Healthy Start projects work collaboratively with com-
munity stakeholders and consumers to leverage existing
service and system resources so that women at risk for
adverse birth outcomes are assured continuity of care
during pregnancy through 2 years postpartum. Since 2001,
all Healthy Start projects have been required to implement
6. nine ‘‘core’’ program components: five service components
(outreach and recruitment, case management, health edu-
cation, interconception care (ICC), perinatal depression
screening) and four systems-building components (con-
sortia, local health systems action plan (LHSAP), coordi-
nation and collaboration with Title V, and a sustainability
plan). Healthy Start projects may also implement other
support services needed in their local communities, such as
breastfeeding support and education, screening for
domestic/intimate partner violence and child abuse, initia-
tives to improve family and/or male involvement, healthy
weight interventions, home visiting, and smoking cessation
[1–3].
National Evaluations of the Federal Healthy Start
Program
The Federal Healthy Start Program has been evaluated
from its inception in the early 1990 s. The first national
evaluation, conducted from 1997 through 1999, examined
7. the implementation of the 15 demonstration project activ-
ities during fiscal years 1992 and 1996 and assessed whe-
ther these projects achieved the Healthy Start Program
goals of reducing infant mortality and improving maternal
and infant health. The second national evaluation was
conducted in two phases from 2002 through 2007 and
sought to obtain information about the implementation of
program components and to identify program features
associated with improved perinatal outcomes. Findings
from this evaluation were summarized in a profile report
presenting the characteristics of all Healthy Start projects
[5] and in case studies that documented the context and
implementation of the Healthy Start Program in eight sites
[6]. The evaluation also collected information on program
implementation and outcomes through a participant survey
that was conducted in four sites [7]. The third national
evaluation is the cross-site evaluation summarized in this
article. It was conducted from 2009 through 2012 to
8. examine relationships between the core program
components and long-term program and birth outcomes, in
addition to factors that influence these relationships. The
primary objective of the evaluation was to assess the effect
of implementation of all nine core program components on
long-term maternal and child health outcomes.
Methods
The evaluation was guided by a logic model (Fig. 1) that
outlined the hypothesized relationships between Healthy
Start project context, implementation of core service and
system program components, and four long-term outcomes
relevant to the Healthy Start Program goals: (1) improved
birth outcomes, (2) improved maternal health, (3)
improved child health, and (4) sustained community
capacity to reduce disparities in health status in the target
community. A cross-sectional design was used to assess the
associations of implementation of the nine core program
components with (1) project characteristics, (2) achieve-
9. ment of intermediate project outcomes, (3) service and
system activities conducted by the Healthy Start project
that made a primary or major contribution to reducing
disparities in maternal and infant health outcomes, and (4)
achievement of long-term birth outcomes.
Data Sources
Self-reported data from the 2010 project director survey
(PD survey) and performance measure (PM) data for 2009
reported to the Maternal and Child Health Bureau (MCHB)
Discretionary Grant Information System (DGIS) were used
in all analyses. The 2010 PD survey was administered via
web to Healthy Start project staff between July and Sep-
tember 2011 and was completed for all 104 projects
(100 % response rate). The survey was designed to collect
information on implementation and features of the nine
core program components as well as additional support
services offered by each Healthy Start project and project
achievements. The DGIS is a Web-based system that
10. MCHB grantees use to report their data online to MCHB
through HRSA’s Electronic Handbook as a part of the
grant application and performance reporting processes; it is
the repository of PM data for all MCHB programs. During
the time period of this evaluation, the MCHB utilized 15
PMs to monitor the progress of all Healthy Start projects
towards the achievement of Program objectives. A list of
current MCHB Healthy Start Program PMs is available
from: https://mchdata.hrsa.gov/DGISReports/PerfMeasure/
default.aspx. Performance measure data for 2009 were
available for 98 projects. After a thorough examination of
the available PM data from the DGIS [8], four PMs (two
birth outcome PMs, 1 service outcome PM and 1 system
Matern Child Health J (2015) 19:1292–1305 1293
123
https://mchdata.hrsa.gov/DGISReports/PerfMeasure/default.asp
x
https://mchdata.hrsa.gov/DGISReports/PerfMeasure/default.asp
x
11. outcome PM; Table 1) were selected for this evaluation
based on the quality and consistency of data as well as the
relevance of the PM to the evaluation objectives. Project
characteristic data that were consistently reported in the
DGIS were also used in our analyses. State Title V birth
outcome PMs (singleton LBW and IMR) and Healthy
People 2010 and 2020 objective targets (LBW and IMR)
[9, 10] were used as benchmarks for comparison.
Measurement of Variables
Variable selection was informed by program components
and expected outcomes, the logic model, and previous
studies of birth outcomes [11]. The primary exposure of
interest was the implementation of all nine core program
components: outreach and recruitment, case management,
health education, interconception care (ICC), perinatal
depression screening, consortia, local health systems action
plan (LHSAP), coordination and collaboration with Title
V, and a sustainability plan. Implementation was deter-
12. mined using data from the 2010 PD survey (yes/no
response for each component). The birth outcomes of
interest were measured using two PMs reported in the
DGIS in 2009: percent singleton low birth weight (PM 51)
and infant mortality rate (PM 52).
We examined characteristics hypothesized to influence
the association of implementation of program components
with birth outcomes. We obtained information on these
characteristics from the 2010 PD survey and the DGIS.
Maternal demographic characteristics were not available
for this analysis. Project characteristics (Table 2) that were
examined were length of funding (initial project funding
received in Phase 1 [1991–1996], 2 [1997–2000], 3
[2001–2004], or 4 [2005–2010]), geographic location
(urban, not urban), and organization type (government
agency, community-based non–governmental agency or
other organization type). Project director report of
achievement of intermediate outcomes (eleven outcomes;
13. see Table 2) (yes/no), service and systems activities that
made a primary or major contribution to reducing dispar-
ities in maternal and infant health outcomes (fourteen
activities; see Table 2) (yes/no), and achievement of long-
term maternal and child health and community capacity
outcomes (five outcomes; see Table 2) (yes/no) were
examined in descriptive analyses and included as covari-
ates in multivariable analyses. One service outcome PM
Fig. 1 Logic model for the cross-site evaluation of Healthy Start
1294 Matern Child Health J (2015) 19:1292–1305
123
(PM 20, the percent of women participants who have an
ongoing source of primary and preventive care services for
women) and one system outcome PM (PM 22, a score
between 0 and 64 representing the degree to which the
project facilitated health providers’ screening of women
participants for eight risk factors) were examined in
14. descriptive analyses and included as covariates in multi-
variable analyses (see Table 1).
Analysis
We calculated descriptive statistics for all variables across
all Healthy Start projects. We then performed bivariate
analyses using Pearson’s Chi square test and Fisher’s exact
test to (1) describe implementation of the nine core Healthy
Start Program components by project characteristics; (2)
examine the association of implementation of all core
components with each of (a) intermediate outcomes,
(b) service and systems activities that made a primary or
major contribution to reducing disparities in maternal and
infant health outcomes, and (c) long-term maternal and
child health and community capacity outcomes; and (3)
examine the association of intermediate outcomes and
service and systems activities that made a primary or major
contribution to reducing disparities in maternal and infant
health outcomes with (a) long-term maternal and child
15. health and community capacity outcomes and with b) birth
outcome PMs. We also compared the birth outcome PM
rates among Healthy Start projects with their state’s Title V
Program rates and with achievement of national Healthy
People (HP) 2010 and 2020 objective targets.
We developed multivariable linear and logistic regres-
sion models to examine the independent associations of
implementation of all core program components with birth
outcomes, adjusting for project characteristics, project
director-reported intermediate outcomes, and service and
system PMs. We developed linear regression models to
examine continuous outcomes (singleton LBW, IMR) and
logistic regression models to examine achievement (yes/
no) of state Title V rates or national HP objectives. We
Table 1 MCHB performance measures (PM) used in multivariate
analyses
Category/
PM
Definition/elements Components/Scale
16. Birth outcomes
PM 51 Percent of live singleton births weighing less than 2,500
g Numerator: Number of live singleton births less than 2,500 g
in
the calendar year to program participants
Denominator: Live singleton births in the calendar year among
program participants
PM 52 The infant mortality rate per 1,000 live births
Numerator: Number of deaths to infants from birth through
364 days of age to program participants
Denominator: Number of live births in the calendar year among
program participants
Service outcomes
PM 20 The percent of women participating in MCHB supported
programs who have an ongoing source of primary and
preventive care services for women
Numerator: The number of women participating in MCHB-
funded projects who have an ongoing source of primary and
preventive care services during the reporting period
17. Denominator: The number of women participating in MCHB-
funded projects during the reporting period
Systems outcomes
PM 22 The degree to which MCHB supported programs
facilitate health
providers’ screening of women participants for risk factors
Total possible score: 0–64
Scoring instructions: Using a scale of 0-2, indicate the degree to
which your grant has performed each activity to facilitate
screening for each risk factor by health providers in your
program
Scale definitions:
0 = Grantee does not provide this function or assure that this
function is completed,
1 = Grantee sometimes provides or assures the provision of this
function but not on a consistent basis,
2 = Grantee regularly provides or assures the provision of this
function
Risk factors
18. 1. Smoking
2. Alcohol
3. Illicit drugs
4. Eating disorders
5. Depression
6. Hypertension
7. Diabetes
8. Domestic violence
A list of all current MCHB Healthy Start Program PMs is
available from:
https://mchdata.hrsa.gov/DGISReports/PerfMeasure/default.asp
x
Matern Child Health J (2015) 19:1292–1305 1295
123
https://mchdata.hrsa.gov/DGISReports/PerfMeasure/default.asp
x
calculated betas or odds ratios with 95 % confidence
intervals. Variables that were included in the models were
those found to be associated with the birth outcomes of
19. interest in previous studies or in the bivariate analyses as
well as any other characteristics of a priori interest
according to the evaluation logic model (Fig. 1). The
multivariate models to examine birth outcomes included
only those projects with PM data.
The model to examine the association of implementa-
tion of all core components with singleton LBW (PM 51)
included the following covariates: initial funding (Phase 1
versus all other phases), urban geographic location, not
urban geographic location, grantee organization type,
Healthy Start project facilitation of provider screening for
risk factors (PM 22, score greater than mean of all pro-
jects), percent of women participants with ongoing source
of primary and preventive care (PM 20), self-reported
improved birth spacing in 2010 (yes/no), self-reported
increased cultural competence of providers (yes/no), and
self-reported increased participant involvement in Healthy
Start decision-making (yes/no). The model to examine the
20. IMR outcome (PM 52) included many of the same covar-
iates, in addition to percent singleton LBW (PM 51), an
independent risk factor for infant mortality.
This evaluation was determined exempt from IRB
review by the Abt Associates Institutional Review Board
on September 1, 2010 (Abt IRB # 0499).
Results
Descriptive Characteristics
Table 2 presents the distribution of project characteristics
as well as project director-reported implementation of the
core components, intermediate project outcomes, service
and systems activities that made a primary or major con-
tribution to reducing disparities in maternal and infant
health outcomes, and long-term maternal and child health
and community capacity outcomes. All 104 Healthy Start
projects implemented all five core service components.
Over two-thirds of projects implemented the four core
systems-building components: 99 % implemented one or
21. more consortium, 91 % implemented a LHSAP, 87 %
collaborated with Title V, and 66 % had a sustainability
plan. Overall, 57 (55 %) projects implemented all nine core
program components; this group includes 10 of the 18
projects that were first funded during Phase 1 (1991-1996)
of the Healthy Start Program. Most projects were in
operation for at least 10 years at the time the PD survey
was administered; 17 % were first funded in Phase 1 and
61 % in Phase 2. Approximately 75 % of projects were
located in urban areas, including cities and metropolitan
areas; and 40 % of grantee organizations were state or local
government agencies.
Approximately two-thirds of all projects reported that in
2010 the project had accomplished a number of interme-
diate outcomes including increased awareness of the
importance of interconception care and of disparities in
birth outcomes as a community priority, increased positive
health behaviors among participants, increased access to
22. available services for participants, and increased number of
participants with a medical home.
More than two-thirds of all projects reported that case
management, enabling services such as transportation and
translation, and interconception care activities conducted
by the project made a primary or major contribution to
reducing disparities in maternal and infant health out-
comes. Less than two-thirds of projects reported that other
service and systems activities conducted by the project,
such as collaboration with consumers, community-based
organizations, and public and private agencies, made a
similar contribution to reducing disparities in maternal and
infant health outcomes.
Sixty-eight percent of project directors reported that the
project had achieved improvements in birth outcomes in
2010 and 39 % reported achieving improvements in
maternal health. Less than one-third of project directors
reported that the Healthy Start project had achieved sus-
23. tained capacity to reduce disparities in health status in the
community (32 % of projects); improvements in child
health (31 %); and increased birth spacing (19 %). A small
proportion (12 %) of project directors reported that the
Healthy Start project had not achieved any long-term out-
comes in 2010.
Bivariate Analyses: Core Program Components
Table 3 presents the results of bivariate analyses examin-
ing the relationship between implementation of the nine
core program components and project characteristics, as
well as relationships between implementation of program
components and three categories of project director-
reported outcomes: (1) intermediate outcomes, (2) activi-
ties that contributed to reducing disparities in maternal and
infant health outcomes, and (3) long-term maternal and
child health and community capacity outcomes. The 57
projects that implemented all core components were used
as the reference group. Only results that were statistically
24. significant (p B 0.05) are reported in the table.
Healthy Start projects whose grantee organizations were
state or local government agencies were significantly
(p B 0.05) less likely to implement all core components
compared with projects whose grantee organizations were
a community-based non-governmental organization or
other type of organization.
1296 Matern Child Health J (2015) 19:1292–1305
123
Table 2 Distribution of Healthy Start project characteristics and
project director-reported implementation of program
components,
intermediate outcomes, service and systems activities that
contributed
to reducing disparities in maternal and infant health outcomes,
and
long-term maternal and child health and community capacity
out-
comes, among all Healthy Start projects (N = 104 projects)
All
25. projects
(N = 104)
n (%)
Project characteristics
a
Length of funding
Initial Funding Phase 1 (1991–1996) 18 17
Initial Funding Phase 2 (1997–2000) 63 61
Initial Funding Phase 3 (2001–2004) 10 10
Initial Funding Phase 4 (2005–2010) 13 12
Geographic location: Urban [urban/central city,
metropolitan area (city and suburbs)]
Yes 78 75
No 26 25
Geographic location: Not urban (suburban, border US-
Mexico, rural)
Yes 28 27
No 76 73
26. HS grantee organization type
Government agency (state agency, community
government agency such as a local health
department)
42 40
Community-based non-governmental organization
(health care or non-health care) or Other
organization (including academic medical center,
non-profit organization, tribal organization,
Federally Qualified Health Center)
62 60
Implementation of all nine core program components
b
Yes 57 55
No 47 45
Intermediate outcomes
c
Increased awareness of the importance of interconception care
Yes 80 77
27. No 24 23
Increased awareness of disparities in birth outcomes as
community
priority
Yes 76 73
No 28 27
Increased positive health behaviors among our participants
Yes 74 71
No 30 29
Increased access to the services available for our participants
Yes 71 68
No 33 32
Increased number of participants with a medical home
Table 2 continued
All
projects
(N = 104)
n (%)
Yes 70 67
28. No 34 33
Increased screening for perinatal depression among providers in
the
community
Yes 51 49
No 53 51
Increased participant involvement in Healthy Start decision-
making
Yes 50 48
No 54 52
Increased integration of prenatal, primary care, and mental
health
services
Yes 47 45
No 57 55
Increased cultural competence of providers in our community
Yes 43 41
No 61 59
Increased participant involvement in other community activities
29. addressing systems change
Yes 39 37
No 65 63
Increased participant involvement in decision-making among
partner
agencies
Yes 22 21
No 82 79
Service and systems activities that contributed to reducing
disparities
in maternal and infant health outcomes
d
Case management
Yes 90 87
No 14 13
Enabling services
Yes 73 73
No 31 30
Interconception care
Yes 70 67
30. No 34 33
Perinatal depression screening
Yes 66 63
No 38 37
Outreach and client recruitment
Yes 64 62
No 40 39
Collaboration with consumers
Yes 60 58
No 44 42
Matern Child Health J (2015) 19:1292–1305 1297
123
Table 2 continued
All
projects
(N = 104)
n (%)
31. Collaboration with community-based organizations
Yes 53 51
No 51 49
Collaboration with public agencies
Yes 49 47
No 55 53
Collaboration with private agencies
Yes 46 44
No 58 56
Consortium
Yes 45 43
No 59 57
Local Health System Action Plan
Yes 43 41
No 61 59
Collaboration with local Title V
Yes 34 33
No 70 67
32. Collaboration with State Title V
Yes 31 30
No 73 70
Provider education
Yes 39 38
No 65 62
Long-term maternal and child health and community capacity
outcomes
e
Improved birth outcomes
Yes 71 68
No 33 32
Improved maternal health
Yes 41 39
No 63 61
Sustained community capacity to reduce disparities in health
status in
the community
Yes 33 32
33. No 71 68
Improved child health
Yes 32 31
No 72 69
Increased birth spacing
Yes 20 19
No 84 81
No long term outcomes were achieved in 2010
Yes 13 12
Table 2 continued
All
projects
(N = 104)
n (%)
No 91 88
a
Data source: Maternal and Child Health Bureau Discretionary
Grant
Information System
34. b
Data source: 2010 Project Director survey. To determine imple-
mentation of core service components, project directors were
asked,
‘‘Which of the following services does your Healthy Start
project
offer?’’ (response options: ‘‘Outreach and participant
recruitment,’’
‘‘Case management,’’ ‘‘Health education,’’ ‘‘Perinatal
depression
screening,’’ and ‘‘Interconceptional services’’). To determine
imple-
mentation of the core systems-building component of having a
con-
sortium, project directors were asked ‘‘Does your Healthy Start
project have at least one active consortium that addresses
maternal
and child health issues’’ (response options: Yes/No). To
determine
implementation of the core systems-building component of
having a
Local Health System Action Plan, project directors were asked
‘‘Does
35. your Healthy Start project have a Local Health System Action
Plan
(LHSAP)?’’ (response options: Yes/No; a follow up question
was
asked to determine if the LHSAP was specific to the Healthy
Start
project). To determine implementation of the core systems-
building
component of coordination and collaboration with Title V,
project
directors were asked to specify the types of collaborative
activities
that their Healthy Start project established with the State Title
V
agency. Projects were classified with a ‘‘yes’’ response if the
project
director indicated that the State Title V agency ‘‘is a member of
the
Healthy Start consortium,’’ ‘‘has a written memorandum of
under-
standing or agreement with Healthy Start,’’ ‘‘provides
contracted
services to Healthy Start,’’ ‘‘hosts out-stationed Healthy Start
staff,’’
36. ‘‘participates in joint training with Healthy Start,’’ ‘‘has a
shared
staffing arrangement with Healthy Start,’’ ‘‘coordinates case
man-
agement or is planning with Healthy Start for shared
participants,’’
‘‘shares protocols with Healthy Start,’’ ‘‘is involved in Healthy
Start
sustainability planning,’’ ‘‘has a data-sharing arrangement with
Healthy Start,’’ ‘‘contributes to pooled funding streams to
support
joint services,’’ ‘‘has a Healthy Start employee on their board,’’
‘‘works with Healthy Start to develop consistent health
messages for
participants,’’ and/or ‘‘receives cultural competence training
from
Healthy Start.’’ To determine implementation of the core
systems-
building component of having a sustainability plan, project
directors
were asked ‘‘Does your Healthy Start project have a
sustainability
plan, that is, a plan to maintain services to the target population
after
37. federal Healthy Start funding ends?’’ (response options:
Yes/No)
c
Data source: 2010 Project Director survey. Project directors
were
asked, ‘‘Which of the following intermediate outcomes did your
Healthy
Start project achieve in 2010?’’. Multiple responses were
allowed
d
Data source: 2010 Project Director survey. Project directors
were
asked, ‘‘To what extent did the following activities conducted
by your
Healthy Start project contribute to reducing disparities in
maternal
and infant health outcomes?’’. Response options included
Primary
contribution, Major contribution, Moderate contribution, Minor
con-
tribution, and No contribution or N/A. Primary contribution and
Major contribution were classified as ‘‘Yes.’’
e
Data source: 2010 Project Director survey. Project directors
38. were
asked, ‘‘Which of the following long term outcomes did your
Healthy
Start project achieve in 2010?’’. Multiple responses were
allowed
1298 Matern Child Health J (2015) 19:1292–1305
123
Table 3 Association of implementation of Healthy Start Program
components with project characteristics and project director-
reported inter-
mediate outcomes, service and systems activities that
contributed to reducing disparities in maternal and infant health
outcomes, and long-term
maternal and child health and community capacity outcomes (N
= 104 projects)
Implementation of all required
core program components
Yes
(n = 57)
No
(n = 47)
39. p value*
n (%) n (%)
Project characteristics
a
HS grantee organization type
Government agency (state agency, community government
agency such as a local health department) 18 32 24 51 0.04
Community-based non-governmental organization (health care
or non-health care) or Other organization
(including academic medical center, non-profit organization,
tribal organization, Federally Qualified Health
Center)
39 68 23 49
Intermediate outcomes
b
Increased access to the services available for our participants
Yes 46 80 25 53 0.00
No 11 20 22 47
Increased screening for perinatal depression among providers in
the community
Yes 33 58 18 38 0.04
40. No 24 42 29 62
Increased integration of prenatal, primary care, and mental
health services
Yes 31 54 16 34 0.03
No 26 46 31 66
Service and systems activities that contributed to reducing
disparities in maternal and child health outcomes
c
Enabling services
Yes 46 81 27 58 0.01
No 11 19 20 42
Interconception care
Yes 44 77 26 55 0.01
No 13 23 21 45
Long-term maternal and child health and community capacity
outcomes
d
Improved child health
Yes 22 39 10 21 0.05
No 35 61 37 79
Increased birth spacing
41. Yes 16 28 4 9 0.01
No 41 72 43 91
* Pearson’s Chi square or Fisher’s exact test
a
Data source: Maternal and Child Health Bureau Discretionary
Grant Information System
b
Data source: 2010 Project Director survey. Project directors
were asked, ‘‘Which of the following intermediate outcomes did
your Healthy
Start project achieve in 2010?’’. Multiple responses were
allowed. Only outcomes with statistically significant (p B 0.05)
relationships with
implementation of all core program components are reported
c
Data source: 2010 Project Director survey. Project directors
were asked, ‘‘To what extent did the following activities
conducted by your
Healthy Start project contribute to reducing disparities in
maternal and infant health outcomes?’’. Response options
included Primary contri-
bution, Major contribution, Moderate contribution, Minor
contribution, and No contribution or N/A. Primary contribution
and Major contribution
were classified as ‘‘Contributed.’’ Only activities with
42. statistically significant (p B 0.05) relationships with
implementation of all core program
components are reported
d
Data source: 2010 Project Director survey. Project directors
were asked, ‘‘Which of the following long term outcomes did
your Healthy Start
project achieve in 2010?’’. Multiple responses were allowed.
Only outcomes with statistically significant (p B 0.05)
relationships with imple-
mentation of all core program components are reported
Matern Child Health J (2015) 19:1292–1305 1299
123
Although projects implementing all core components more
frequently reported achievement of the majority of interme-
diate outcomes than projects that did not implement all core
components, the intermediate outcomes for which the rela-
tionship between implementation of all core components and
the outcome were statistically significant were (1) increased
access to services available for participants, (2) increased
43. integration of prenatal, primary care and mental health ser-
vices and (3) increased screening for perinatal depression.
Projects implementing all core components were signifi-
cantly more likely to report that enabling and interconception
care services conducted by the project made a primary or
major contribution to reducing disparities in maternal and
infant health, when compared with projects that did not
implement all required core components. Additionally, pro-
jects implementing all core components were significantly
more likely to report that their project had achieved increased
birth spacing and improved child health in 2010, compared
with projects that did not implement all core components.
Bivariate Analyses: Intermediate Outcomes, Service
and Systems Activities that Contributed to Reducing
Disparities in Maternal and Infant Health Outcomes,
and Long-Term Maternal and Child Health
and Community Capacity Outcomes
Results of the bivariate analyses examining the relationship
44. between project director-reported intermediate outcomes,
service and systems activities that made a primary or major
contribution to reducing disparities in maternal and infant
health outcomes and long-term outcomes revealed many
significant associations (data not shown). Intermediate out-
comes that were significantly associated (p B 0.05) with
project director-reported improvements in birth outcomes
and/or maternal health included: increased cultural compe-
tence of providers in the community; increased number of
participants with a medical home; increased awareness of the
importance of interconception care; increased screening for
perinatal depression; and increased participant involvement
in community activities addressing systems change. Healthy
Start project activities, such as interconception care, peri-
natal depression screening, enabling services, collaboration
with consumers, and LHSAP, that made a primary or major
contribution to reducing disparities in maternal and infant
health outcomes were each significantly (p B 0.05) associ-
45. ated with project director-reported improvement in birth,
maternal, and/or child health outcomes (data not shown).
Descriptive and Comparative Analyses: Birth Outcome
Performance Measures
In 2009, 20 % of Healthy Start projects had singleton LBW
rates and 59 % had IMR that were less than or equal to the
Healthy People 2010 (HP2010) LBW targets of 5 % and
4.5 per 1,000 live births [9], respectively. The Healthy
People 2020 (HP2020) targets were revised to 7.8 % (LBW
rate) and 6 per 1,000 live births (IMR) [10], and a higher
proportion of Healthy Start projects achieved these targets
than achieved the HP2010 targets (33 % achieved the
LBW target and 60 % achieved the IMR target) (data not
shown). Compared with Healthy Start projects that did not
meet the HP2020 LBW target, projects that achieved the
HP2020 target were significantly (p B 0.05) more likely to
report achieving increased access to services available for
participants and increased integration of prenatal, primary
46. care, and mental health services. Similarly, these projects
were significantly more likely to report that their outreach
and client recruitment, collaboration with community-
based organizations, collaboration with private and public
agencies, and/or collaboration with local Title V activities
made a primary or major contribution to reducing dispar-
ities in maternal and infant health outcomes. Achieving the
HP2020 target for IMR was not significantly associated
with project director-report of achieving intermediate out-
comes or of (conducting) service or system activities that
made a primary or major contribution to reducing dispar-
ities in maternal and infant health outcomes (Table 4).
Similar results were observed when comparing Healthy
Start project PM rates with state birth outcome rates. In
2009, over one quarter (27 %) of all Healthy Start projects
had a singleton LBW rate less than the rate in their state,
and 62 % had an IMR that was less than the rate in their
state. Healthy Start projects that had a lower singleton
47. LBW rate in 2009 than the rate reported for their state were
significantly (p B 0.05) more likely to report achieving
increased positive health behaviors among participants and
increased number of participants with a medical home in
2010 (data not shown).
Multivariate Analyses
The results of the multivariate analyses are presented in
Tables 5 and 6. After controlling for project characteristics,
project director-reported intermediate outcomes and other
covariates consistent with the logic model, there were no
significant associations of implementation of all core pro-
gram components with singleton LBW and/or infant mor-
tality rates. Urban project setting and state/local
government agency grantee organization were significantly
associated with higher rates of LBW, and non-urban pro-
ject setting was significantly associated with higher IMR.
As expected, LBW rates were significantly associated with
higher IMR. Intermediate and long-term program outcomes
48. reported in the 2010 PD survey were not significantly
associated with either singleton LBW or infant mortality.
1300 Matern Child Health J (2015) 19:1292–1305
123
Table 4 Association of percent singleton low birth weight
(LBW) and infant mortality rates (IMR) among project
participants’ infants meeting
HP2010 and HP2020 objective targets with Healthy Start project
director-reported achievement of intermediate outcomes and
conduct of service
and systems activities that contributed to reducing disparities in
maternal and infant health outcomes (N = 104)
PM 51 (% singleton LBW) PM 52 (IMR)
Less than
HP2010 LBW
target of 5 %
(n = 20
projects)
Less than
HP2020 LBW
49. target of
7.8 %
(n = 32
projects)
Less than
HP2010 IMR
target of 4.5
deaths per
1,000 live
births
(n = 58
projects)
Less than
HP2020 IMR
target of 6
deaths per
1,000 live
births
51. Intermediate outcomes
a
Increased awareness of the importance of interconception care
85 15 84 16 79 21 80 20
Increased awareness of disparities in birth outcomes as
community priority 75 25 75 25 71 29 71 29
Increased positive health behaviors among our participants 85
15 84 16 71 29 71 29
Increased access to the services available for our participants
85* 15 81* 19 69 31 69 31
Increased number of participants with a medical home 85 15 75
25 76 24 76 24
Increased screening for perinatal depression among providers in
the community 60 40 59 41 48 52 49 51
Increased participant involvement in Healthy Start decision-
making 45 55 50 50 47 53 47 53
Increased integration of prenatal, primary care, and mental
health services 60 40 66* 34 40 60 41 59
Increased cultural competence of providers in our community 55
45 53 47 36 64 37 63
Increased participant involvement in other community activities
addressing systems
change
20 80 31 69 34 66 36 64
52. Increased participant involvement in decision-making among
partner agencies 10 90 16 84 24 76 25 75
Service and systems activities that contributed to reducing
disparities in maternal and child health outcomes
b
Case management 90 10 88 12 93 7 93 7
Enabling services 75 25 69 31 78 22 78 22
Interconception care 65 35 63 37 66 34 66 34
Perinatal depression screening 60 40 56 44 69 31 69 31
Outreach and client recruitment 50 50 47** 53 62 38 63 37
Collaboration with consumers 60 40 50 50 57 43 58 42
Collaboration with community-based organizations 30* 70 28**
72 55 45 56 44
Collaboration with public agencies 35 65 31* 69 50 50 51 49
Collaboration with private agencies 30 70 25** 75 48 52 49 51
Consortium 35 65 37 63 41 59 42 58
Local Health System Action Plan 30 70 31 69 40 60 41 59
Collaboration with local Title V 15 85 19* 81 40 60 39 61
Collaboration with state Title V 30 70 25 75 36 64 36 64
Provider education 35 65 31 69 41 59 42 58
53. Note that Healthy People (HP) LBW targets are for LBW among
all live births, whereas Healthy Start PM 51 and State Title V
HSI 01B
measures the singleton LBW rate
* Pearson’s Chi square or Fisher’s exact test p value B 0.05
** Pearson’s Chi square or Fisher’s exact test p value B 0.01
a
Data source: 2010 Project Director survey. Project directors
were asked, ‘‘Which of the following intermediate outcomes did
your Healthy
Start project achieve in 2010?’’. Multiple responses were
allowed. A ‘‘yes’’ response indicates that the project director
reported that the project
achieved the intermediate outcome. A ‘‘no’’ response indicates
that the project director did not report that the project achieved
the intermediate
outcome
b
Data source: 2010 Project Director survey. Project directors
were asked, ‘‘To what extent did the following activities
conducted by your
Healthy Start project contribute to reducing disparities in
maternal and infant health outcomes?’’. Response options
included Primary contri-
bution, Major contribution, Moderate contribution, Minor
54. contribution, and No contribution or N/A. A ‘‘yes’’ response
indicates that the project
director reported that the service or system activity made a
primary or major contribution to reducing disparities in
maternal and infant health
outcomes. A ‘‘no’’ response indicates that the project director
reported that the service or system activity did not make a
primary or major
contribution to reducing disparities in maternal and infant
health outcomes
Matern Child Health J (2015) 19:1292–1305 1301
123
Table 5 Adjusted associations of implementation of Healthy
Start Program components with singleton low birth weight
(LBW) among Healthy
Start project participants’ infants (N = 98 projects)
Project characteristic %
Singleton
LBW
a
% Singleton LBW
less than State
55. Title V rate
b
% Singleton LBW less
than HP2010 LBW
target of 5 %
b
% Singleton LBW less
than HP2020 LBW
target of 7.8 %
b
Implemented all 5 core service components and all 4 core
systems components versus did not implement all core
components
c
0.4 0.5 0.4 0.4
Initial funding received in Phase 1 (1991–1996) versus
initial funding received in Phase 2, 3, or 4
d
1.2 1.3 0.6 0.4
Urban geographic location [urban/central city,
metropolitan area (city and suburbs)] versus not urban
56. d
2.9 0.4 0.8 0.4
Not urban geographic location (suburban, border US-
Mexico, rural) versus not not urban
d
1.6 0.6 1.7 0.4
State or local government agency grantee organization
versus community-based non-governmental
organization (health care or non-health care) or other
organization (including academic medical center, non-
profit organization, tribal organization, Federally
Qualified Health Center)
d
1.5 0.1 0.1 0.4
PM 20 (% women participants with an ongoing source of
primary and preventive care for women) (%, 2009)
d
0.0 1.0 1.0 1.0
PM 22 (degree to which Healthy Start project facilitates
health providers’ screening of women participants for
57. risk factors) (score greater than mean of all projects,
2009)
d
0.8 1.3 0.5 1.1
Achieved increased birth spacing
e
0.5 0.4 0.8 2.1
Achieved increased cultural competence of providers in
the community
f
-1.3 2.1 2.4 1.9
Achieved increased participant involvement in Healthy
Start decision-making
f
0.9 0.9 0.8 0.6
Results based on multivariable linear or logistic regression
models (separate models for each outcome), with each model
adjusted for the other
variables in the table. Bold font indicates effect estimate was
significant at p 0.10 or 95 % confidence interval [1
a
Linear model: values are b coefficients. The effect estimate
58. represents the effect per percent increase of LBW
b
Logistic model: values are odds ratios. The effect estimate
represents the effect of having a rate less than the state Title V
rate or less than the
Healthy People (HP) target. Note that HP2010 and HP2020
LBW targets are for LBW among all live births, whereas
Healthy Start PM 51 and
State Title V HSI 01B measures the singleton LBW rate
c
Data source: 2010 Project Director survey. To determine
implementation of core service components, project directors
were asked, ‘‘Which of the
following services does your Healthy Start project offer?’’
(response options: ‘‘Outreach & participant recruitment,’’
‘‘Case management,’’ ‘‘Health
education,’’ ‘‘Perinatal depression screening,’’ and
‘‘Interconceptional services’’). To determine implementation of
the core systems-building com-
ponent of having a consortium, project directors were asked
‘‘Does your Healthy Start project have at least one active
consortium that addresses maternal
and child health issues’’ (response options: Yes/No). To
determine implementation of the core systems-building
component of having a Local Health
System Action Plan, project directors were asked ‘‘Does your
Healthy Start project have a Local Health System Action Plan
59. (LHSAP)?’’ (response
options: Yes/No; a follow up question was asked to determine if
the LHSAP was specific to the Healthy Start project). To
determine implementation of
the core systems-building component of coordination and
collaboration with Title V, project directors were asked to
specify the types of collaborative
activities that their Healthy Start project established with the
State Title V agency. Projects were classified with a ‘‘yes’’
response if the project director
indicated that the State Title V agency ‘‘is a member of the
Healthy Start consortium,’’ ‘‘has a written memorandum of
understanding or agreement with
Healthy Start,’’ ‘‘provides contracted services to Healthy
Start,’’ ‘‘hosts out-stationed Healthy Start staff,’’ ‘‘participates
in joint training with Healthy
Start,’’ ‘‘has a shared staffing arrangement with Healthy Start,’’
‘‘coordinates case management or is planning with Healthy
Start for shared participants,’’
‘‘shares protocols with Healthy Start,’’ ‘‘is involved in Healthy
Start sustainability planning,’’ ‘‘has a data-sharing arrangement
with Healthy Start,’’
‘‘contributes to pooled funding streams to support joint
services,’’ ‘‘has a Healthy Start employee on their board,’’
‘‘works with Healthy Start to develop
consistent health messages for participants,’’ and/or ‘‘receives
cultural competence training from Healthy Start.’’ To determine
60. implementation of the
core systems-building component of having a sustainability
plan, project directors were asked ‘‘Does your Healthy Start
project have a sustainability
plan, that is, a plan to maintain services to the target population
after federal Healthy Start funding ends?’’ (response options:
Yes/No)
d
Data source: Maternal and Child Health Bureau Discretionary
Grant Information System
e
Data source: 2010 Project Director survey. Project directors
were asked, ‘‘Which of the following long term outcomes did
your Healthy Start
project achieve in 2010?’’. Multiple responses were allowed
f
Data source: 2010 Project Director survey. Project directors
were asked, ‘‘Which of the following intermediate outcomes did
your Healthy Start
project achieve in 2010?’’. Multiple responses were allowed
1302 Matern Child Health J (2015) 19:1292–1305
123
Table 6 Adjusted associations of implementation of Healthy
Start Program components with infant mortality rate (IMR)
61. among Healthy Start
project participants’ infants (N = 98 projects)
Project characteristic
a
Infant
mortality
rate
a
Infant mortality
rate less than
State Title V
IMR
b
Infant mortality rate less than
HP2010 IMR target of 4.5
deaths per 1,000 live births
c
Infant mortality rate less
than HP2020 IMR target of
6 deaths per 1,000 live
births
62. c
Implemented all 5 core service components and
all 4 core systems components versus did not
implement all core components
d
-0.7 1.2 1.1 1.1
Initial funding received in Phase 1 (1991–1996)
vs. initial funding received in Phase 2, 3, or 4
e
4.9 0.4 0.5 0.4
Urban geographic location (urban/central city,
metropolitan area [city and suburbs]) versus
not urban
e
-4.1 1.6 1.5 1.3
Not urban geographic location (suburban, border
US-Mexico, rural) versus not urban
e
7.4 0.5 0.6 0.5
State or local government agency grantee
63. organization versus community-based non-
governmental organization (health care or non-
health care) or other organization (including
academic medical center, non-profit
organization, tribal organization, Federally
Qualified Health Center)
e
0.7 1.0 0.9 1.1
PM 51 (% low birth weight) (%, 2009) 0.5 0.9 0.9 0.9
PM 20 (% women participants with an ongoing
source of primary and preventive care for
women) (%, 2009)
0.0 1.0 1.0 1.0
PM 22 (degree to which Healthy Start project
facilitates health providers’ screening of
women participants for risk factors) (score
greater than mean of all projects, 2009)
e
-3.0 0.7 0.8 0.6
64. Achieved increased birth spacing
f
3.8 0.6 0.3 0.5
a
Results based on multivariable linear or logistic regression
models (separate models for each outcome), with each model
adjusted for the other
variables in the table. Bold font indicates effect estimate was
significant at p 0.10 or 95 % confidence interval [1
b
Linear model: values are b coefficients. The effect estimate
represents the effect per increase in the infant mortality rate
(deaths per 1,000 live
births)
c
Logistic model: values are odds ratios. The effect estimate
represents the effect of having a rate less than the state Title V
rate or less than the
Healthy People (HP) target
d
Data source: 2010 Project Director survey. To determine
implementation of core service components, project directors
were asked, ‘‘Which of
the following services does your Healthy Start project offer?’’
(response options: ‘‘Outreach and participant recruitment,’’
‘‘Case management,’’
‘‘Health education,’’ ‘‘Perinatal depression screening,’’ and
65. ‘‘Interconceptional services’’). To determine implementation of
the core systems-
building component of having a consortium, project directors
were asked ‘‘Does your Healthy Start project have at least one
active consortium
that addresses maternal and child health issues’’ (response
options: Yes/No). To determine implementation of the core
systems-building
component of having a Local Health System Action Plan,
project directors were asked ‘‘Does your Healthy Start project
have a Local Health
System Action Plan (LHSAP)?’’ (response options: Yes/No; a
follow up question was asked to determine if the LHSAP was
specific to the
Healthy Start project). To determine implementation of the core
systems-building component of coordination and collaboration
with Title V,
project directors were asked to specify the types of
collaborative activities that their Healthy Start project
established with the State Title V
agency. Projects were classified with a ‘‘yes’’ response if the
project director indicated that the State Title V agency ‘‘is a
member of the Healthy
Start consortium,’’ ‘‘has a written memorandum of
understanding or agreement with Healthy Start,’’ ‘‘provides
contracted services to Healthy
Start,’’ ‘‘hosts out-stationed Healthy Start staff,’’ ‘‘participates
66. in joint training with Healthy Start,’’ ‘‘has a shared staffing
arrangement with
Healthy Start,’’ ‘‘coordinates case management or is planning
with Healthy Start for shared participants,’’ ‘‘shares protocols
with Healthy Start,’’
‘‘is involved in Healthy Start sustainability planning,’’ ‘‘has a
data-sharing arrangement with Healthy Start,’’ ‘‘contributes to
pooled funding
streams to support joint services,’’ ‘‘has a Healthy Start
employee on their board,’’ ‘‘works with Healthy Start to
develop consistent health
messages for participants,’’ and/or ‘‘receives cultural
competence training from Healthy Start.’’ To determine
implementation of the core
systems-building component of having a sustainability plan,
project directors were asked ‘‘Does your Healthy Start project
have a sustainability
plan, that is, a plan to maintain services to the target population
after federal Healthy Start funding ends?’’ (response options:
Yes/No)
e
Data source: Maternal and Child Health Bureau Discretionary
Grant Information System
f
Data source: 2010 Project Director survey. Project directors
were asked, ‘‘Which of the following long term outcomes did
your Healthy Start
67. project achieve in 2010?’’. Multiple responses were allowed
Matern Child Health J (2015) 19:1292–1305 1303
123
Discussion
This evaluation of the Federal Healthy Start Program using
both data from a survey of project directors and Healthy
Start project birth, service, and system outcome perfor-
mance measures data revealed a mixed set of relationships
between implementation of core program components and
long-term maternal and child health outcomes. Analyses of
the 2010 PD survey data indicate that implementation of all
core components was associated with better project direc-
tor-reported intermediate and long-term project outcomes.
This is the first analysis to use MCHB performance mea-
sure data in a national evaluation to assess Healthy Start
projects’ progress toward achieving outcomes that are
expected to occur if program elements are successfully and
68. completely implemented. Results from this evaluation are
consistent with our hypothesis (illustrated in the logic
model, Fig. 1) of a progression of achievement of inter-
mediate outcomes leading to long-term outcomes. For
example, increased screening for perinatal depression, case
management and interconception care services may have
led to PD-reported improvement in maternal health. In
addition, we found that Healthy Start projects that reported
an increase in the number of participants with a medical
home in 2010 and an increase in positive behaviors among
participants had a significantly better (lower) singleton
LBW rate among project participants’ infants than the rate
in their state.
Our analyses used state and national benchmarks, and
our findings are reinforced by the results of previously
published evaluations that were conducted by Healthy Start
projects using vital records, clinical services and program
data. Site-specific evaluations conducted by individual
69. Healthy Start projects have identified components of the
program that show a positive effect on birth outcomes of
participants’ infants when compared with demographically
similar women who did not participate in the program. For
example, evaluations of individual Healthy Start projects
found that services provided to high risk participants
resulted in improved birth outcomes such as reduced rates
of LBW, preterm birth, and infant mortality [12–14] in
addition to lower rates of sexually transmitted diseases
[15].
Although previous national evaluations of the Federal
Healthy Start Program helped to establish the importance
of the Healthy Start program components for achieving
Program goals, these evaluations relied solely on grantees’
perspectives because objective performance measure data
were not adequate for use in national evaluations. A thor-
ough examination of the PM data reported by Healthy Start
projects revealed that the quality of reported data is suffi-
70. cient for evaluation activities but also identified several key
challenges to using these data for program evaluation [8].
Our review of the notes and detailed explanations that
accompanied the PM data that grantees submitted to the
DGIS revealed data quality issues, including: 1) inconsis-
tency in the definition of the measure used by the project
with the definition specified by MCHB; 2) lack of verifi-
cation of some measures, e.g. PM 52, due to the timing of
the completion of birth–death linked files prepared by the
state vital records department; and 3) missing and incom-
plete data. These data limitations may introduce bias if the
projects that had missing data or provided incomplete data
are different from those who provided accurate and com-
plete data, or if the under-reporting or erroneous reporting
is related to the performance measures used as the out-
comes for this analysis (PM 51 and PM 52).
A potential limitation of these analyses was the possible
variation in the information source(s) used to complete the
71. PD survey. Healthy Start project staff, including the project
director and other project staff, were asked to complete the
survey, and the staff member(s) who provided responses
could have varied by project. The survey was pilot-tested
with representatives of different Healthy Start project staff
roles, but allowing survey completion by more than one
type of respondent can increase the potential for variation
in the interpretation of the survey questions and lead to
variation in responses. Responses may also have varied
based on the length of time the respondent had been with
the project, in addition to the length of time that the project
had been in operation and the program components that
were implemented. We did not have access to complete,
reliable information about other project characteristics and
program components needed to perform a comprehensive
evaluation of project implementation in a variety of com-
munity settings and to conduct analyses that adequately
addressed all of the relationships outlined in the logic
72. model. For example, participant demographic data cap-
tured by the MCHB DGIS were not available for use in
these analyses. The eligibility criteria for participation in
Healthy Start lead to some demographic similarities across
project sites; however, other important differences in the
populations served by sites may exist. More detailed
information about program implementation and outcomes
achieved by individual Healthy Start projects is needed to
improve the specificity of future evaluations.
Healthy Start projects provide services to high risk
women in the most vulnerable communities in our country.
Improving birth outcomes for project participants requires
intensive and focused services and policies that will assure
quality services within communities. Ongoing monitoring
and assessment of the implementation of these programs
and routine, standardized collection of essential birth out-
come and project implementation data will provide critical
information for evaluating what is and is not working in
73. individual Healthy Start projects and the Program as a
1304 Matern Child Health J (2015) 19:1292–1305
123
whole. MCHB could provide Healthy Start Program staff
with online tools and training to improve the reliability of
data collection and reporting. Future Healthy Start Program
evaluations should build on more robust local evaluations
at the project level as well as employ a set of focused
questions for the national evaluation that specifically
address the major issues of interest to state and national
policy-makers. Improved capacity for data collection and
documentation by individual projects would help assure
that comprehensive cross-site evaluations could be con-
ducted in the future. Resources should be provided to
assure that the systems required to conduct this type of
evaluation are in place.
Based on our experience conducting national evalua-
74. tions of the Federal Healthy Start Program, we recommend
that future evaluations explicitly connect to local, state, and
national frameworks and agendas for improving birth
outcomes and reducing health inequities. The evaluation
plan should incorporate analyses at multiple levels to
provide a robust and comprehensive examination of
Healthy Start Program activities and achievements. Most
importantly, monitoring and evaluation activities con-
ducted by individual Healthy Start projects must be
strengthened to help ensure systematic and standardized
annual reporting to MCHB of performance measure data,
program activities and accomplishments, and other data
needed for evaluation.
Acknowledgments Financial support for this study was provided
by
the Health Resources and Services Administration, Maternal and
Child Health Bureau under Contract No. HHSH250200646015I
Task
Order HHSH25034002T: An Evaluation of the Core
Components of
75. the Federal Healthy Start Program: A Cross-site Examination.
The
authors would like to acknowledge the contributions of the
Healthy
Start Grantees who participated in this evaluation, the staff of
the
National Healthy Start Association, the Healthy Start Project
Officers
at MCHB, especially Dr. David de la Cruz and Dr. Keisher
High-
smith, and the Healthy Start project team at Abt, including Dr.
Chanza Baytop, Ms. Meredith Eastman, Ms. Carolyn Robinson,
and
Dr. Meghan Woo.
References
1. Health Resources and Services Administration. (2001).
Healthy
Start Guidance, 2001. Rockville, MD: Health Resources and
Services Administration, U.S. Department of Health and Human
Services.
2. Health Resources and Services Administration. (2005).
Healthy
76. Start Guidance, 2005. Rockville, MD: Health Resources and
Services Administration, U.S. Department of Health and Human
Services.
3. Health Resources and Services Administration. (2009).
Healthy
Start Guidance, 2010. Rockville, MD: Health Resources and
Services Administration, U.S. Department of Health and Human
Services.
4. Maternal and Child Health Bureau. (2013). DGIS Reports:
Pro-
gram Data Reports. Retrieved https://perf-data.hrsa.gov/MCHB/
DGISReports/PerfMeasure/PerfMeasureReports.aspx?Report=
ProgramPerfMeasures&Archived=0.
5. Health Resources and Services Administration. (2006). A
Profile
of Healthy Start: Findings From Phase I of the Evaluation,
2006.
Rockville, MD: Health Resources and Services Administration,
U.S. Department of Health and Human Services.
6. Brand, A., Walker, D. K., Hargreaves, M., & Rosenbach, M.
77. (2010). Intermediate outcomes, strategies, and challenges of
eight
Healthy Start Projects. Maternal and Child Health Journal,
14(5),
654–665.
7. Rosenbach, M., O’Neil, S., Cook, B., Trebino, L., & Walker,
D.
K. (2010). Characteristics, access, utilization, satisfaction, and
outcomes of healthy start participants in eight sites. Maternal
and
Child Health Journal, 14(5), 666–679.
8. Abt Associates. (2012). Methodology for Analysis of Health
Resources Service Administration, Maternal and Child Health
Bureau Healthy Start Performance Measures, September 2012.
Cambridge, MA: Abt Associates.
9. Centers for Disease Control and Prevention. National Center
for
Health Statistics. Healthy People Objective Targets: Maternal,
Infant, and Child Health. Retrieved http://www.cdc.gov/nchs/
data/hpdata2010/hp2010_final_review_focus_area_16.pdf.
10. Centers for Disease Control and Prevention. National Center
78. for
Health Statistics. Healthy People Objective Targets: Maternal,
Infant, and Child Health. Retrieved http://www.healthypeople.
gov/2020/topicsobjectives2020/overview.aspx?topicid=26.
11. Taylor, Y. J., & Nies, M. A. (2012). Measuring the impact
and
outcomes of maternal child health federal programs. Maternal
Child Health Journal, 17(5), 886–896. doi:10.1007/s10995-012-
1067-y.
12. Will, J. A., Hall, I., Cheney, T., & Driscoll, M. (2005).
Flower
Power: Assessing the impact of the Magnolia Project on
reducing
poor birth outcomes in an at-risk neighborhood. Journal of
Applied Sociology/Sociological Practice, 22.2/7(2), 74–90.
13. Salihu, H. M., Mbah, A. K., Jeffers, D., Alio, A. P., &
Berry, L.
(2009). Healthy Start program and feto-infant morbidity out-
comes: Evaluation of program effectiveness. Maternal and Child
Health Journal, 13(1), 56–65. doi:10.1007/s10995-008-0400-y.
79. 14. Kothari, C. L., Wendt, A., Oemeeka, L., Overton, J., &
Sweezy,
L. C. (2011). Assessing maternal risk for fetal-infant mortality:
A
population-based study to prioritize risk reduction in a Healthy
Start community. Maternal and Child Health Journal, 15(1),
68–76. doi:10.1007/s10995-009-0561-3.
15. Livingood, W. C., Brady, C., Pierce, K., Atrash, H., Hou,
T., &
Bryant, T, 3rd. (2010). Impact of pre-conception health care:
evaluation of a social determinants focused intervention. Mater-
nal and Child Health Journal, 14(3), 382–391.
Matern Child Health J (2015) 19:1292–1305 1305
123
https://perf-
data.hrsa.gov/MCHB/DGISReports/PerfMeasure/PerfMeasureRe
ports.aspx?Report=ProgramPerfMeasures&Archived=0
https://perf-
data.hrsa.gov/MCHB/DGISReports/PerfMeasure/PerfMeasureRe
ports.aspx?Report=ProgramPerfMeasures&Archived=0
https://perf-
data.hrsa.gov/MCHB/DGISReports/PerfMeasure/PerfMeasureRe
ports.aspx?Report=ProgramPerfMeasures&Archived=0
http://www.cdc.gov/nchs/data/hpdata2010/hp2010_final_review
_focus_area_16.pdf
81. L o g i c M o d e l W o r k b o o k
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
L o g i c M o d e l W o r k b o o k
T a b l e o f C o n t e n t s
P a g e
Introduction - How to Use this Workbook
.....................................................................2
Before You Begin
...............................................................................................
..................3
Developing a Logic Model
...............................................................................................
..4
Purposes of a Logic Model
...............................................................................................
5
The Logic Model’s Role in Evaluation
............................................................................ 6
Logic Model Components – Step by Step
82. ....................................................................... 6
Problem Statement: What problem does your program address?
......................... 6
Goal: What is the overall purpose of your program?
.............................................. 7
Rationale and Assumptions: What are some implicit underlying
dynamics? ....8
Resources: What do you have to work with?
......................................................... 9
Activities: What will you do with your resources?
................................................ 11
Outputs: What are the tangible products of your activities?
................................. 13
Outcomes: What changes do you expect to occur as a result of
your work?.......... 14
Outcomes Chain
....................................................................................... 16
Outcomes vs. Outputs
............................................................................. 17
Logic Model Review
...............................................................................................
............18
Appendix A: Logic Model Template
83. Appendix B: Worksheet: Developing an Outcomes Chain
Logic Model Workbook
Page 2
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
I n t r o d u c t i o n - H o w t o U s e t h i s W o r k b o o k
Welcome to Innovation Network’s Logic Model Workbook. A
logic model is a commonly-used
tool to clarify and depict a program within an organization. You
may have heard it described as
a logical framework, theory of change, or program matrix—but
the purpose is usually the same:
to graphically depict your program, initiative, project or even
the sum total of all of your
organization’s work. It also serves as a
foundation for program planning and
evaluation.
This workbook is a do-it-yourself guide to
the concepts and use of the logic model. It
describes the steps necessary for you to
create logic models for your own
programs. This process may take
anywhere from an hour to several hours or
even days, depending on the complexity of
the program.
We hope you will use this workbook in the way that works best
84. for you:
• As a stand-alone guide to help create a logic model for a
program in an organization,
• As an additional resource for users of the Point K Learning
Center, and/or
• As a supplement to a logic model training conducted by
Innovation Network.
You can create your logic model online using the Logic Model
Builder in Innovation Network’s
Point K Learning Center, our suite of online planning and
evaluation tools and resources at
www.innonet.org. This online tool walks you through the logic
model development process;
allows you to save your work and come back to it later; share
work with colleagues to review
and critique; and print your logic model in an attractive, one-
page presentation view for sharing
with stakeholders. Free registration is required.
For those of you who prefer to work on paper or who don’t have
reliable Internet access, a logic
model template is located in Appendix A of this workbook. You
may want to make several
copies of this template, to allow for adjustments and updates to
your logic model over time.
This checklist icon appears at points in the workbook at which
you should record
something – either write something in your template, or enter it
into your online Logic Model
Builder.
85. Why evaluate?
Evaluation serves many purposes:
• Supports program and strategic
planning
• Helps communicate your goals and
progress
• Serves as a basis for ongoing learning
to make your work stronger and
more effective.
http://www.innonet.org/�
Logic Model Workbook
Page 3
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Ongoing Learning Cycle
Evaluation is an ongoing learning cycle; a process that starts
with planning, leads into data
collection, analysis and reflection, and then to action and
improvement. Logic models are the
foundation of planning and the core of any evaluation process.
As you make strategic decisions
based on evaluation findings, you move right back into the
planning stage.
86. B e f o r e Y o u B e g i n
In preparing to create a logic model, you may want to consider:
What stakeholders should I involve?
The development of a logic model offers an opportunity to
engage your program’s stakeholders
in a discussion about the program. Stakeholders might include
program staff, clients/service
recipients, partners, funders, board members, community
representatives, and volunteers.
Their perspectives can enrich your program logic model by
clarifying expectations for the
program.
What is the scope of this logic model?
• Identify a timeframe for the logic model you are about to
create. It will help you frame
short-, intermediate, and long-term outcomes and make better
decisions about resources
and activities. Many groups design logic models for a funding
or program cycle, a fiscal
year, or a timeframe in which they believe they can achieve
some meaningful results.
• This logic model structure is intended for program planning.
Define the parameters of
your program clearly. If your organization is small and only
has one program, you can
also use this structure for small-scale strategic planning.
87. Logic Model Workbook
Page 4
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
D e v e l o p i n g a L o g i c M o d e l
Many different logic model formats exist, but they all contain
the same core concepts. The
format we use in this workbook and in our online tools has
proven useful and manageable for
the nonprofit partners we have worked with, and is the result of
more than fifteen years of
program planning and evaluation experience in the field.
It’s not necessary to create your logic model all in one sitting.
It will almost certainly be useful
to talk to other program stakeholders and get their input along
the way. You can work through
the process as we have it laid out here – starting with the
problem your program is meant to
solve, and ending with your intended outcomes – or, if it’s
easier for you, you can work in
reverse, starting with outcomes and working your way
backwards.
Similarly, the names of key components may also vary among
different logic models used in the
field, but the underlying concepts are the same. In this
workbook, we identify other terms used
in the field for similar concepts. As you develop your logic
model, we encourage you to find a
88. common language to use among key stakeholders, whether that
language reflects the terms
used here or elsewhere. The important thing is that everyone
involved uses the same terms.
The components of the logic model used by Innovation Network
are:
Logic Model Workbook
Page 5
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
A series of “if-then” relationships connect the components of
the logic model: if resources are
available to the program, then program activities can be
implemented; if program activities are
implemented successfully, then certain outputs and outcomes
can be expected.
As you draft each component of the logic model, consider the
if-then relationship between the
components. If you cannot make a connection between each
component of the logic model, you
should identify the gaps and adjust your work. This may mean
revising some of your activities
89. to ensure that you are able to achieve your outcomes, or
revising intended outcomes to be
feasible with available resources.
P u r p o s e s o f a L o g i c M o d e l
The logic model is a versatile tool that can support many
management activities, such as:
• Program Planning. The logic model is a valuable tool for
program planning and
development. The logic model structure helps you think through
your program
strategy—to help clarify where you are and where you want to
be.
• Program Management. Because it "connects the dots" between
resources, activities, and
outcomes, a logic model can be the basis for developing a more
detailed management
plan. Using data collection and an evaluation plan, the logic
model helps you track and
monitor operations to better manage results. It can serve as the
foundation for creating
budgets and work plans.
• Communication. A well-built logic model is a powerful
communications tool. It can
show stakeholders at a glance what a program is doing
(activities) and what it is
achieving (outcomes), emphasizing the link between the two.
• Consensus-Building. Developing a logic model builds common
understanding and
90. promotes buy-in among both internal and external stakeholders
about what a program
is, how it works, and what it is trying to achieve.
• Fundraising. A sound logic model demonstrates to funders that
you have purposefully
identified what your program will do, what it hopes to achieve,
and what resources you
will need to accomplish your work. It can also help structure
and streamline grant
writing.
The logic model you create with this workbook can be used for
any or all of the above purposes
– any time you need to show or refer to a clear and succinct
picture of your program.
If…..
Resources Activities Outputs Outcomes
ThenThenThen
Logic Model Workbook
Page 6
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
T h e L o g i c M o d e l ’ s R o l e i n E v a l u a t i o n
The cornerstone of effective evaluation is a thorough
understanding of the program you are
91. trying to evaluate: What resources it has to work with, what it is
doing, what it hopes to
achieve, for whom, and when. In conducting an evaluation, it is
tempting to focus most of your
attention on data collection. However, your evaluation efforts
will be more effective if you start
with a logic model. Going through the logic model process will
help ensure that your
evaluation will yield relevant, useful information.
The figure below illustrates how the logic model you will build
can serve as the foundation for
future evaluation plans. (Our Evaluation Plan Workbook and
online Evaluation Plan Builder offer
guidance for creating evaluation plans.)
Logic Model
Evaluation Plan:
Process
OutputsActivities Outcomes
OutputsActivities
Evaluation Plan:
Outcomes
Outcomes Indicators
Resources
Data Collection
Data Collection
92. C o m p o n e n t s – S t e p b y S t e p
A note about our “Home Buying” example: People often ask for
examples that relate directly to
their program area—but examples for one programmatic area
can be difficult to “translate” to
another programmatic area. We use the example of becoming a
homeowner to give a more
general conceptual framework.
Problem Statement
The first step in creating a logic model is to clearly
articulate the problem your work is tring to solve—
that is, frame a particular challenge for the
population you serve. problem that frames a
particular challenge for the population your work
will try to solve.
Other Terms for
“Problem Statement”
You might also hear a problem statement
called an "issue statement" or "situation."
Logic Model Workbook
Page 7
93. I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Your problem statement should briefly explain what needs to
change: why is there is a need for
an intervention? Your problem statement answers the question,
“What problem are we working
to solve?” Include “who, what, why, where, when, and how” in
your statement.
Sample problem statements:
I do not own my own home, so I do not experience the many
financial and emotional
benefits of home ownership.
A growing number of women in Highland Falls lack the
confidence and know-how to
obtain employment and be self-sufficient due to low literacy in
our region.
In Townsville, low-income residents with bad or no credit do
not have resources available
to help them improve their current living situations.
Build Your Logic Model: When you have identified your
problem statement, insert it into
the Problem Statement box in your logic model template, or on
the “Problem/Goals” tab of the
online Logic Model Builder.
Goal
Next, think about the overall purpose of what you are
trying to measure (your program, intervention, etc).
What are you trying to accomplish? The answer to this
94. question is the solution to your problem statement, and
will serve as your goal.
Goals serve as a frame for all elements of the logic model that
follow. They reflect
organizational priorities and help you steer a clear direction for
future action.
Goals should:
• Include the intended results—in general terms—of the
program or initiative.
• Specify the target population you intend to serve.
Examples of goal statements include:
To increase my financial independence and security through
home ownership.
Significantly increase the literacy rates among children with
reading difficulties at Yisser
Elementary School by implementing a teen-tutored reading
program.
Assist clients in their effort to become economically self-
sufficient.
Improve the health status of children, ages birth to 8 years, in
Harrison County.
Other Terms for
“Goal”
95. You might also hear a goal called an
"objective" or a "long-term outcome."
Logic Model Workbook
Page 8
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Goal Tips:
• All logic model components should be connected to your goal.
Having a clear goal helps
fight the temptation to implement an interesting program that
doesn’t really “fit.”
• It’s tempting to have more than one goal, but we recommend
that you articulate one clear
solution to your problem statement. Other goals of your
program may be long-term
outcomes, rather than goals.
• Phrase your goal in terms of the change you want to achieve
over the life of your
intervention, rather than a summary of the services you are
going to provide.
• Don’t make your statement so broad and general that it
provides no guidance for your
project.
96. Build Your Logic Model: Insert your goal statement(s) into the
Goal box in your logic
model template, or on the “Problem/Goals” tab of the online
Logic Model Builder.
Rationales
A program’s rationales are the beliefs about how change occurs
in your field and with your
specific clients (or audience), based on research, experience, or
best practices. For example:
Home ownership increases a person’s options for financial
stability and wealth-building.
Current research on women leaving public income support
systems indicates that targeted job
training, partnered with a menu of support and coaching
services, can help women get and keep
living wage jobs
Success in moving into higher-paying jobs and achieving
economic self-sufficiency is closely
related to the availability of opportunities for training and
education.
These rationales all demonstrate a core set of beliefs based on
knowledge about how changes
occur in the field.
Build Your Logic Model: If you choose to include Rationales
in your logic model, record
them in the “Rationales” box on the template, or on the
97. “Rationale/Assumptions” tab in the
online Logic Model Builder.
http://www.innonet.org/?module=lmb.rationales&set_v=d3BfaW
Q9MjEwJmxtYl9kYXRhX3R5cGVfaWQ9MyZsbWJfZGF0YV9p
ZD0yMjU5Jm1vZGU9ZWRpdF9sbWJfZGF0YQ==�
http://www.innonet.org/?module=lmb.rationales&set_v=d3BfaW
Q9MjEwJmxtYl9kYXRhX3R5cGVfaWQ9MyZsbWJfZGF0YV9p
ZD0yMjU5Jm1vZGU9ZWRpdF9sbWJfZGF0YQ==�
Logic Model Workbook
Page 9
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Assumptions
The assumptions that underlie a program’s theory are conditions
that are necessary for success,
and you believe are true. Your program needs these conditions
in order to succeed, but you
believe these conditions already exist – they are not something
you need to bring about with
your program activities. In fact, they are not within your
control.
These assumptions can refer to facts or special circumstances in
your community, region, and/or
field. Examples of program assumptions are:
98. There are houses for sale for which potential homebuyers will
qualify.
There are living wage jobs available within a reasonable
distance of this neighborhood, with
adequate public transportation to reach those jobs.
Two counselors can serve a client population of approximately
40.
The first assumption demonstrates that there is a circumstance
within the community that will
enable a homebuyer to successfully purchase a home. The third
example shows that the
program manager has clearly thought out how many counselors
are needed to support the
number of participants the program will serve.
Build Your Logic Model: If you choose to include the
Assumptions behind your
program choices in your logic model, record them in the
“Assumptions” box on the template, or
on the “Rationale/Assumptions” tab in the online Logic Model
Builder.
Resources
Identify the available resources for your program. This
helps you determine the extent to which you will be
able to implement the program and achieve your
intended goals and outcomes.
List the resources that you currently have to support
99. your program. (If you intend to raise additional resources for
the program during this program
timeframe, account for them under "Activities.")
An exception: If you’re building your logic model as part of a
proposal or to justify a funding
request, list all the resources you will need for a successful
program, whether or not you have
them in hand. (You may wish to separate resources under
headings for “need” and “have.”)
Other Terms for
“Resources”
You might also hear resources called
“inputs” or “program investments”.
Logic Model Workbook
Page 10
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Common types of resources include:
- and part-time staff, consultants (e.g.,
fundraising, technical
support, strategic planning, communications), pro bono staff
services, and volunteers
cial resources: Restricted grants, operating budget, and
other monetary resources
100. infrastructure (email,
website)
inters, copiers) and
equipment specific to the
program
materials), insurance, etc.
Resource Tips:
• Identify the major resource categories for your program.
• Be specific about these resources, but do not spend a lot of
time developing a detailed
list of all actual or anticipated program expenditures.
Not specific enough Just right Too specific
Home-buying resources
Clear financial records
W2 forms
1099s
Tax returns
Bank statements
Pay stubs
Utilities bills
Credit report
101. Staff
3 full-time staff
1 part-time
1 project lead @ 40 hrs/wk
2 project associates @ 40 hrs/wk
1 part-time support person @ 20 hrs/wk
Supplies Art Supplies
25 paintbrushes
50 bottles of paint
250 sheets of paper
25 coffee cans
Dishwashing liquid
• Remember to include resources such as technology, materials,
and space: these are often
overlooked at the program planning stage, which can cause
trouble later.
• You can use your resource list as the foundation for
developing a program budget.
• Do you receive in-kind contributions? List those among your
resources.
Build Your Logic Model: List your resources statement(s) in
the Resources box in your
logic model template, or on the “Timeframe/Resources” tab of
the online Logic Model Builder.
102. Logic Model Workbook
Page 11
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Activities
Activities are the actions that are needed to implement
your program—what you will do with program
resources in order to achieve program outcomes and,
ultimately, your goal(s).
Common activities are:
• Developing products (e.g., promotional materials and
educational curricula),
• Providing services (e.g., education and training, counseling or
health screening),
• Engaging in policy advocacy (e.g., issuing policy statements,
conducting public
testimony), or
• Building infrastructure (e.g., strengthening governance and
management structures,
relationships, and capacity).
It is often helpful to group related activities together. The
number of activity groups depends on
your program’s size and how you administer it. For a large
program, there might be numerous
activity groups; smaller programs may consist of just one or
two. Each activity group will have
more specific activities under it—but remember, this isn’t a to-
do list. Getting too specific can
103. overwhelm your audience.
Examples:
For our homebuying example, we use the activity groups of
preliminary research, financial
preparation, homebuyer’s education, identify a neighborhood,
secure mortgage loan, choose a
house, and make the purchase.
A program with the goal of reducing the teen pregnancy rate in
its city might have the following
activity groups: family planning education, mentoring, and
providing individual and group
counseling.
A program with a goal of increasing organizational capacity
through strategic use of technology
might have the following activity categories: technology
planning, selecting and implementing
technology infrastructure, staff assessment and training, and
network support.
Other Terms for
“Activities”
You might also hear activities called
“processes,” “strategies,” “methods,” or
“action steps.”
Logic Model Workbook
104. Page 12
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Activities Tips:
• You can use the activities you identify here as an outline for a
work plan. Use the
activities as headings in a more comprehensive work plan that
includes staff
assignments and a timeline.
• Providing a complete list of activities helps people who are
not familiar with your
understand what it really takes to implement it—but getting too
specific can overwhelm
them. The chart below gives some examples of what level of
specificity is useful.
Activity Group: Identify a neighborhood
ACTIVITIES:
• Hire real-estate agent
• Drive around the city
This set of activities is not detailed enough. It
omits a number of key steps needed to implement
mentor training.
Activity Group: Identify a neighborhood
ACTIVITIES:
105. • Conduct Google search
• Interview friends and family
• Choose three books from the local library
about neighborhoods
• Read three books
• Hire a driver to tour neighborhoods
• Try neighborhood restaurants
• Set up review meeting
• Take friends and family on neighborhood
tours
o Send out Invitations
o Arrange transportation
This is too detailed. It would more appropriately
belong in a work plan.
Activity Group: Identify a neighborhood
ACTIVITIES:
• Research local neighborhoods--amenities
and prices
• Hire a real-estate agent
• Tour priority neighborhoods
This is just about the right level of detail for a logic
model.
Build Your Logic Model: List all activities required to
implement your program, and
group related activities together. Record them in your template
106. or on the “Activities/Outputs”
tab of the online Logic Model Builder.
Logic Model Workbook
Page 13
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Outputs
Outputs are the measurable, tangible, and direct products
or results of program activities. They lead to desired
outcomes—benefits for participants, families,
communities, or organizations—but are not themselves the
changes you expect the program will produce. They do
help you assess how well you are implementing the
program.
Whenever possible, express outputs in terms of the size and/or
scope of services and products
delivered or produced by the program. They frequently include
quantities or reflect the
existence of something new.
Examples of program outputs include numbers and descriptions
of:
• Number of home buying workshops attended
• Number of neighborhoods researched
• Number of program participants served
• Hours of service provided
107. • Number of partnerships or coalitions formed
• Focus groups held
• Policy briefings conducted
An output statement doesn’t reveal anything about quality. You
will assess the quality of your
outputs in your evaluation.
Outputs Tips:
• Make sure your outputs have activities and resources
associated with them. This is one
way a logic model is useful—to check whether a program has
planned how it will create
a product or deliver a service.
• Many people identify specific numbers for their outputs.
Begin with an estimate, based
on experience, desired impact, and resources available. Don’t
get stuck on exact
numbers; you can adjust them later.
Build Your Logic Model: List all the outputs you expect your
program’s activities will
produce. Place these in the Outputs box of the logic model
template or on the
“Activities/Outputs” tab of the online Logic Model Builder.
Other Terms for
“Outputs”
108. You might also hear outputs called
“deliverables,” “units of service,” or
“products.”
Logic Model Workbook
Page 14
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Outcomes
Outcomes express the results that your program intends to
achieve if implemented as planned. Outcomes are the
changes that occur or the difference that is made for
individuals, groups, families, organizations, systems, or
communities during or after the program.
Outcomes answer the questions: “What difference does the
program make? What does success
look like?” They reflect the core achievements you hope for
your program.
Outcomes should:
• Represent the results or impacts that occur because of program
activities and services
• Be within the scope of the program’s control or sphere of
reasonable influence, as well as
the timeframe you have chosen for your logic model
• Be generally accepted as valid by various stakeholders of the
program
• Be phrased in terms of change
109. • Be measurable. (It may take work to translate them into
measurable indicators.)
Types of Change: Organizations with diverse missions and
services share common categories
of outcomes, because outcomes are about change: changes in
learning, changes in action, or
changes in condition.
Changes in Learning:
o New knowledge
o Increased skills
o Changed attitudes, opinions, or values
o Changed motivation or aspirations
For example:
• Potential homeowners increase their understanding of the
home buying process
• Teens ages 15-18 increase their commitment to community
service.
Changes in Action:
o Modified behavior or practice
o Changed decisions
o Changed policies
For example:
• Potential homeowners have purchased their first home.
• Teens ages 15-18 participate in community service.
110. Other Terms for
“Outcomes”
You might also hear outcomes called
“results”, “impacts”, or “objectives”.
Logic Model Workbook
Page 15
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Changes in Condition:
o Human (e.g., from oppression to freedom; from
malnourishment to food
security)
o Economic (e.g., from unemployed to employed)
o Civic (e.g., from disenfranchised to empowered)
o Environmental (e.g., from polluted to clean)
For example:
• Potential homeowners have purchased their first home.
• Teens ages 15-18 have improved employment prospects
because of community
service.
Focus of Outcomes: Clarify who or what will experience the
intended changes.
1. Individual, Client-Focused Outcomes: These reflect the
111. difference the program will make in the
lives of those directly served by the program. Examples
include:
• Potential homebuyer has purchased a home (change in
status/condition)
• Parents use alternative discipline approaches (behavior)
• Participants are better able to organize and advocate for their
rights (skills)
• Children are better prepared to enter school (changed
status/condition)
2. Family or Community Outcomes: Some programs intend to
create change for families,
neighborhoods, or whole communities. Examples include:
• Higher percentage of homeowners as opposed to renters in a
low-income community
• Improved communication among family members
• Increased parent-child-school interactions
• Decreased neighborhood violence
• Community group has an inclusive membership policy, work
group practices, and
democratic governance
3. Systemic Outcomes: These illustrate changes to overall
systems and might include cases where
agencies, departments, or complex organizations work in new
ways, behave differently, share
resources, and provide services in a coordinated fashion.
Examples include:
• Integrated system of services or interagency resource sharing
• Greater coordination among partners in a system
112. 4. Organizational Outcomes: Some programs lead to internal
outcomes—both individual and
institutional—that affect how well a program can achieve
external outcomes. These produce
improvements in program management and organizational
effectiveness. Examples of
organizational outcomes include:
• Increased efficiency
• Increased staff motivation
• Increased collaboration with other organizations
Logic Model Workbook
Page 16
I N N O V A T I O N N E T W O R K , I N C .
www.innonet.org • [email protected]
Chain of Outcomes. Not all outcomes can occur at the same
time. Some outcomes must occur
before others become possible. This is referred to as the “chain
of outcomes.” (See Appendix B
for a worksheet.)
-term Outcomes: What change do you expect to occur
either immediately or in the near
future? Short-term outcomes are those that are the most direct
result of a program’s
activities and outputs. They are typically not ends in
themselves, but are necessary steps
toward desired ends (intermediate or long-term outcomes or
113. goals)
te Outcomes: What change do you want to occur
after that? Intermediae outcomes
are those outcomes that link a program’s short-term outcomes to
long-term outcomes.
-term Outcome: What change do you hope will occur
over time? Long-term outcomes are
those that result from the achievement of your short- and
intermediate-term outcomes. They
are also generally outcomes over which your program has a less
direct influence. Often
long-term outcomes will occur beyond the timeframe you
identified for your logic model.
Outcomes Chain Example
Good Health for Kids is an advocacy organization that educates
parents and guardians about
the importance of immunizing children. The staff has identified
the following program
activities:
• Develop educational literature
• Disseminate literature to social service agencies
• Develop public service announcements (PSAs)
• Identify and work with radio stations to air radio spots
The outcomes associated with these activities fall into three
categories:
114. Short-Term
LEARNING: The knowledge
parents and guardians gain
from the literature & PSAs.
• Increased understanding
among targeted parents
of the importance of
childhood immunization
• Increased knowledge
among targeted parents
of where to go to have
their children immunized
Intermediate
BEHAVIOR: The actions
parents & guardians take as
a result of that knowledge.
• Increased number of
targeted parents who take
their children to be
immunized
Closer in Time
Easier to Measure
More Attributable to Program
Long-Term
CONDITION: The conditions
that change as a result of
those actions.
• Increased number of