5671 Peachtree Dunwoody Road, NE, Suite
700
Atlanta, Georgia 30342
Contact: Jayme Hannay
Phone: (404) 851-5712
Key words: childhood immunization; patient/parent
education
The purpose of this study was to determine
the effectiveness of an enhanced patient
education program by easuring childhood
immunization rates and parental knowledge.
METHODS
A total of four sites were utilized for
this study two intervention and two control.
The intervention sites were the
Atlanta Day Shelter for Women (ADS), serving
a predominately Black population and Casa
San Jose (CSJ), serving primarily a Hispanic
population. The control site for ADS was
the Moreland Avenue Clinic (MAC). The
control site for CSJ was La Mision Catolica
de las Americas (LMCA). Patient enrollment
for the study was conducted from January
1997 through July 1997; 109 children between
the ages of 1 month and 5 years were enrolled
in the Homeless Immunization Project (HIP).
Thirty-six percent were Black (n=39);
60 percent Hispanic (n=66); and 4 percent
were biracial or white (n=4). Forty children
were from ADS; 90 percent of these children
were Black. The majority of children from
ADS were homeless or highly transient,
living in shelters, or public housing
communities. Sixty-nine children from
CSJ were enrolled, 96 percent of whom
were Hispanic. The majority of the CSJs
children were living in low-income apartment
complexes frequently shared by two or
more immigrant families. A health educator
was available 1 day per week at both intervention
sites (ADS and CSJ) to screen and enroll
eligible children whose mothers attended
weekly Women, Infants, and Children clinics.
The HIP intervention protocol consisted
of the following steps: 1) the health
educator invited women with eligible children
to participate in the study and completed
and encounter form for all women who agreed
to enroll; 2) a written pretest of immunization
knowledge and intention was administered;
3) the health educator delivered a five
minute overview on the importance of immunization
and identified nine vaccine-preventable
diseases; 4) a written post-test was completed
by the client; 5) the client was referred
to an onsite immunization nurse and immunization
services were provided as needed; 6) the
childs immunization status was confirmed
through MATCH (the states immunization
database), medical record review, or the
determination of a registered nurse and
future vaccination(s) were scheduled;
7) reminder cards were sent out at appropriate
times (when possible); and 8) the health
educator followed up with a telephone
call (when possible). At the time of enrollment,
the health educator collected demographic
and immunization history from the childs
mother and recorded it on a one-page encounter
form. Information included the childs
age, shelter/residence, immunization history,
and source of primary care. Pre- and post-tests
of parental knowledge were administered
by the health educator after completion
of the encounter forms. The tests, developed
by Mercy Mobile Health Care (MMHC) staff
with input from a CDC consultant, were
designed to determine the following: 1)
the parent_s intention to immunize her
child; 2) the parents knowledge of the
immunization schedule; and 3) the parents
attitude toward the importance of immunization.
A translator was available as needed for
Spanish-speaking clients. For the construct
of parental knowledge and attitude, data
analysis was performed on a group of 81
women (32 at ADS and 49 at CSJ) who completed
the pre and post-tests. Parents were considered
to have knowledge of immunization if,
following education by the health educator,
they were able to name at least five of
the nine major diseases that are preventable
through immunization. Mean pre and post-test
scores were compared for women at each
site. Comparisons were made between ADS
and CSJ to determine if there was a significant
difference with regard to race. To conduct
Chi square analyses, a new variable was
created (knowshot) in order to group women
into two categories according to their
test score: those scoring five or above
were considered knowledgeable (i.e., knowshot=yes)
and those with scores less than five were
not considered knowledgeable (i.e., knowshot=no).
Chi square tests of significance were
performed to determine if immunization
knowledge was associated with the childrens
up-to-date status. Because knowledge alone
does not necessarily affect behavior,
the construct intention to immunize was
also measured. Parents who responded that
immunizing their children was important
or very important were considered to have
attitudes/intentions supportive of immunization.
The standards for the analysis of immunization
status for the 109 children were based
on the states MATCH program; up-todate
was, therefore, defined as having all
the DPT (and its variants), Hib, Polio,
and MMR shots recommended for a person
of the clients age. A 1 month grace period
is allowed. At the conclusion of the study,
children were assigned to one of four
immunization status categories: up-to-date
(category 1); up-to-date on assessment
(category 2); not up-to-date or delayed
(category 3); and lost-to-follow-up (category
4). In order to perform Chi square analyses
of differences between the sites, a new
variable (shotok) was created; children
who fell into category 1 or 2 were considered
to be up-to-date (shotok=yes) while children
in categories 3 or 4 were considered not
up-to-date (shotok=no). To determine the
whether or not HIP was associated with
improved immunization compliance, a comparison
was made between the intervention and
control sites. Chart reviews were performed
on all children immunized at MAC during
the study period (21 charts). Chart reviews
were also conducted at LMCA (31 charts).
A focus group and one-on-one interviews
were also held with three clients to explore
their attitudes and knowledge about immunization,
perceptions of barriers, as well as their
assessment of HIP. EpiInfo 6 was used
to analyze the data.
Parental Knowledge and Intention to Immunize
Eighty-one parents or guardians (all female)
completed the pre
and post-test of immunization knowledge
and intention and received the educational
intervention from the health educator.
On the pre-test, 90.7 percent of parents
indicated that they intended to have their
children immunized in the next month or
two; on the post-test, 96.5 percent of
parents stated that they planned to have
their children immunized on schedule.
Vaccine knowledge tests were scored from
zero to nine with one point given for
each disease the parent was able to name.
The mean pre-test score was 4.5 (median
4) and the mean post-test score was 6.9
(median 7). Analysis by site showed that
the mean pretest scores for CSJ and ADS
were 5.3 (median 5) and 3.2 (median 3.5),
respectively. Defining parental knowledge
of immunization as the ability to name
five or more vaccine preventable diseases
and dividing parents into two groups based
on that definition (e.g., knowshot yes
or no), 48 percent of the parents sampled
were knowledgeable about immunization
on the pre-test and 86 percent were knowledgeable
on the post-test. The percentage of parents
knowledgeable about immunizations differed
significantly by site for both the pre
and post-tests. At ADS, only 28 percent
of parents were able to name five or more
vaccine preventable diseases at the pretest,
but at CSJ, 61.2 percent of parents were
knowledgeable at pretest (difference in
mean pretest scores was significant: Kruskal-Wallis
8.251, p-value = .004). On the post-test
75 percent of parents at ADS were able
to name 5 or more vaccine preventable
diseases, while at CSJ 94 percent were
able to do so (difference in mean post
test score was significant; p-value =
.04). The mean knowledge gain for women
at ADS was 3.03 points, compared to 2.0
points for women at CSJ (Kruskal-Wallis
3.738, df 1, p = .053). Immunization status
Seventy-three percent (N=80) of the children
under age five enrolled in the study were
up-to-date according to the MATCH definition
as of November 1997. Of these, 44 had
received immunizations from MMHC and were
being followed as appropriate, while 36
were determined to be up-to-date on assessment
and required no further immunizations
for the duration of the study period.
Twenty-seven percent of children were
not up-to-date (N=29). Nine of these children
were lost to follow-up (e.g., had moved
from a shelter or apartment). The majority
of those lost to follow-up (N=7) were
homeless children at the ADS. Of the study
sample, 73 children were age 2 or younger
as of January 1, 1997. Of these, 67 percent
were up-to-date. Thirty-seven had received
shot(s) from MMHC (51 percent) while 12
were determined to be up-to-date on assessment
(16 percent). The remaining 33 percent
of children under two were not up-todate
or lost. A total of 51 children received
shots from MMHC during the study period
(47 percent of the total HIP enrollment).
Thirtyfour of these were immunized at
CSJ and 17 at ADS. As of November 1, 86
percent (N=44) of the children who received
shots were up-to-date. All seven who were
not up-to-date were seen at CSJ. Seventy
percent of the children at ADS were up-to-date
and 75 percent of the those at CSJ were
up-to-date. When a two-group comparison
was made between children who were up-to-date
(either on assessment or by being immunized
by MMHC) and those who were not (e.g.,
lost to follow-up and delayed immunization),
there was no significant difference between
the two sites. Similarly, when a two-group
comparison was made between Blacks and
Hispanics, there was no significant difference
found between the groups. Of the 31 children
whose charts were reviewed at LMCA (control
site for CSJ), only three (9.7 percent)
were verifiably up-todate as indicated
by immunization status data in their charts
(all had been immunized by the nurse).
This compares with 75 percent at CSJ who
were up-to-date with their immunizations
as verified by the nurse. While all children
at CSJ had their immunization status assessed,
almost half (N = 15, 48 percent) of the
children at LMCA had not (no information
was present in their chart on immunization
status). Thirteen children were reported
as up-to-date by their parents, but their
actual immunization status was not verified.
At MAC (control site for ADS), 66 percent
of children (N=14) were up-to-date; 29
percent (N=6) were not up-to-date or lost.
One child was reported by the mother to
be up-to-date, but no information on the
childs status was present in the medical
record. The difference in immunization
levels between ADS and MAC was not significant.
Overall, 73 percent of children age five
and under enrolled in the study were up-to-date
for immunizations. Children age 2 and
under, however, had an immunization level
of only 67 percent. A more encouraging
trend can be discerned among the subset
of children who received immunization
services from MMHC. Of the 51 children
immunized by MMHC, 86 percent (44 children)
were up-to-date. Among the 40 children
age 2 or younger in this group, 82.5 percent
(33 children) were up-todate. This suggests
that once children are established in
the _Mercy system, they are likely to
return for services. In one-on-one interviews
with Black and Hispanic women enrolled
in HIP, women expressed a strong approval
of MMHC, its convenient hours, and the
cultural sensitivity of staff. Hispanic
women particularly liked the availability
of Spanish-speaking staff and the willingness
of even non-Spanish speaking staff to
communicate with them respectfully. They
reported encountering prejudice and a
lack of respect from staff at some community
health centers and health department clinics.
This study asked two questions: 1) is
the parent's immunization knowledge and
intention to immunize positively associated
with the child's immunization level; and
2) do education and follow-up services
provided by a health educator result in
higher immunization levels at intervention
sites than at control sites?
Question #1--Association between parental
knowledge and up-todate status Sixty (74
percent) of the 81 women who received
immunization education from the health
educator were up-to-date for their child's
immunization. Results suggest an
association between parental knowledge
and immunization status. Ninety-two percent
of the women whose children were up-to-date
were also knowledgeable about immunization;
only five (8 percent) of the women with
up-to-date children were unable to name
five or more vaccine preventable diseases
on the post-test. Of the women whose children
were not up-to-date, only 71 percent were
knowledgeable about immunizations; 29
percent (N=6) failed to name five vaccine
preventable diseases. The association
between immunization knowledge and up-to-date
status was significant (Chi square 3.84,
p-value=.05, Yates corrected). This association
remained when results were stratified
by site. This suggests that education
about immunization is valuable despite
the difficulty (reported by health educators)
of providing education in crowded shelters
to women with limited time and multiple
"distractions." However, it
is difficult to conclude that the HIP
educational intervention itself was responsible
for the women's level of knowledge. A
high percentage of women (particularly
Hispanics) were knowledgeable about immunization
prior to the intervention. Although the
study supports the assertion that education
is valuable, further study is needed to
determine how to identify and target those
most in need of education and the most
effective content and presentation method
for the population to be targeted. One
of the most striking findings of the study
is the significant difference between
Hispanic and Black women with regard to
knowledge of immunization. More than 60
percent of Hispanic women met the criteria
for knowledge of immunizations on the
pretest. In interviews, Hispanic women
born outside the U.S. noted the strong
government-sponsored campaigns in their
countries of origin to educate the public
about the importance of immunization.
Although HIP results confirmed a significant
difference in immunization knowledge between
Hispanic women at CSJ and Black women
at ADS, there was no difference with respect
to their childrens up-to-date status.
This suggests the possibility that barriers
of transportation, lack of insurance,
and fear of mainstream providers due to
undocumented status cancel out any advantage
conferred by greater immunization knowledge.
Question #2_Do education and follow-up
services provided by a health educator
result in higher immunization levels at
intervention sites than at control sites?
Seventy percent of children at the ADS
were up-to-date compared to 66 percent
at MAC; 75 percent were up-to-date at
CSJ compared to only 9 percent confirmed
up-to-date at LMCA. These results are
inconclusive because intervention and
control sites differed markedly with respect
to many variables other than the intervention.
Although children served at ADS and MAC
were predominately Black and homeless,
MAC clients are women residing in an adjacent
transitional shelter and are more stable
than ADS women who reside in various shelters
and are highly mobile. A pediatric nurse
practitioner is located at MAC four days
per week and provides assertive immunization
services in contrast to ADS where the
immunization nurse is available only one
day per week. These differences suggest
that the effect of the health educator
intervention may have been more evident
if a control site more comparable to ADS
had been available. Similarly, at CSJ
and LMCA, although clients were predominately
Hispanic and marginally housed (or doubled
up), differences in service delivery procedures
prevented a meaningful comparison of the
sites. At LMCA, children were treated
for illnesses by multiple providers (including
volunteers) who did not routinely assess
for immunization levels. Except for three
cases of children immunized by the Registered
Nurse (RN), no information on immunization
status (except for the mother's self-report)
was available. It was impossible to classify
the majority of children as up-to-date
or not because no information (other than
self-report) was available. A possibility
of mis-classification is a limitation
of the study at all sites. Children categorized
as not up-to-date may include children
who received services at another site
and are actually up-to-date. Staff used
the state immunization registry (MATCH)
to track children who did not return but
since a limited number of county health
departments and private physicians participate
in MATCH, only a small number of "lost"
children were subsequently found in the
registry. The ADS children were difficult
to track by telephone due to their transience.
Similarly, a change of ownership in apartment
complexes served by CSJ resulted in the
exodus of a number of Hispanic families
(which may partially account for the large
number of Hispanic children who did not
return for subsequent shots during the
study period). If tracking had been more
successful, the percentage of up-to-date
children may have been higher.
The HIP found that education was associated
with up-to-date immunization status. However,
since many women in this study were already
knowledgeable to some degree about immunization
at the start of the project, this finding
also suggests that education efforts should
focus on identifying those women least
knowledgeable about immunization to avoid
duplication. Like many clients served
by Health Care for the Homeless programs,
these women were also likely to have other
problems and barriers that require intensive
follow-up, support, and education. Follow-up,
which was limited to phone calls and mailed
reminders, was most effective with the
Hispanic population served at CSJ, who
live doubled up in low-income apartments
surrounding the intervention site and
had telephones. Not surprisingly, it was
less effective for women at ADS who were
highly mobile, often changing shelters,
and unable to leave telephone numbers
or forwarding addresses. The significance
of these types of barriers cannot be overestimated.
These barriers impede both the delivery
of services to clients as well as efforts
to conduct research to help improve service
delivery.