January 2009, Vol. 16, No. 1 Cancer Control 79
Storytelling for Promoting Colorectal Cancer
Screening Among Underserved Latina Women:
A Randomized Pilot Study
Linda K. Larkey, PhD, Ana Maria Lopez, MD, MPH, FACP, Archana Minnal, MPH,
and Julie Gonzalez, MPA
storytelling (ST) as a culturally aligned narrative method to promote colorectal cancer (CRC) prevention and
screening, compared to a risk tool (RT)-based intervention.
Seventy-eight women were randomized in this pilot study to one of two brief interventions to
communicate CRC risk reduction options: ST or an RT. Measures of behavioral intentions relative to CRC
prevention and screening were obtained following the intervention.
Results: Mean scores for intent to obtain and recommend endoscopy to others were significantly better for
participants receiving ST than RT (P = .038 and P = .011, respectively). All participants expressed intent to
increase fruit and vegetable consumption and physical activity in response to interventions. Post-intervention
perceptions of cancer risk and fear of CRC were not significantly different for participants receiving ST
compared with RT. Pre- to post-intervention perceptions of risk increased in ST and decreased in RT, while
decreases in fear were similar across both intervention groups.
Conclusions: Storytelling may be an effective approach for changing CRC risk-related behavioral intentions
among Latinas. Mediating factors (such as perceived risk or fear) often used to predict behavior change may
not adequately explain the potential persuasive mechanisms of storytelling.
In a low socioeconomic-status population of Latina women, we evaluated the potential of
develops, endoscopic screening becomes even more
important as a prevention strategy.5
In addition to screening, the American Cancer Soci-
ety and the National Cancer Institute both recommend
increasing activity and consumption of fruits and veg-
etables as strategies to reduce the risk for cancer, two
important risk factors associated in general with cancer
and especially for CRC.6Arizona data from the 2003
Behavioral Risk Factor Surveillance Survey7indicate a
larger percentage of Latinos are not meeting the goal of
5 or more servings of fruits and vegetables daily (80.7%
compared with 77.4% non-Hispanic), and 57.9% of
Latinos (compared with 49.9% of all respondents) in
Arizona report not meeting recommendations for phys-
ical activity levels. Latinos are more likely to be obese
(24.4% compared with 19.2% of non-Hispanics). Thus,
the need to develop programs that will be effective in
promoting healthier lifestyles and screening behaviors
regarding CRC is becoming increasingly important in
The Promotora Approach and Storytelling
Community health advisors (CHAs) have been used for
decades to improve health knowledge and behaviors in
underserved populations and rural communities.8One
of the greatest strengths of this approach is that CHAs
(or promotoras de salud,the term used among Latinos,
Although rates of most cancers are generally lower
among Latinos, certain cancer-related risk factors and
screening deficits place this group at increasing risk for
Breast cancer and colorectal cancer (CRC)
place first and third, respectively, for cancer mortality
among Latina women.2Lower rates of cancer screen-
ing contribute to the trend of late-stage detection
among Latinos, especially for CRC and breast cancer.2,3
The screening rates of CRC and the readiness to
obtain these screenings are particularly low in the Lati-
no population, most especially among those with lower
socioeconomic status and poor access to health care.2,4
Given the potential for colonoscopy to prevent cancer
when polyps are detected and removed before cancer
From the Department of Family and Community Medicine (LKL,
AM, JG) and the Department of Hematology and Oncology (AML)
at the University of Arizona College of Medicine; and the Arizona
Cancer Center at the University of Arizona (LKL, AML, JG).
Archana Minnal, MPH, is now with the Maternal, Child and Adoles-
cent Health Program at the California Department of Public Health
Submitted January 8, 2008; accepted July 1, 2008..
Address correspondence to Linda K. Larkey, PhD, Department of
Family and Community Medicine, University of Arizona College of
Medicine, Arizona Cancer Center, 10510 N. 92nd Street, Suite 100,
Scottsdale, AZ 85258. E-mail: email@example.com
Abbreviations used in the paper: CRC = colorectal cancer, ST = story-
telling, RT = risk tool, HCRI-G = Harvard Cancer Risk Index-General.
January 2009, Vol. 16, No. 180 Cancer Control
often shortened to promotoras) are members of their
own communities and have the natural ability to reach
others with culturally sensitive approaches by tailoring
their methods and messages to meet the special needs
of their friends and neighbors.9,10
CHAs are considered by many researchers to be
more effective than professionals in influencing behav-
ior change and overcoming these barriers because of
the large number of characteristics they share with the
These commonalities include
social, historical, environmental,and ethnic characteris-
tics as well as verbal and nonverbal language, an under-
standing of the target population’s health beliefs and
barriers to preventive behaviors,8,11and enhanced
empathy with and responsibility toward the population
and its unique and complex health care needs.12Thus,
CHAs are able to communicate more effectively in the
cultural language of their community. This inherent cul-
tural understanding and the face-to-face flexibility
brought by the CHAs allows them to promote health
messages in ways that parallel the cultural norms.
Specifically related to many traditional cultures, they
present a more culturally integrated approach by pre-
senting health-related information in the oral tradition
in the contexts of their lives, families, and communities.
The CHA approach is commonly used for cancer
prevention and screening promotion programs. In gen-
eral, programs such as Por la Vida13and Compañeros
en la Salud14that reflect cultural norms of the com-
munity and develop trust through local representatives
from those communities have been shown to be suc-
cessful in promoting breast and cervical cancer screen-
ing among Latina women of the southwestern United
States. Less has been done to test culturally relevant
ways to encourage CRC screening specifically, but
lessons learned from cancer screening in general might
apply to some degree to CRC screening.
One important way that cultural elements are rep-
resented and communicated is through the stories told
by promotoras and by participants in the classes or
groups receiving the educational interventions. Sharing
personal experiences and examples from the lives of
friends and family members has been observed among
Latina women for advocating for clinical trials, encour-
aging dietary changes, and utilizing cancer screen-
ing.12,15We propose using storytelling (ST) as a more
specific mode of sharing personal experiences, based
on evidence of its value as an educational tool and its
use of role modeling, socialization, and positive accul-
turation in a variety of behavior change contexts.16-19
Stories are narrative accounts of causally linked
sets of events and characters, whether true or fictitious,
and closing with a sense of completeness. Vivid, emo-
tionally engaging stories “enable a leap in understand-
ing by the audience so as to grasp the possibility of
change” for self or the community.20Using fiction has
also been shown to impact memory and influence atti-
tudes through vivid characters and story lines more
powerfully than unadorned factual information.16,17,21
Thus, stories may not only serve to communicate cul-
turally familiar details, but also have inherent qualities
that may enhance persuasive impact.
Risk Tool-Based Information
Most health behavior theories suggest that perception
of personal risk is necessary (but not sufficient) for a
person to change behavior (eg, screening for early
detection or action to prevent).22-24
efforts to promote health behaviors often begin with a
message that communicates information about risk,
including information on behaviors that increase or
decrease chances of a negative outcome. Such infor-
mation may not only increase knowledge, but also
“facilitate [informed] decision making, motivate new
[healthy] behaviors, and change existing [unhealthy]
behaviors.”22Perceived risk for cancer is generally seen
as a factor motivating cancer screening but with some
mixed results and with variation according to type
Similar uneven associations between
perceived risk and screening behavior seem to hold for
low-income and ethnic minority populations.26
In both community and clinical settings, use of a
risk assessment tool to communicate either an
absolute, numeric risk or a relative risk has been gen-
erally effective for informing and correcting percep-
tions of risk for CRC, but it is not necessarily effective
for achieving screening outcomes. Among the usual
choices for communicating risk based on an indexing
of risk factors, “gambling odds” or frequency formats
(eg, 1 chance in x), visual formats, or relative terms
(chances compared to others your age) seem to be
most readily understood.27,28
An example of a cancer risk tool (RT) adapted for
wide application is the Harvard Cancer Risk Index-Gen-
eral (HCRI-G). It serves as a quick assessment tool and
as instant feedback of personalized cancer risk in
numeric, relative terms (ie, risk level compared to simi-
lar others the same age). At the same time, the HCRI-G
can be used as an educational tool that teaches about
specific risk factors that are important for prevention
or early detection of cancer.
Despite the growing popularity and appeal of
using parsimonious risk factor-based tools for commu-
nicating risk information,29,30there are potential pit-
falls. Communicating personalized risk information via
numeric data may not be well understood among pop-
ulations with low numeracy skills27,31and, when it is
understood, may or may not correct misperceptions of
risk.32,33Many lay people, especially those less educated,
do not accurately interpret or recall most of the numer-
ical presentations of risk.28Moreover, efforts to correct
perceptions of risk occasionally backfire when initial,
For this reason,
January 2009, Vol. 16, No. 1 Cancer Control 81
uninformed estimates of personal risk for cancer are
higher than the corrected risk perception after educa-
tion, thus resulting in decreased motivation to screen.
ST for Risk Information
ST methods of communicating risk and prevention
information stand in contrast to the usual numeric
methods. Rather than evoking the logical language of
numbers, comparisons, and lists of weighted risk fac-
tors, a story can embed information in the context of
fictitious but believable characters and situations. Spe-
cific information about risk factors and the effects of
changing behavior on cancer risk can be conveyed
through narrative rather than utilizing lists of factors
and numbers. The same information can be communi-
cated while being framed by culturally relevant context
and characters. Identification with the story with cul-
turally familiar details and circumstances may be key to
engagement in the message,34making it more likely
that individuals will continue to process the informa-
tion as being personally relevant rather than ignore it.
For example, stories have been used successfully in
the PROMISE program focusing on HIV prevention to
communicate risk and suggest ways to alter behavior to
lower risk. Lay members in the HIV target population
deliver the stories. The method is theorized to work by
tailoring the interventions to the community profile
and by including elements in the story to not only help
participants identify with the characters and story
lines, but also provide social modeling of effective
action to reduce risk.35
Other examples show promise in reaching special
populations through culturally aligned stories. In a
study that utilized ST to teach African American women
about breast health, women were able to identify fears,
learn correct information, and validate their experi-
ences with others who shared their culture and val-
ues.36There is also evidence that using stories may
help overcome resistance to messages that might cre-
ate anxiety and promote prevention as an important
practice.18Particularly for people who have had low
prior involvement in the health issue,37capturing atten-
tion may better be achieved through an engaging story
rather than with direct risk information.
Our proposed methods for communicating about
cancer risk and prevention are drawn from years of
observing and documenting methods used by promo-
toras. Numeric data are rarely used in these contexts;
rather, personal stories are shared from the heart with
content familiar to and emotionally evocative for the lis-
tener. We chose ST to test as an alternative, potentially
more culturally aligned method for communicating risk
information and the call to change behavior. This method
was compared in a pilot study to a standard RT-based
method, the HCRI-G, to examine effects on behavioral
intentions regarding CRC prevention and screening.
We examined whether three key behaviors — CRC
screening, fruit and vegetable consumption, and physical
activity — related to prevention of CRC might be more
effectively promoted using ST rather than RT-based com-
munication methods. The CRC screening method pro-
moted in our interventions included both colonoscopy
and flexible sigmoidoscopy. Although the evidence is
strong for primary prevention when polyps are found
and removed, and colonoscopy is considered the best
currently available tool to detect lesions and remove
polyps,38a high proportion of our target population was
expected to be uninsured or underinsured. Colonoscopy
is generally less accessible than flexible sigmoidoscopy in
underinsured populations, so both were suggested and
encouraged as options. Evidence of vegetable consump-
tion associated with reduced risk for CRC39,40led us to
choose the more broadly promoted target of increasing
fruits and vegetables in the diet. Physical activity was also
selected for promotion as a factor for reducing CRC risk,
given the strong evidence of the association6and the rec-
ommended clinical guidelines of the American Gastroen-
terological Association.41The following hypotheses were
proposed to be tested in a pilot study to examine the
potential of ST for promoting CRC prevention behaviors
in this cultural context:
Hypothesis 1: Latina women aged 50 years or
older who are due for CRC screening and are exposed
to ST vs an RT-based educational intervention will
demonstrate greater intent to obtain endoscopy.
Hypothesis 2: Latina women of any age who are
exposed to ST vs an RT-based educational intervention
will express greater intent to eat more fruits and veg-
etables and increase physical activity.
As our past work with promotoras and the Latino
population taught us, a likely outcome of teaching pre-
vention information is that women would then pro-
mote this important prevention message to others
within the family and friendship network, especially in
the culturally socializing context of ST. We added this
Hypothesis 3: Latina women of any age exposed
to ST vs an RT-based cancer prevention intervention
will more often express intent to recommend
endoscopy to others.
Development of Interventions and Translations
The story (the central feature of the ST intervention)
was written by a team of promotoras and other female
staff through an iterative process of discussing person-
al experiences as well as stories from friends and rela-
tives and creating a composite story of a single fiction-
al family. The elements of the story were selected by
this team to reflect values and themes of the Latino cul-
ture, with an emphasis on women’s experiences to
January 2009, Vol. 16, No. 182 Cancer Control
keep the stories aligned to one gender’s point of view.
The choice to limit this pilot study to the female/Latina
point of view was based on an effort to keep the study
population more homogenous, with a narrow field of
ethnic and gender culture to which to tie the story con-
tent. Events and dialogue were designed to teach about
general cancer risk factor information and the potential
for changing risk for cancer when modifiable risk fac-
tors are adjusted. CRC-specific risk information and
screening guidelines were included,with details of fam-
ilies and situations common to Latinos. For example,
during a family crisis involving two adult daughters’ con-
versations about their father’s scheduled colonoscopy,
one of their children brings home a school assignment
where he has to keep track of the fruits and vegetables
eaten during family meals. In this story line, drama and
tension are created by proposing a potential threat (pos-
sible CRC) to the father, suspense is created as the infor-
mation about risk factors is conveyed through a number
of interactions in the family and in health care encoun-
ters, and eventually tension relief is created as the lis-
tener discovers the father does not have cancer. The
point of the story was that all family members, including
women, need to be screened at age 50 years and older
for early detection, and even prevention,of CRC.
The RT-based intervention was based on the HCRI-
G, a risk factor assessment tool communicating relative
risk for cancer, also specifying CRC risk factors and
screening guidelines. This tool has been tested with
low-literate and Latino populations42and provides a
composite risk assessment tool weighted across the
most prevalent cancers and associated risk factor
weights.6This method of communicating personal risk
is visual, easy to understand, and personalized (defining
one’s own risk level relative to similar age others,
marked as a number on a bar graph). The primary pur-
pose for using the HCRI-G was as a teaching tool to
communicate the key modifiable risk factors and to ini-
tiate a discussion about what could be done to improve
the CRC risk profile.
The HCRI-G and the story were translated to Span-
ish and back-translated to English for comparison of
meaning across translations. The use of back-translation
is a well-accepted method of double checking that the
meaning of the translation is equivalent to the origi-
nal.43We adapted the method so that a team of 4 to 5
bilingual evaluators, including both English- and Span-
ish-language dominant speakers, participated in the
process. Wherever any wording differences occurred
in the back-translation compared with the English orig-
inal, both English and Spanish versions were reviewed
by this team of evaluators. Discrepancies in any subtle
meaning across translations were adjusted in Spanish
until all members agreed that the word choices were
clearly representative of the English version. The same
methods were used to translate the survey instrument
(described below), including demographic information
and all scales and outcome measures.
Recruitment and Intervention
Latina women were contacted through a number of
recruitment sites that were part of a larger promotora-
led intervention study. At these sites, women were
invited to participate in this brief intervention pilot
study prior to entering the larger intervention. Women
were eligible for the current study if they were Latina,
over 18 years of age, due for CRC screening if over age
50, and willing to be randomized and participate in a
session that included a short educational intervention
and a pre- and post-questionnaire. A total of 78 women
(47 women aged ≥ 50 years) agreed to participate in
this pilot study.
Once informed consent was obtained, participants
were randomized to either ST (n = 38) or RT (n = 40)
intervention and were asked a series of questions to
assess baseline levels of perceived cancer risk, CRC risk
(and CRC perceived severity), fear of CRC, and intent to
recommend or obtain endoscopy (colonoscopy or flexi-
ble sigmoidoscopy), if due. Initially, 32 women were
recruited, and after the first data were reviewed, three
other questions were added to the survey to enrich the
study and clarify endpoints (see “Measures” below). This
type of midstream change in design is consistent with a
participatory action research method in which feedback
from participants shapes the research in progress, a pat-
tern that is possible and preferred with studies in com-
Women randomized to the ST intervention listened
to a story in an individual, face-to-face context in Eng-
lish or Spanish (depending on stated preference of the
participant) read by a promotora who was culturally
similar to the participant (Latina). The promotoras read
the story with dramatic style, playing the parts of the
characters in dialogue. The story reflected the socio -
economic profile and culture of participants in the
pilot study and was read to them in their preferred lan-
guage, either English or Spanish.
The HCRI-G questions were read to individual par-
ticipants, also in an individual, face-to-face context, in
English or Spanish according to the preference of the
participant. After a participant answered all questions,
the promotora assisted the individual through scoring,
discussing relative risk for cancer compared to others the
same age and gender (ie, pointing out the total score on a
graded 5-point Likert-style ruler scale showing a range
from Much Below Average to Much Above Average to
describe the risk level). Each of the risk factors addressed
in the HCRI-G were also mentioned in the story to assure
parallel information in both arms of the study.
After each standardized intervention, story,or HCRI-
G, the promotora discussed with each participant the
modifiable risk factors that, if changed, could reduce
January 2009, Vol. 16, No. 1 Cancer Control 83
risk for cancer and, more specifically, for CRC. In both
interventions, the promotora described endoscopic
options and included the message that CRC may be pre-
vented when polyps are removed during colonoscopy.
All participants selected Spanish for the interven-
tion and questionnaires. Each study participant took 8
to 15 minutes to experience the intervention. Together
with informed consent and pre- and post-question-
naires, participant involvement was 30 to 40 minutes
for the entire procedure.
Mediation of Variable Measures
Studies that include risk information often examine
responses such as perceptions of risk and affective
response to the message. ST has not been used specifi-
cally in the context of communicating cancer risk infor-
mation, so the variables that might make a difference
have not yet been defined. We chose to add measures
to assess perceptions of risk and fear to examine possi-
ble differences in mediating factors that may affect how
the two interventions impact intentional outcomes.
Perceived Risk for Cancer
Participants’ perceived risk for cancer was assessed pre-
and post-intervention, using the question used in the
original Internet-based HCRI teaching/risk assessment
instrument45that asks: “Compared to other people my
age, my chances for getting cancer are...” Response
choices were “much below average” and “below aver-
age,” up to “much above average” arranged on a 5-point
Fear of Colorectal Cancer
Fear of CRC was assessed pre- and post-intervention
using three items drawn from the breast cancer fear
scale46and adapted for CRC: “When I think about colo -
rectal cancer...” Response choices were “I feel nervous,”
“It scares me,” and “I feel uneasy,” with a 5-point Likert-
style agree/disagree response scale.
Intent to Obtain Endoscopy
Intent to obtain endoscopy within the next 2 years was
assessed by asking “Do you plan on having a sigmoid -
oscopy or colonoscopy in the next two years?” with
“yes” or “no” response options.
Physical Activity and Fruit and
Baseline levels of each were obtained with questions
adapted from measures to provide global estimate serv-
ings per day and amount of moderate-to-vigorous physical
activity. Baseline levels of fruit and vegetable consump-
tion were estimated using a food frequency format.47
Physical activity was estimated from a yes/no response to
a question regarding activities that “make your heart rate
go up or make you breathe harder” and estimates of times
per week and minutes per session.48
After 32 women completed the pilot study, we
noted that participants were spontaneously and enthu-
siastically reporting that they planned to discuss what
they learned with others, family and friends. We recon-
sidered what else might be important to understanding
CRC-related responses to the interventions, particularly
the social context of family and friendship ties so
important in Latino culture, and included additional
questions, discussed below.
Intent to Recommend Colorectal Cancer Screening
A single question was asked to evaluate intentions to
“recommend CRC screening to any of your female rela-
tives or friends” with “yes” or “no” response options.
Data were analyzed using SPSS 14.049(SPSS Inc,
Chicago, IL) using t tests to examine hypothesized dif-
ferences in outcome measures. The 3-item scale for
measuring fear of CRC was reliable (Cronbach’s alpha,
.82 and .94 in pre- and post-intervention administra-
Of the 78 women completing the intervention, 47 were
aged 50 years or older. Mean age of all participants was
49.84 years (ST group mean = 50.52 years, RT group
mean = 49.16). Among the 78 participants, 42 (54%)
were uninsured and 45 (64%) did not have a regular
health care provider. Twenty-two (42%) reported less
than $15,000 annual household income, and 11 (22%)
reported under $25,000. A sixth-grade education or
less was reported by 47%, while 10% completed
7th–8th grade and 30% attended high school. These
demographic data and distributions across arms of
study are presented in Table 1.
Baseline levels of fruit and vegetable consumption
were much lower than the Behavioral Risk Factor Sur-
veillance Survey (BRFSS) assessments of Latinos in Ari-
zona, indicating 19.3% eat 5 or more servings per day.
In our population, only 2.1% stated they eat 5 or more,
and the mean servings per day were 1.70 (SD = 1.061).
Mean minutes per week of moderate to vigorous activ-
ity for all participants across intervention groups was
71.53 (SD = 117.285). The large standard deviation is
due to 8 participants whose activity levels were very
high (250 to 600 minutes per week, consistent with com-
ments from some of these participants who described
that they held physically active jobs. When these are
eliminated, mean minutes per week were 30.87 (SD =
32.52), with 16 women reporting no activity.
Baseline Assessment of Group Differences
The t tests run to examine equivalence for age, income,
fruit and vegetable consumption, physical activity, and
January 2009, Vol. 16, No. 184 Cancer Control
regular provider and insurance status indicate that ST
and RT groups were not significantly different at base-
line. Similarly, perceptions of risk for cancer and fear of
CRC were equivalent for both groups at baseline.
The primary outcome variable — intent to screen for
CRC via endoscopy — was tested for significant differ-
ence in scores between ST and RT groups for women
over 50 years of age (n = 47). Mean score for intent to
obtain endoscopy was significantly lower (with lower
score indicating greater intent) for the ST group than
for the RT group (P = .038) (Table 2). Intent to increase
physical activity and increase consumption of fruits
and vegetables was examined for 78 participants, all
ages. Responses were “yes” for all ST and RT partici-
pants, so evaluation of difference between groups was
not possible for the second and third hypotheses.
Intent to recommend CRC screening to others (the
research question added after the first 32 participants,
so that n = 43; 3 with missing data) was also signifi-
cantly higher for women in the ST group than in the RT
group (P = .011).
After the interventions, the mean score for per-
ceived risk for cancer was greater in the ST group com-
pared with the RT group, but not significantly so. Post-
intervention scores for fear of CRC were essentially the
same comparing these two groups, but fear decreased
significantly (as indicated by higher scores) for both
groups combined (P = .015). Although there was no
difference in fear of CRC between groups after expo-
sure to the intervention, fear was substantially reduced
for participants in both arms of study.
Our pilot data suggest that intent to obtain endoscopy
among women aged 50 years and older and,interesting-
ly, to recommend endoscopy for others (when women
of all ages are asked, including those younger than CRC
screening age) was greater for Latinas exposed to the ST
intervention than the RT-based intervention.
The staff who implemented this pilot study report-
ed a great deal of emotional involvement from the
women exposed to the story. Anecdotally we can report
that women often would become so involved in the
story that they would exclaim their relief when they dis-
covered that “Papa” indeed did not have CRC. Percep-
tions of risk seemed to be differently affected by the two
interventions, with the more emotionally charged ST
intervention resulting in higher perceptions of risk than
the RT intervention. Even so, both seemed to generate a
reduction in fear so it is not clear what mechanisms
influence intentions for screening relative to these inter-
ventions or how stories might potentially affect actual
behavior. Additional factors not yet assessed that are
unique to narrative may explain the difference.
Table 1. — Participant Demographics by Study Arm*
(N = 78)
(n = 38)
Risk Tool Participants
(n = 40)
Hispanic/Female 78 3840
Age (mean)49.84 yrs 50.52 yrs49.16 yrs
Annual Household Income
$15,000 to $24,999
$25,000 to $34,999
$35,000 to $44,999
Covered by insurance
Regular Health Care Provider
0 – 6th grade
7th – 8th grade
9th – 12th grade
Beyond high school
* Valid percents reported, calculated without missing data.
January 2009, Vol. 16, No. 1 Cancer Control 85
As a small pilot study in a community setting, there
were limitations to the study. For example, our mea-
sures needed to be brief assessments based on larger
multiple-item, validated scales, but these slimmer tools
have not been validated. The assessment of fear with a
3-item scale adapted directly from a validated measure
of breast cancer fear demonstrated initial reliability, but
otherwise our single-item indicators depend on face
validity. Although significant effects were distinguished
between the interventions, the chosen assessments of
proposed mediating variables do not provide much
explanation for the effects. Also, in the community set-
ting and with limited time for engagement, we lost the
opportunity to fully follow up with those participants
completing the HCRI-G. For those at higher levels of
risk,there could have been more focus on colonoscopy
as a better option for screening and diagnostic visual-
ization of the colon. Larger studies with more elaborate
interventions should be designed to include more
detailed recommendations tailored to risk level.
Intent to screen is not always the best predictor of
actual behavior, particularly for a difficult test such as
flexible sigmoidoscopy or colonoscopy that requires
planning, preparation, transportation, substantial time,
and cost (for uninsured persons). Our results suggest
only the first step of the process, the decision step,
might be enhanced with a culturally relevant method of
communication. Barriers to follow-through, such as
lack of insurance or transportation, would likely dilute
the effects of initial intentions. A program that encour-
ages a decision to screen or change lifestyle behaviors
would likely require logistical, financial,and social sup-
port, skills enhancement, and ongoing persuasive mes-
sages to achieve behavioral results.
Future studies need to explore more fully a set of
proposed mediating factors (including factors implicated
by the narrative paradigm such as the potential of story
to build identification and engagement or to enhance
cognitive processing)50,51on CRC screening intentions as
well as factors that may be specific to changing percep-
tions of social norms. Moreover, as the ST intervention
model is explored further, actual screening behavior will
need to be assessed among Latino men and women. A
similar pilot study was recently completed that included
men and provided more sensitive measures of dietary
and physical activity change intentions.52
showed positive effects of ST compared to a RT inter-
vention, suggesting further support for the potential of
this method of health promotion while underscoring the
need to take this research to the next step. Given the ini-
tial favorable results of these pilot studies, additional
work to examine theoretically based variables’ effects
(going beyond the risk communication model) on actual
behavioral outcomes is critical.
A recent survey of key Latino opinion leaders
(researchers, scientists, and health service organization
leaders) regarding cancer prevention and control for
Latinos listed CRC as the second most important cancer
site to address among this population, with recommen-
dations for special emphasis on culturally competent
risk communication.53As a first step to meet this chal-
lenge, we have taken the elements of cancer risk infor-
mation and, more specifically, CRC risk factor informa-
Table 2. — Comparisons of Outcomes in Storytelling (ST) and Risk Tool-Based (RT) Interventions
Pretest MeanPosttest Mean Difference Between RT and ST
Intent to screen (over age 50 yrs)
RT, n = 23
ST, n = 22
*1.35 (SD .487)
*1.09 (SD .294)
Intent to recommend (add’l)†
RT, n = 24
ST, n = 19
*1.25 (SD .442)
*1.00 (SD 1.00)
Perceived risk for cancer (all)‡
RT, n = 40
ST, n = 37
2.80 (SD 1.07)
2.81 (SD 1.02)
2.57 (SD .846)
2.98 (SD 1.19)
Fear of CRC (All)‡
RT, n = 37
ST, n = 34
*1.78 (SD 1.01)
*1.98 (SD 1.08)
*2.31 (SD 1.30)
*2.35 (SD 1.31)
* Lower score indicates higher attitude levels (ie, greater fear, greater intent).
** P < .05, equal variances not assumed.
† “Add’l” indicates second phase participants only, after additional question was added.
‡ “All” indicates all participants, all ages and in first and second phase, were surveyed.
January 2009, Vol. 16, No. 186 Cancer Control
tion and incorporated these messages in the text of a
story created by and for members of Latino culture.
Numeric-oriented risk messages have been used in
community settings with mixed results regarding can-
cer prevention and screening results. In low-income
Latino populations, health literacy and numeracy may
limit the effectiveness of these messages. We stepped
outside this numeric-oriented risk data paradigm to test
a lay model for influencing behavioral intentions com-
pared to the standard RT-based communication and
didactic, factual teaching. ST may present a more
coherent, culturally consistent method of communicat-
ing, fitting with the norms of promotora interventions.
The population of Latina women recruited into
this pilot study was particularly “underserved”: they
had low incomes and low levels of education, lacked
insurance, and spoke mostly Spanish. The risk profile
of this population, including low levels of fruit and veg-
etable consumption and poor compliance with CRC
screening recommendations, poses even greater chal-
lenges for promoting cancer prevention and screening
behaviors. Thus,the special profile of the test popula-
tion needs to be taken into account as next steps are
developed, recognizing that while the story approach
was effective in this group, it may require different
story elements in other groups or it may not be as
effective. Our finding that women are encouraged to
recommend screening to others, family, and friends,
with only this single exposure to a story, is an impor-
tant platform from which to design and further test
Appreciation is expressed to the promotoras de salud
(Latina community health advisors) for their creative
conceptual contribution and for telling their stories.
No significant relationship exists between the authors and the com-
panies/organizations whose products or services may be refer-
enced in this article.
This pilot project was conducted within the context of a larger study
funded by the American Cancer Society, Juntos en la Salud: Cancer
Prevention and Screening for Latinas #TURSG-03-080-01-PBP.
1. Ramirez AG, Suarez L. The impact of cancer on Latino populations.
In: Aguirre-Molina M, Molina CW, Sambrana RE, eds. Health Issues in the
Latino Community. San Francisco, CA: Jossey-Bass; 2001:211-244.
2. American Cancer Society. Cancer Facts and Figures for Hispanic/
Latinos 2003-2005. 2003;1-14. http://www.cancer.org/downloads/STT/
CAFF2003HispPWSecured.pdf. Accessed August 7, 2008.
3. Ward E, Jemal A, Cokkinides V, et al. Cancer disparities by race/
ethnicity and socioeconomic status. CA Cancer J Clin. 2004;54(2):78-93.
4. American Cancer Society. Arizona Cancer Facts and Figures 2004-
2005: A Sourcebook for Planning and Implementing Programs for Cancer
Prevention and Control. Phoenix, AZ: American Cancer Society, Great
West Division, Inc; 2004. http://www.cancer.org/downloads/COM/AZ%20
Facts%20and%20Figuresletter.pdf. Accessed August 7, 2008.
5. Pignone M, Rich M, Teutsch SM, et al. Screening for colorectal can-
cer in adults at average risk: a summary of the evidence for the U.S. Pre-
ventive Services Task Force. Ann Intern Med. 2002;137(2):132-141.
6. Colditz GA, Atwood KA, Emmons K, et al. Harvard report on cancer
prevention volume 4: Harvard Cancer Risk Index. Risk Index Working
Group, Harvard Center for Cancer Prevention. Cancer Causes Control. 2000;
7. Arizona Department of Health Services. BRFSS Prevalence Data.
http://www.azdhs.gov/plan/brfs/reports/rpt03.pdf (state-specific) and http://
apps.nccd.cdc.gov/brfss. Accessed August 7, 2008.
8. Rosenthal EL. A Summary of the National Community Health Advi-
sor Study: A Policy Research Project of the University of Arizona. Annie E.
Casey Foundation; 1998.
9. Meister JS, Warrick LH, de Zapién JG, et al. Using lay health work-
ers: case study of a community-based prenatal intervention. J Community
10. Castro FG, Coe K, Gutierres S, et al. Designing health promotion
programs for Latinos. In: Kato PM, Mann T, eds. Handbook of Diversity
Issues in Health Psychology: Issues of Age, Gender and Orientation, and
Ethnicity. New York, NY: Plenum Press; 1996:319-345.
11. Larkey LK, Staten LK, Ritenbaugh C. Communication strategies
used by Latina lay advocates of the Women’s Health Initiative. Women Cancer.
12. Larkey L. Las mujeres saludables: reaching Latinas for breast, cer-
vical and colorectal cancer prevention and screening. J Community Health.
13. Navarro AM, Senn KL, McNicholas LJ, et al. Por La Vida model inter-
vention enhances use of cancer screening tests among Latinas. Am J Prev
14. Castro FG, Elder J, Coe K, et al. Mobilizing churches for health pro-
motion in Latino communities: Compañeros en la Salud. J Natl Cancer Inst
15. Larkey LK, Alatorre C, Buller DB, et al. Communication strategies for
dietary change in a worksite peer educator intervention. Health Educ Res.
16. Nwoga IA. African American mothers use stories for family sexuali-
ty education. MCN Am J Matern Child Nurs. 2000;25(1):31-36.
17. Bergman P. Storytelling as a teaching tool. Clin Excell Nurse Pract.
18. Costantino G, Malgady RG, Rogler LH. Storytelling through pictures:
culturally sensitive psychotherapy for Hispanic children and adolescents. J
Clin Child Psychol. 1994;23(1):13-20.
19. Hodge FS, Pasqua A, Marquez CA, et al. Utilizing traditional story-
telling to promote wellness in American Indian communities. J Transcult Nurs.
20. Steve Denning: The Website for Business and Organizational Story-
telling. What Are the Main Types of Stories and Narratives? 2000. http://
stevedenning.com/Main_types_story.html. Accessed August 7, 2008.
21. Green MC, Garst J, Brock TC. The power of fiction: determinants and
boundaries. In: Shrum LJ, ed. The Psychology of Entertainment Media:
Blurring the Lines Between Entertainment and Persuasion. Mahwah, NJ:
Lawrence Erlbaum Associates, Inc, Publishers; 2004:161-176.
22. Rimer BK, Glassman B. Is there a use for tailored print communica-
tions in cancer risk communication? J Natl Cancer Inst Monogr. 1999;(25):
23. Janz NK, Becker MH. The Health Belief Model: a decade later.
Health Educ Q. 1984;11(1):1-47.
24. Vernon SW. Risk perception and risk communication for cancer
screening behaviors: a review. J Natl Cancer Inst Monogr. 1999;(25):101-119.
25. McGarvey EL, Clavet GJ, Johnson JB 2nd, et al. Cancer screening
practices and attitudes: comparison of low-income women in three ethnic
groups. Ethn Health. 2003;8(1):71-82.
26. Lipkus IM, Green LG, Marcus A. Manipulating perceptions of col-
orectal cancer threat: implications for screening intentions and behaviors. J
Health Commun. 2003;8(3):213-228.
27. Hopwood P, Howell A, Lalloo F, et al. Do women understand the
odds? Risk perceptions and recall of risk information in women with a fam-
ily history of breast cancer. Community Genet. 2003;6(4):214-223.
28. Royak-Shaler R, Blocker DE, Yali AM, et al. Breast and colorectal
cancer risk communication approaches with low-income African-American
and Hispanic women: implications for healthcare providers. J Natl Med
29. Woloshin S, Schwartz LM, Welsh HG. Risk charts: putting cancer in
context. J Natl Cancer Inst. 2002;94(11):799-804.
30. Gurmankin AD, Baron J, Armstrong K. The effect of numerical state-
ments of risk on trust and comfort with hypothetical physician risk communi-
cation. Med Decis Making. 2004;24(5):265-271.
31. Woloshin S, Schwartz LM, Moncur M, et al. Assessing values for
health: numeracy matters. Med Decis Making. 2001;21(5):382-390.
32. Weinstein ND, Atwood K, Puleo E, et al. Colon cancer: risk percep-
tions and risk communication. J Health Commun. 2004;9(1):53-65.
33. Lipkus IM, Skinner CS, Green LS, et al. Modifying attributions of col-
orectal cancer risk. Cancer Epidemiol Biomarkers Prev. 2004;13(4):560-566.
34. Strickland CT, Squeoch MD, Chrisman NJ. Health promotion in cer-
vical cancer prevention among the Yakama Indian women of the Wa’Shat
Longhouse. J Transcult Nurs. 1999;10(3):190-196.
35. Community PROMISE: Peers Reaching Out and Modeling Interven-
January 2009, Vol. 16, No. 1 Cancer Control 87
tion Strategies for HIV/AIDS Risk Reduction in their Community Guidelines.
Accessed August 7, 2008.
36. Williams-Brown S, Baldwin DM, Bakos A. Storytelling as a method
to teach African American women breast health information. J Cancer Educ.
37. Kirby SD, Ureda JR, Rose RL, et al. Peripheral cues and involvement
level: influences on acceptance of a mammography message. J Health
38. Levin B, Smith RA, Feldman GE, et al. Promoting early detection
tests for colorectal carcinoma and adenomatous polyps: a framework for
action. The strategic plan of the National Colorectal Cancer Roundtable.
Cancer. 2002;95(18):1618-1628. Erratum in: Cancer. 2002;95(12):2580.
39. Norat T, Riboli E. Fruit and vegetable consumption and risk of can-
cer of the digestive tract: meta-analysis of published case-control and
cohort studies. IARC Sci Publ. 2002;156:123-125.
40. Satia-Abouta J, Galanko JA, Martin CF, et al. Food groups and colon
cancer risk in African-Americans and Caucasians. Int J Cancer. 2004;
41. Burt RW, Winawer SJ, Bond JH, et al. Preventing Colorectal Cancer:
A Clinician’s Guide. The American Gastroenterological Association. 2004.
Accessed August 7, 2008.
42. Larkey LK, Colditz G, Wilson J, et al. Test of a VideoGuide for Har-
vard Cancer Risk Index in low literate population. J Biocommun. 2005;31(2).
http://www.jbiocommunication.org/archives.html. Accessed August 7, 2008.
43. Marin G, Marin BV. Research With Hispanic Populations. Newbury
Park, CA: Sage Publications Inc; 1991.
44. Teram E, Schachter CL, Stalker CA. The case for integrating ground-
ed theory and participatory action research: empowering clients to inform
professional practice. Qual Health Res. 2005;15(8):1129-1140.
45. Fairfield KM, Chen WY, Colditz GA, et al. Colon cancer risk counseling
by health-care providers: perceived barriers and response to an internet-
based cancer risk appraisal instrument. J Cancer Educ. 2004;19(2):95-97.
46. Champion VL, Skinner CS, Menon U, et al. A breast cancer fear
scale: psychometric development. J Health Psychol. 2004;9(6):753-762.
47. Centers for Disease Control and Prevention (CDC). Behavioral Risk
Factor Surveillance System Survey Questionnaire. Atlanta, GA: US Depart-
ment of Health and Human Services, Centers for Disease Control and Pre-
48. Staten LK, Taren DL, Howell WH, et al. Validation of the Arizona Activ-
ity Frequency Questionnaire using doubly labeled water. Med Sci Sports
49. Green SB, Salkind NJ. Using SPSS for Windows and Macintosh:
Analyzing and Understanding Data. 4th ed. Upper Saddle River, NJ: Pearson/
Prentice Hall; 2005.
50. Green MC, Brock TC, Kaufman GF. Understanding media enjoy-
ment: the role of transportation into narrative worlds. Commun Theory.
51. Slater MD, Rouner D. Entertainment-education and elaboration-like-
lihood: understanding the processing of narrative persuasion. Commun
52. Larkey LK, Gonzalez J. Storytelling for promoting colorectal cancer
prevention and early detection among Latinos. Patient Educ Couns. 2007;
67(3):272-278. Epub 2007 May 23.
53. Ramirez AG, Gallion KJ, Suarez L, et al. A national agenda for Latino
cancer prevention and control. Cancer. 2005;103(11):2209-2215.