Journal of Nursing Research VOL. 17. MO.

4, DECEMBER 2009

A Tailored Web-Based Intervention to

Promote Women's Perceptions of and
Intentions for Mammography
Zu-Chun Lin Shu-Fang Wang

ABSTRACT
Background: The Internet is rapidly becoming an accepted
tandard for disseminating and obtaining health information.
However, health information presented on the Web frequently
is neither tailored to patients' characteristics nor theoretically
driven. This study used Prochaska's transtheoretical model to
guide the design and evaluation of a tailored Web-based mammography educational program.
Purpose: This study used a pretest-posttest design to examine the impact of a tailored Web-based educational interveniion on women's perceptions of and intentions to obtain a
mammography. We hypothesized that a compiete tailored intervention (CTI) would result in a significantly higher score than
that of tailored message intervention (TMI) in terms of women's
perceptions of mammography and intentions to receive a
mammography. Both TMI and CTI have been found to have
significantly higher scores in terms of women's perceptions
of mammography and intentions to receive a mammography
than those of the standard intervention (SI),
Methods: One hundred eighty-five women were assigned
landomly to one of three groups: CTI (n = 61}. TMI (n = 63), or
SI (n = 61). Online questionnaires were completed by participants at baseline and after completion of intervention.
Results: Study results supported the hypothesis of percepiions of mammography. The CTI group had significantly higher
scores on intention to obtain a mammogram than those of
the Si group. However, the differences between CTI and TMI
groups and between TMI and SI groups were not significant.
Conctusions and Implications for Practice: Results of the
study validated the importance of utilizing theory for tailored
intervention design to enhance individuals' positive perceptions of health behaviors. Additional studies should consider
the effect of various processes and activities on women at
different stages of acceptance of the need to obtain a mammogram to further determine outcomes and apply such to
mammography screening behavior.
KEY WORDS:
tailoted intervention, Web-based, transtheoretical model,
mammography, screening.

Introduction
Breast cancer had the highest rate of morbidity among cancers in women in 2007 and ranks as the fourth most sig-

nificant cause of death among Taiwanese women. Breast

cancer is prevalent between the ages of 45 and 5S years in
Taiwan, and age ar diagnosis has been decreasing in recent
years (Department of Health, Executive Yuan, Tiiiwan, ROC,
2008). Increased use of screenin^i mammography has contributed to a decrease in breast cancer mortaiit). Early detection can reduce breast cancer mortality from 20% to 4.5%
(Department of Health, Executive Yuan, Taiwan, ROC, 2008;
Duffy et al., 2002). The effective use i>f mammography in
the United States has helped identify 56.4% of cases of
hreast cancer in the early stages (i.e., stage 0 or 1). However, only 35.6"/ii of Taiwan's hreast cancer cases are detected
in early stages {Department of Health, Executive Yuan,
Taiwan, ROC, 2008).
Despite the effectiveness of mammography for breast
cancer detection, the ratio of women between 50 and 69
years in Taiwan who have had at least one mammography
in the past 2 years remains low (17.8%) in comparison with
75% found in the United States and United Kingdom (Department of Health, Executive Yuan, Taiwan, ROC-, 2008).
This would suggest that Taiwanese women need to he
educated regarding breast cancer risks and the benefits of
screening mammography. Alrhcmgh mammography is useful and effective for breast cancer detection, it is not recommended for Taiwanese women before the age of 50 years or
for those with high-density hreasts. Sontigraphy is recommended for this latter group (Chang, Kuo, s: Wang, 2008).
1 he increasing incidence of breast cancer in younger age
groups in Taiwan has helped increase mammography and sonography utilization in hreast cancer screening (Chang et al.,
2008). Women who are at high risk for hreast cancer or who
have been identified as having abnormal hreast tissue are encouraged to seek professional help. Once a mammography
has been recommended by a health professional, cost for the
procedure may he reimbursed through the National Heatth Insurance program of Taiwan, regardless of the woman's age.
RN, PhD, Associate Professor, Department of Nursing, Tzu-Chi
College of Technology.
Received: April 2, 2009 Revised: July 15, 2009
Accepted: August 20, 2009
Address correspondence to: Zu-Chun l.in. No. 880, Chien-Kuo Rd.
Sec. 2, Hualien 97005, Taiwan, ROC. Tel: +886 (3) 857-2158
ext. 414; Fax: +886 {3) 857-7962.
E-mail: etaine@tccn.edu.rw; zuchulin.elaine@gtnail.com

249

loumal of Nursing Research

Although early diagnosis is a key to reducing breast cancer mortality, several factors are associated with low or
irregular mammography utilization, including inadequate
knowledge about breast cancer or screening guidelines, financial and/or insurance concerns, education background,
place of residence, inaccessibility of health facilities, personal beliefs, or social and cultural determinants (Blackman
& Masi, 2006). Misperception s about mammography as
well as prior unfavorable experiences with mammography,
fear of radiation or pain, embarrassment, inconvenient scheduling, or long waiting times in outpatient departments also
have been reported as factors that contribute to mammography nonadherence (Han, William, & Harrison, 2001; Liao,
2003). Even when inventive and resourceful interventions
are given, such as offering screening free of charge, providing medical vans, and distributing printed promotional materials, adherence rates still remain low. This illustrates the
need to reexamine current available health education and
interventions offered and delivered to women.

Applying Tailored and Theory-Based Approach

in Educational Intervention
Internet technology is reshaping the ways in which health
education is delivered. However, Web-based education is
typically neither tailored to individual needs nor theoretically driven (Evers, Prochaska, Prochaska, Cummins, &
Velicer, 2003; Lauver et al., 2002). Currently, available Webbased interventions are more likely provide information in
"one size fits all" portions, similar to a "brochure" format
(Lauver et al., 2002). This uniform approach is relatively
cost-effective in terms of providing general information about
mammography screening to the public. However, this approach may miss opportunities to provide culturally sensitive and/or customized information that may be critical to
the individual decision-making process. Studies have shown
greater promise to increase mammography adherence in
providing tailored interventions rather than generalized care
(Allen &: Bazargan-Hejazi, 2005; Champion et al., 2002;
Rimer et al., 2002).
Theory guides both current and future understanding of
health behaviors and offers directions for research and suggestions for intervention development (Redding, Rossi, Rossi,
Velicer, &c Prochaska, 2000). The transtheoretical model
(TTM; Prochaska & DiClemente, 1982) is one of the most
widely used behavior theories and has been tested with a
variety of patient interventions, including smoking cessation, sun protection, mammography screening, and physical
activity. Yet, studies utilizing TTM concepts have failed to
consider the full scope of the theory and only addressed
TTM concepts related to change stage and/or decisional
balance. Bridle et al. (2005) suggested that incomplete
model application may lead to inappropriate intervention
design and unsatisfactory outcomes such as inefficient use
of financial resources with little to no increase in knowledge
or motivation to act on health behaviors. A theory-based
250

Zu-Chun Lin et al.

tailored intervention is essential to maximize health intervention outcomes.

Overview of TTM
The TTM was originally developed by Prochaska Ik
DiClemente (1982) and tested in the context of smoking
cessation interventions (Prochaska et al., 2005). T1"M assumes that behavioral change is influenced by an individual's readiness, beliefs, and confidence in his or her ability
to make a change, as well as strategies used to progress
through change stages. TTM encompasses four major components, including stage i)f change (also known as stage of
adoption), processes of change, decisional balance, and selfefficacy (Rakowski, Dube, & Goldstein, 1996; Redding et al.,
2000). Stage of cbange or adoption describes an individual's
current behavior and intention to change that behavior.
Rakovvski, Dube, et al. (1996) found that mammography
screening and smoking cessation behaviors have different
charaaeristics, specifically adoption versus addiction and
detection oriented versus prevention oriented, and thus
redefined stage of change as stage of adoption. This stage includes p recontempla tion, relapse, risk for relapse, contemplation, action, and maintenance (Rakowski, Dube, et al., 1996).
Processes of change represent the covert and overt activities that an individual uses to proceed to the next stage.
Specific techniques must be incorporated to provide stagematched content to facilitate processes of change and meet
stage-specific goals (DiCiementc, 200.^). Prwhaska and Velicer
(1997) suggested consciousness raising, dramatic relief, and
environmental rvaluation processes of change to move
women beyond the precontemplation stage (i.e., never having
contemplated mammography and not planning to undergo
one in the coming 1- to 2-year period) to the contemplation stage (i.e., never having undergone a mammography
but intending to schedule one in the coming 1 to 2 years;
Rakowski, Dube, et al., 1996). Recommended techniques
used to raise consciousness include feedback and education.
Personal testimony is suggested as providing dramatic relief, whereas role modeling helps facilitate environmental
rvaluation (Pnichaska c Velicer, 1997).
Decisional balance assesses the reasons for (pro) and
against (con) behavioral change. Positive decisional balance
(more pros than cons) occurs more frequently as an individual moves through the stages of adoption (Rakowski,
Dube, et al., 1996). Self-efficacy describes the level of an
individual's confidence and inclination toward behavioral
change {Prochaska &: DiClemente, 1982; Rakowski,
Dube, et al., 1996; Redding et al., 2000). Self-efficacy is
deemed less appropriate in adoptive than in addictive (i.e.,
smoking cessation) behaviors and is thus usually excluded in
mammography adoption research.

Tailored Intervention and Outcomes

Tailored interventions arc those that are ik'veloped for a
specific recipient based on that individual's personal

Tailored Web-Based Intervention

VOL. 17, NO. A. DECEMBER 2009

charaaeristics such as knowledge, skills, beliefs, or intention

to engage in a recommended health behavior (Kreuter &
Wray, 200.^). Tailored interventions are distinct from personalized and targeted interventions due to the increasing
complexity and specificity of interventions. A personalized
intervention can be as simple as sending a generic letter or
e-mnil to an individual using his or her name. A targeted
intervention is used to reach a group of people who share
similar characteristics such as age, gender, or ethnicity
(Krcutcr, Straher, & Cilassman, 1999; Kreuter & Wray, 200.^}.
Targeted interventions are used extensively in community
health (l-auver et al-, 2002). Tailored interventions have the
potential for greater advantages over hoth personalized and
targeted interventions as they are able to provide a greater
level of customization to address variability' among targeted
individuals (Kreuter ik Wray, 2003).

be attributable to the longer times necessary to receive tailored printed messages after assessments and inadequate
application of behavioral theories. The inconsistent effectiveness of tailored interventions shown in studies may
result from the behavior theory concept levels utilized in
tailored interventions. Given the high rate of breast cancer,
clear need for earlier detection in Taiwanese women, and
inadequate emphasis and application of theory in health
educational Web sites, we investigated whether Web-based
mammography education using tailored instructional interventions based on comprehensive TTM concepts could increase women's perceptions of and intentions to obtain a
mammography.

The outcomes of tailored interventions in promoting maminography screening or other heath behaviors are mixed.
Prochaska et al. (2005) found that tailored interventions
based upon stage change assessment results, decisional balance, change processes, and efficacy level produced higher
adherence to mammogriiphy (PrcKhaska et al., 2005). Women
in the action stage for mammography were given materials
based on adoption stage and TTM constructs. They were
more likely to obtain repeat screening mammography than
were those who used no intervention but did not differ
significantly from the standard material group (Clark et al.,
2002).
Use of a computer-tailored message corresponding to
individual characteristics was compared with sending general information and pursuing no intervention in terms of
impact on intention and behavior with regard to early
cancer detection. It was found that the computer-tailored
intervention had positive effects on passive detection and
help-seeking intentions and behaviors in the short-term
hut not in the long-term (de Nooiier, Lenchner, Candel, &
Vries, 2002). African American women who received both
behavioral construct tailoring (i.e., knowledge, beliefs, perceived barriers, and stage of readiness) and culturally relevant tailoring (i.e., religiosity, colleaivism, and racial pride)
were more likely to receive a mammography than were
those subjected to behavioral construct tailoring, culturally
relevant tailoring, or normal care groups (Kreuter et al.,
2005). Participants in the three intervention groups (i.e.,
tailoretl telephone vs. tailored mail vs. tailored telephone
and mail) had significantly l^etter mammography adherence
rates than the rates of those in the control group (Saywell,
(Champion, Skinner, Menon, &: Oaggy, 2004).
Mler-Riemenschneider, Reinhold, Nacon, and Willich
(2008) conducted a systematic review of the effectiveness
of tailored interventions in prom<jting physical activity. Results did not provide strong evidence for such. Rakowski
ct al. (2003) found that nontailored materials (i.e., reminder)
performed as well as or better than tailored communications
in mammography adherence. The finding that tailored interventions were not more efficacious than regular care may

The purpose of this study was to develop and evaluate the

effects of three types of mammogram adherence interventions; complete tailored intervention (CTI) versus tailored
message inten'ention (TM!) versus standard inter\'ention (SI).
The CTI intervention was guided by major "ITM concepts:
stage of adoption, decisiona! balance and prcKesses of change.
The hypotheses that were tested were the following: (a) The
CTI would result in significantly higher scores in women's
perceptions of mammography and intentions to receive mammography than would TMI, (b) TMI would result in significantly higher scores in women's perceptions of mammography
and intentions to receive mammography than would SI, and
(c) CTI would result in significant differences in women's
perceptions of mammography and intentions to receive mammography than would SI.

Purpose

Methods
Sample and Setting
To address the generally low rate of mammography acceptance in Taiwan, interventions were targeted on those who
expressed no intention to receive mammography within the
next 1- to 2-year period (precontemplator). The sample comprised 202 precontemplators between ^S and 69 years
who had no history of breast cancer, had Internet access at
home or their place of employment, and had experience
using computers.
A convenience sample was recruited from Taiwan counties by distributing recruitment flyers by hand and e-mail.
A snowballing technique was then employed to expand
recruimient to a larger pool of potential participants. Participants accessed the designated Web site either at home or
their place of employment. Participant pr(tection approval
for the research was obtained from the Tzu-C;hi Hospital
Institutional Review Board (IRB097-36).

Design
The study used a pretest-posttest design. Data were collected at baseline and immediately after completion of the
251

Joumal of Nursing Research

Zu-Chun Lin et al.

TABLE 1.

Exampies of Tailored Messages

Types of Responses and Corresponding
trongly Disagree

12

Disagree

Undecided

Mammography is
not a useful procedure
for women my age.

Correct !
Breast cancer is prevalent
between the ages of 35
to 55 in Taiwan. Recently,
the onset age has been
falling. Therefore, women
at any age should be caution.

Mammography is
necessary even when
there is no history of
breast problems in
my family.

Caution!
Many factors can increase your risk of breast
cancer. Some (e.g., diet, life style, and exposure
to radiation) can be controlled or avoided, but your
family history can't be avoided. Women can help
protect themselves by staying away from known
risk factors whenever possible, Women with a
family history of breast cancer are 5 times more
likely to have breast cancer than those without a
family history of the disease. But even without a
family history of breast cancer, it is possible to get
breast cancer. Hence, having mammograms are
still necessary.

gly Agree

Caution!
Breast cancer threatens women's health
worldwide. It may occur at any age.
In Taiwan, breast cancer is most
commonly seen in women between
the ages of 35 and 55, with a peak
incidence between the ages of 40 to
50. The incidence rate drops after age
55. but this doesn't mean worrien after
menopause have zero chance of having
breast cancer. This shows that women
at any age are threatened by breast cancer
and should have regular mammograms for
preventive detection.
Correct!
Family history is only one
of the risk factors. Having
no family history of breast
cancer doesn't guarantee
that you will not have
breast cancer.

Note DBMI = Decisional Balance for Mammography Inventory.

educational intervention. Participants were assigned randomly to one of three groups., designated as CTI, TMI, and
SI. The CTI group received a variety of educational program tailored to the precontemplation stage of mammography adoption. Specific educational programs were developed
based upon TTM's change processes (i.e., consciousness raising, dramatic relief, and environmental rvaluation) and
used recommended techniques for each process of change
(education, feedback, personal testimonies, and role modeling). The TMI group obtained tailored messages only. The
SI group received breast cancer and mammography brochures. A similar intervention component design has been
previously tested {Lin, Effken, Li, &: Kuo, in press).

Measurement
Questionnaires were modified from ones used in previous
studies, including (a) a 5-point Likert scale 15-item Decisional Balance for Mammography Inventory (DBMI) designed to assess women's perceptions of mammography. Of
252

the 15 items, 6 measured favorable and 9 measured unfavorable perceptions of mammography (Rakowski et al.,
1997). (b) A dichotomous one-item Stage of Adoption for
Mammography Questionnaire (SAMQ) assessed women's
intention to obtain a mammography (Rnkiiwski, Diibc, ct al.,
1996). (c) The Demographics Inventory inclutk'd 12 items
that colleaed infonnation regarding participant age, ducation,
employment status, and previous computer experience. Both
SAMQ and DBMI had demonstrated adequate validity
(.43 to .82 for the pros index and .69 to .77 for the cons
index as examined using principal components analysis
and reliability (Cronbach's a =.77 .80 for pro and .78-.8.
for con items) in previous studies (Rakowski, Ehrich, et al.,
1996; Rakowski et al., 1997). SAMQ and DBMI were translated into Chinese by the first author and back-translated
into English by a bilingual expert ro ensure translation
accuracy. A panel of experts comprised of three nursing
instructors with breast cancer expertise verified instrument
content validity. One expert suggested adding two items
("I am afraid to receive mammography because it is painful" and "1 am afraid to receive mnmmography because of

VOL. 17, NO. 4, DECEMBER 2009

Tailored Web-Based Intervention

a role-modeling activity in a tailored intervention. Feedback

was provided through a computerized assessment tool that
used women's perceptions of the pros and cons of mammography to create user-tailored messages. As participants completed the DBMI, they received computer-generated messages
tailored to their responses. Table 1 illustrates examples of such
Procedures
messages.
Parricipanrs all accessed the same Weh site URL and were
Personal testimonies consisted of interviews with two
asked several questions prior to the interventions, includbreast cancer survivors and a woman who went for reguing intent to have a mammography, breast cancer history,
lar mammography. Role modeling utilized women's diantammography history, and so on to confirm eligibility. Etilogues. Issues related to personality (i.e., embarrassment)
jiible participants then were randomly assigned to the CTl,
and conditions specific to Taiwan (i.e., medical fees, prevaTMI, or SI group by the computer program as follows:
lence of breast cancer in younger women, and scheduling)
The first eligible visitor was assigned to the CTI group; the
were addressed here. Such content was mandatory for
second one, to the TMI; and the third, to the Si group. The
participants to review. Two additional interventions used
sequence was repeated from the fourth participant onward.
education techniques. These consisted ot a minilecture on
Completing the SI, TMI, and CTI took approximately 15
breast cancer and demonstration of a mammography proto ?>{) minutes. To avoid participants skipping educational cedure. Figures 1 and 2 are screen shots of CTI.
materials, required times on tasks in each section were set.
HBMI and SAMQ were administered at baseline, and DBMI
Tailored message intervention
and SAMQ were again, along with Demographics Inventory,
Tailored
message intervention was designed based upon
assessed online immediately upon completion of the educaTTM's principle of feedback. However, it did not incortional intervention. Survey responses were saved in a dataporate techniques or activities associated with the change
base on the Web site and could be downloaded for analysis.
process such as education, personal testimony, and role modeling. The TMI group received computer-generated messages
Complete tailored intervention
tailored to their DBMI responses.
lo motivate precontemplators to bectinie conteniplators,
ITM's techniques and activities incorporated a consciousnessStandard intervention
raising tecbniqtie that applied a feedback or education activThe SI group was also asked to respond to the DBMI but
ity, dramatic relief technique utilizing a personal testimony
activity, and an environmental rvaluation technique using
received only educational brochures on breast cancer and
radiation exposure") to the DBMI. The content validity
index for SAMQ was 1.0 and .87 for DBMI. Cronhach's a
was .82 for the 6-item pros suhscale and .84 for the 9-item
cons suhscale. The overall reliability ofthe DBMI was .87.

The dialogue shows Mrs. Li serving ;is a

role model lo convince Mn;. Wang to
receive m;immiijir;ipliy.

c;ise remain m Mu-' ^L\^on ,i

iree and half minutes before

proceeding to the next section.

*
Figure 1. Role modeling.

253

Journal of Nursing Research

Zu-Chun Lin et al,

I his setlitin demonstrates mammogtaphy

procedures to decrease women's anxiety
Linii increase iheir motivation for

'1

mamniDgraphy

Figure 2. Demonstration of mammography procedures.

tnammography developed by the S.Y. Dao Memorial Fund
and the Bureau of Health Promotion, Department of Health
(see Figure 3). These materials were saved in PDF format and
then posted to the research Web site.

Data Analysis
SPSS Version 15 (SPSS Inc, Chicago, TL) was used for data
analysis. Con items in the DBMI were reverse coded. Potential DBMI scores ranged from 15 to 75, with higher
scores representing more positive perceptions of mammography. A chi-square test was performed to assess the
equivalence of groups in terms of demographic characteristics. Analyses of covariances and chi-square tests were
used to analyze the answers to the questions.

Results
A total of 185 participants (61 CTI, 63 TMI, and 61 SI)
completed the questionnaire, for a 91.6% response rate.
There were no statistically significant differences {p > .05)
in demographics among the groups (Table 2). Hence, no
confounding factors (covariates) were present to potentiallv interact with intervention effects.
254

CTI earned a higher mean postintervention score

(M = 60.56, .SD = 6.26) than that of both TMI (M = 57.95,
SD = 7.79) and SI (M = 53.26, SD = 8.58). When baseline
perception was controlled, analysis of covariance revealed
that the interventions had a significantly different impact
on women's perceptions, F(2,179) = 4.902, p < .05. The
CTI group had higher mean ptsttest scores (M = 60.09,
SE = 1.16, 95% confidence interval [CIl = 56.92-61.42)
than those of the TMI group (M = 57.60, SH = 0.88, 95%
CI = 55.85-59.31) and SI group (M = 55.09, SE = 0.82,
95% CI = 53.45-56.63). Posteriori comparisons showed
that the CTI produced results that differed significantly
from those of TMI (mean difference =2.49,5/) = 1.73, 95%
CI = 1.82-5.00, p < .05), TMI achieved results that differed
significantly from those of S! (M = 2.51, SD = 1.03, 9 5 %
CI = 0 . 5 0 ^ . 5 8 , p < .05), and CTI generated differences
that were significantly different from those of SI [M =5.0,
SD = 1.61, 95% CI = 0.95-7.30, p < .05) in terms of an
increase in positive perceptions (Table 3).
A chi-square test showed that intervention type had a
significant impact on intention to receive mammography
{df= 2, r = 12.909, p < .05). Haberman (1978) suggested
using adjusted residual (AR) 1.96 as the critical value to

VOL 17, NO. 4, DECEMBER 2009

Tailored Web-Based Intervention

(ti

Understanding

na

SS
n
9

Si

-U

is
8

's

i*

Mammography

I
Why do Breasts Need
U) be Compressed?

Preparation for
Mammography

Is Mammography a
Safe Exam?

Bit

Proceed
Figure 3. Mammography brochure.
tk-rermine significance between the groups in posteriori
comparison. Posteriori comparisons showed that CTI (28.1%,
AR = 3.7) had a significantly higher intention compared with
that of SI {15.7%, AR = -4.0). However, the difference
hetween CTI (28.1%) and TMI (2.3.2%) was not significant,
as was the difference between TMI (23.2%) and SI (15.7%)
in terms of intention to have a mammography {Table 4).

Discussion
Healthcare has shifted its focus over time from acute and
i-hronic diseases to health promotion and the need to understand personal, social, and cultural determinants of
liealth. Tailored interventions have been introduced to
support the notion of individualized care. TTM has gained
widespread popularity, yet little is known about its effectiveness as a basis for Web-based tailored educational inrcrventinn. Results of this study support the hypotheses
that C ri is more powerful than TMI and that TMi is more
powerful than SI in positively changing women's perceptions regarding mammography. Findings were somewhat
consistent with previous studies on mammography and
health-related behavior promotion (Clark et al., 2002; Dutton,

Provost, Tan, & Smith, 2008; Lin, 2008; Prochaska et al.,

2005; Saywell et al., 2004).
The emphasis of major components of tailored CTI intervention may be one of the factors contributing to increasingly positive perceptions of mammography. First, the
CTI included an assessment of key characteristics of each
person, stage of adoption for mammography screening (intention), and decisional balance for mammography (perception).
Second, a decision algorithm was used to match a particular message to an individual's characteristics. Third, educational content was organized in small segments designed
to match these characteristics. Also, a variety of content designs (e.g., dialogue between two women and exploring
mammography procedures) based on TTM's change processes {i.e., role modeling and education) were used. Finally,
Internet and multimedia teclinology were used in the various
tailored intervention designs to provide more flexible, efficient, diverse, and attractive health education than TMI
and SI to increase the woman's intention to participate in
the intervention. Although TMI lacks the direct social presence that one may get from direct, face-to-face interventions
by health professionals, it still mimics to some extent classic
interpersonal counseling via an expert system (Oenema, Brug,
255

Journal of Nursing Research

2u-Chun Lin ei al.

TABLE 2.

Sample Demographics by Group

LCTI (n = 611
\ variables

Si{n

B61)

r%

^^^^^^^^^^^^^^^^^H

Age (years)
35-15
46-55
Over 65
Ethnicity
Taiwanese (Holo)
Mainlander
Other (Hakka, aboriginal)
Education
Junior high
High school
College
Postcollege
Missing
Marital status
Married
Other (single, divorced)
Children
0-2
>2
Work status
Full-time
Housewife
Other (part-time, unemployed)
Income (NTD)
400,000
400,001 -600,000
600,001-800,000
800,001-1000,000
More than 1000,000
Missing
Average hours spent per day on the computer
Less than 3
3-5
6-fi
Over 8
Computer experience (years)
Less than 3
3-5
6-8
Over 8
Experience searching online health information
Yes
No
Bandwidth
High speed
Dial-up
Computer access
Yes
No
Missing

TMI (ne 63)

^

,065

40
16
5

65.6
26.2
8.2

50
8
5

79,4
12,7
7,9

50
6
5

81.9
9.9
8.2

49
7
5

80,3
11.5
8.2

46
9
8

73.0
14.3
12.7

49
9
3

80,3
14,8
4,9

18
10
11
22

29.5
16.4
18.0
36.1

9
9
18
27

14.3
14.3
28,9
42,9

8
7
27
18
1

13.2
11,5
44.3
29.5
1.5

52
9

85,2
14,8

56
7

88,9
11.1

52
9

85.2
14,8

43
18

70.5
29.5

52
11

82.5
17.5

48
13

78.6
21.4

27
30
4

44.3
49.2
6.6

38
21
4

60,3
6,3
33.3

31
5
25

50,8
8,2
41,0

20
21
8
7
5

32,8
34,4
13.1
11.5
8.2

22
13
12
5
10
1

34,9
20.6
19,0
7.9
15,9
1.6

14

23,0
29,5
21,3
13,1
13,1

31
15
8
7

50,8
24,6
13.1
11.5

22
16
15
10

34,9
25,4
23,8
15,9

22
12
17
10

36,1
19.7
27.9
16,4

22
5
7
27

36.1
8.2
11.5
44,3

16
6
7
34

25,4
9.5
11,1
54,0

8
9
S
36

13,1
14,8
13.1
69.0

47
14

77.0
23,0

54
9

85,7
14,3

53
8

86,9
13,1

57
4

93.4
6.6

60
3

95,2
4,8

59
2

96,7
3,3

49
12

80,3
19,7

55
7

87,3
11.1
1.6

54
7

88,5
11,5

366

,112

,793

.267

,181

,345
18
13
8
8

.091

,063

.285

.705

.319

Note CTI = complete tailored intervention; TMI = tailored message (nterveniion; SI = standard intervention, NTD = New Taiwan Dolldt.

256

VOL. 17, NO 4, DECEMBER 2009

Tailored Web-Based Intervention

TABLE 3.

Comparison of Mean Scores Among the Three Groups

Group
CTI
TMI
SI

mm

n
61
63
61

60.5G
57.95
53.26

1, Adjusted Mean
El

26

GO 09

7.79
8.58

57.60
55,09

1.16
0.88
0.82

4.902

CTI - TMI > SI

/Vole. CTI = complota tailored interventton; TMI = lailoied message imervention; SI = standard intervention.
'Analysis of covariance used the baseline score as the covariate.
p< .05.

s: Ix'chner, 2001 ). This niiiy explain why TMI resulted in

Limitations and Suggestions for Future
higher positive perceptions of mammography than did SI.
Research
CTI performed significantly better than did SI. However, C'TI was not significantly superior to TMI and TMI Limitations of the study include the following: Participants
were all from Taiwan, mostly from Hualien and Taipei
was not better than SI in terms of promoting women to
counties,
which limits sttidy generaiizability. Self-reporting
move from the preconremplation to the contemplation stage.
was
used
to collect participant intentions related to obtainReasons to explain these effects include the following: First.,
ing
mammography
screening. Although self-reporting is
there may be some women who were already located in n
generally
considered
accurate, it cannot ascertain whether
transition state between the two stages. The stage of adopor
not
participants
actually
obtain mammography screening.
tion or change ust-d in this study may oversimplify the
A
follow-up
study
would
be needed to evaluate whether
complexities of behavioral change by imposing artificial
participants
actually
acted
on stated intentions. One SI
categories on a continuous process (Littel! &C Girvin, 2002).
"dosage"
was
administered.
However, differences in dosSecond, perceptions, intentions, and behavioral adoptions
age
are
needed
to
reflect
truly
tailored education.
may fluctuate independent of one another (Littell c Girvin,
Stage
of
adoption
or
change
has been defined and mea2002). This implies that a continuously measured stage of
sured
in
a
variety
ways.
Stage
algorithms
(mostly "yes" or
;id<>ption or change may reflect a woman's intention level
"no"
answers)
are
based
on
either
past
behavior
or arbimore accurately. Third, existing studies have not discussed
trary
time
frames
(i.e.,
3
months,
6
month,
and
1 to 2
which prix:esses of change are associated with progress
years),
which
are
inconsistent
across
studies.
Any
shift
in
across stages for different health behavioral problt-ms and
such
time
frames
would
alter
the
distribution
of
people
which processes of change are more successful than are
across stages. A consistent definition of stage of change or
others in terms ofa particular health behavior. Our protoadoption and measurement using continuous items is rectype of a comprehensive tailored intervention requires more
ommended
to generate more accurate results. Future studies
extensive testing to validate outcomes. Last, tailored intermay
also
compare
the effectiveness (i.e., stage of adoption
ventions should be repeated until the individual achieves or
and
decisional
balance
for mammography) of tailored inmaintains desired behavioral change. In that way, interterventions
for
mammography
screening for women at
ventions evolve and adapt in response to an individual's
different
stages,
particularly
in
terms
of risk for relapse and
characteristics (Bridle et al., 2005). Our tailored intervenrelapse,
as
these
two
stages
are
less
studied. Internet acand TMI) were given at one point in time and, as
tions
counts for a small percentage of interventions were used as
such, were static, incomplete, and only partially tailored.
the delivered method. More Web-based research is needed
to verify the efficacy of theory-based tailored interventions.
TABLE 4.

Number of Participants Intending to

Receive Mammography by Group
(Postintervention)

No.
% of total
AR

52
28.1
3.7

43
23.2
0.3

29
15.7
-4.0

12.909'

Noie CTI = complete tailored inter/emion: TMI = tailored messagu

intervention; SI = standard intervention; AR = adjusted residual.
p<.05.

Implications and Conclusions

Nurses have always been at the forefront of patient education. However, nurses and healthcare providers in busy
practical settings seldom have time to educate their patients
on disease management or health promotion issues. A Webbased tailored program can be used as a powerful and
supportive educational tool to fulfill their obligations in this
area and provide continuous, customized care. We tested
the ability ofa theoretically Web-based tailored intervention
to meet the needs of an increasing population seeking health
information on the Internet and to address flaws in advice
257

Journal of Nursing Research

given on contemporar)- health education Web sites. Our

findings partially supported the use of comprehensive concepts of TTM in tailored intervention design. Future tailored
educational Web sites are recommended to take the strengths
and weakness identified in this paper to design more individualized and resourceful programs.

Zu-Chun Lin et al,

lored printed-based physical activity intervention for patient

with type 2 diabetes, Preventive Medicine, 47(4), 409-411.
Evers, K. E., Prochaska, J. M., Prochaska, M. M. D., Cummins,
C. 0., & Velicer, W. F, (2003), Strengths and weakness of
health behavior change programs on the Internet. Journal
of Health Psychology. S(l), 63 70,
Haberman, S. J. (1978). Analysis of qualitative data. New York:
Academic Press.

Acknowledgments
The study was supported by Tzu-Chi College of Technology (TCCT-961A02). The authors thank volunteers who
participated in the study.

References

Han, Y., William, R. D., & Harrison, R. (2000), Breast cancer

screening knowledge, attitudes, and practice among Korean
American women. Oncology Nursing Forum. 27(10), 1585 1591.
Kreuter, M. W., Strecher, V. J., & Glassman, B. (1999). One size
does not fit all: The case for tailoring print materials. Annals
of Behavioral Medicine. 27(4), 276 283,

Allen, B., & Bazargan-Hejazi, S. (2005). Evaluating a tailored

intervention to increase screening mammography in an
urban area. Journal of the National Medical Association,
57(10), 1350-1360.

Kreuter, M, W., Sugg-Skinner, C, Holt, C. L, Clark, E. M,, HaireJoshu, D., Fu, 0., et al. (2005). Cultural tailoring for mammography and fruit and vegetable intake among low-income
African-American women in urban public health centers. Preventive Medicine. 47(1), 53-62.

Blackman, D. J., & Masi, C. M. (2006). Racial and ethnic

disparities in breast cancer mortaiity: Are we doing enough
to address the root causes? Journal of Clinical Oncology,
24(14). 2170-2178.

Kreuter, M, W,, & Wray, R. J. (2003), Tailored and targeted

health communication: Strategies for enhancing information
relevance. American Journal of Health Behavior, 27(Suppl,),
S227 S232.

Bridle, C, Riemsma, R. P., Pattenden, J., Sowden, A. J.,

Mather, L, & Watt, I, S., et al. (2005). Systematic review of
the effectiveness of health behavior interventions based on
the transtheoretical model. Psychology and Health. 20(3).
283-301.

Lauver, D. R., Ward, S. E., Heidrich, S. M., Keller, M. L, Bowers, B.,

Brennan, P, F., et al. (2002). Patient-centered interventions.
Research in Nursing & Health, 25(4), 246- 255.

Champion, V. L., Skinner, C. S., Menon, U., Seshadri, R.,

Aznalone, D. C, & Rawl, S. M. (2002). Comparisons of
tailored mammography interventions at two months post
intervention. Annais of Behavioral Medicine, 24(3), 211-218.
Chang, K. J., Kuo, W. H., & Wang, M. Y. (2008). The epidemiology of breast cancer in Taiwan. Journal of Oncology
Society, 24{2], 85 93. (Original work published in Chinese)
Clark, M. A., Rakowski, W., Ehrich, B., Rimer, B. K., Velicer,
W. F., Dube, C. E., et al. (2002). The effect of a stage-matched
and tailored intervention on repeat mammography. American
Journal of Preventive Medicine, 22{'\). 1-7.
de Nooijer, J., Lenchner, L, Candel, M., & de Vries, H. (2002). A
randomized controlled study of short-term and long-term
effects of tailored information versus general information
on intention and behavior related to early detection of cancer. Cancer Epidemiology, Biomarkers & Prevention. 77(11 ),
1489-1491,

Liao, M. C. (2003), Research on factors affecting acceptance of

mammography by female patients. Unpublished master's
thesis, Kaohsiung Medical University, Taiwan, ROC. (Original
work published in Chinese)
Lin, Z. C, 12008). Effects of a tailored Web-based educational
intervention on Taiwanese women's mammography-related
perceptions and intention. Unpublished doctoral dissertation.
University of Arizona,
Lin, Z. C, Effken, J. A., Li, Y. J., & Kuo, C. H. (in press).
Designing a tailored Web-based educational mammography program. Computers, Informatics, Nursing.
Littell, J. H., & Girvin, H. (2002). Stage of change: A critique.
Behavior Modification, 26(2), 223 273,
Mller-Riemenschneider, F., Reinhold, T., Nacon, M., & Willich,
S. N. (2008). Long-term effectiveness of interventions
promoting physical activity: A systematic review. Preventive
Medicine. 47(4), 354-368.

Oenema, A., Brug, J., & Lechner, L (2001). Web-based tailored

nutrition education: Results of a randomized controlled trial.
Department of Health, Executive Yuan, Taiwan, ROC. (2008,
Health Education Research, 7616), 647 660.
October 15). Breast cancer. Women's leading cancer. Three
steps to prevent breast cancer. Retrieved May 3, 2009, from Prochaska, J. O., & DiClemente, C. C. (1982). Transtheoretical
http://www.doh.gov.tw/CHT2006/DM/SEARCH_RESULT.aspx.
therapy: Toward a more integrative mode! of change.
(Original work published in Chinese)
Psychotherapy: Theory, Research and Practice. 19(2). 276 288.
DiClemente, C. C- (2003). The process of human intentional
Prochaska, J, O,, & Velicer, W. F. (1997). The transtheoretical
behavior change. In C. C. DiClemente (Ed.), Addiction and
model of health behavior change. American Journal of
change: How addictions develop and addicted people
Health Promotion, 2(1), 38-48.
recover (pp. 22-43). New York: Guilford.
Prochaska J. O., Velicer, W. F., Redding, C, Rossi, J. S.,
Duffy, S. W., Tabor, L., Chen, H. H., Holmqvist, M,, Yen, M. F.,
Goldstein, M., Depue, J., et al, (2005). Stage-based expert
Abdsalah, S., et al. (2002). The impact of organized mamsystems to guide a population of primary care patients to
mography service screening on breast carcinoma mortality in
quit smoking, eat healthier, prevent skin cancer, and receive
seven Swedish counties: A collaborative evaluation. Cancer,
regular mammograms. Preventive Medicine, 47(2), 406 416,
95{3), 458 469.
Rakowski, W,, Clark, M. A,, Pearlman, D. N., Ehrich, B., Rimer,
Dutton, G. R., Provost, B. C, Tan, F., & Smith, D. (2008). A taiB. K., Goldstein, M. G., et al. (1997), Integrating pros and

258

Tailored Web-Based Intervetitlon

VOL. 17, NO. 4, DECEMBER 2009

cons for mammography and pap tasting: Extending the

construct of decisional balance to two behaviors. Preventive
Medicine. 26{5). 664 673.

care, and self-choice comparison for repeat mammography. American Journal of Preventive Medicine. 25(4),
308-314.

Rakowski, W., Dube, C. E., & Goldstein, M. G. (1996). Considerations for extending the transtheoretical model of behavior change to screening mammography. Health Education
Research. {1), 77-96.

Redding, C A., Rossi, J. S., Rossi, S. R., Velicer, W. F., &

Prochaska, J. 0. (2000). Health behavior models. International
Electronic Journal of Health Education. 3(3), 180-193.

Rakowski, W., Ehrich, B., Dube, C. E., Pearlman, D. N.,

Goldstein, IV). G., Peterson, K. K., et al. {1996). Screening
mammography and constructs from the transtheoretical
model: Associations using two definitions of the stage of
adoption. Annals of Behavioral Medicine. 18{2), 91-100,
Rakowski, W., Lipkus, I. M., Clark, M. A., Rimer, B., Ehrich, B.,
Lyna, P. R., et al. (2003). Reminder letter, tailored stepped-

Rimer, B. K., Halabi, S., Sugg-Skinner, C, Lipkus, 1. M., Strigo,

T. S-, Kaplan, E. B., et al. (2002). Effects of a mammography
decision-making intervention at 12 and 24 months. American Journal of Preventive Medicine, 2Z.A), 247-257.
Saywell, R. U., Champion, V. L, Skinner, C. S., Menon, U., &
Daggy, J. A. (2004). Cost-effectiveness comparison of three
tailored interventions to increase mammography screening.
Journal of Women's Health. 73(8), 909 918.

259

mmtmm^m

Journal of Nursing Research VOL. 17. NO. 4, DECEMBER 2009

ical Model) hMi

J f 'mx^
15 ;

( precontemplation stage )

mmmm
mEM-mj $i

m ' mmmm - mmm^

260

Copyright of Journal of Nursing Research is the property of Taiwan Nurses Association and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.