G Model

PEC-5016; No. of Pages 6

Patient Education and Counseling xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Patient Education and Counseling

journal homepage: www.elsevier.com/locate/pateducou

Patient Education

Changing knowledge, attitudes, and behaviors regarding cervical

cancer screening: The effects of an educational intervention
in rural Kenya
Joelle I. Rosser a,*, Betty Njoroge b, Megan J. Huchko c
a
Department of Internal Medicine, University of Washington, Seattle, USA
b
Centre for Microbiology Research, Kenya Medical Research Institute (KEMRI), Nairobi, Kenya
c
Department of Obstetrics, Gynecology and Reproductive Sciences, Bixby Center for Global Reproductive Health, University of California, San Francisco, USA

A R T I C L E I N F O A B S T R A C T

Article history: Objective: Cervical cancer screening uptake may be influenced by inadequate knowledge in resource-
Received 16 June 2014 limited settings. This randomized trial evaluated a health talk’s impact on cervical cancer knowledge,
Received in revised form 26 February 2015 attitudes, and screening rates in rural Kenya.
Accepted 21 March 2015
Methods: 419 women attending government clinics were randomized to an intervention (N = 207) or
control (N = 212) group. The intervention was a brief health talk on cervical cancer. Participants
Keywords: completed surveys at enrollment (all), immediately after the talk (intervention arm), and at three-
Cervical cancer screening
months follow-up (all). The primary outcomes were the change in knowledge scores and the final
Knowledge
Risk perception
screening rates at three-months follow-up. Secondary outcomes were changes in awareness about
Behavior cervical cancer screening, perception of personal cervical cancer risk, cervical cancer and HIV stigma, and
Health education screening acceptability.
Sub-Saharan Africa Results: Mean Knowledge Scores increased by 26.4% (8.7 points increased to 11.0 points) in the
intervention arm compared to only 17.6% (8.5 points increased to 10.0 points) in the control arm
(p < 0.01). Screening uptake was moderate in both the intervention (58.9%; N = 122) and control (60.9%;
N = 129) arms, with no difference between the groups (p = 0.60).
Conclusion: A brief health talk increased cervical cancer knowledge, although it did not increase
screening over simply informing women about free screening.
Practical implications: Screening programs can increase patient understanding with just a brief
educational intervention.
ß 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction attitudes. In South Africa, where women are legally entitled to free
cervical cancer screening, screening is estimated at less than 10–
Cervical cancer is a leading cause of cancer in Sub-Saharan 20% [2]. Although this low rate is partially attributable to limited
Africa [1]. Screening coverage in Africa is estimated to be only 2– access, surveys also indicate that one to two thirds of women are
20% in urban areas and 0.4–14% in rural areas; in Kenya, screening unaware of screening availability [3]. Similarly, women in
rates are approximately 4% in urban areas and only 2.6% in rural Botswana reported that the main barrier to screening was
areas [2]. Low screening rates are due not only to limited ‘‘inadequate knowledge about the test’’ (46.7%), ranking even
availability of screening services, but also to barriers to screening above ‘‘limited access to doctors’’ (33.3%) [5]. Women in Limuru,
uptake such as inadequate knowledge, not feeling at risk for Kenya also cited ‘‘lack of knowledge’’ and ‘‘lack of concern’’ as
cervical cancer, and stigma [3–6]. barriers to screening uptake and suggested cervical cancer
Alongside improving access, achieving maximal screening education at the health facility would increase screening [7].
uptake will depend on understanding patient knowledge and In addition to inadequate knowledge, another barrier to
screening is not feeling at risk for cervical cancer. At two large
hospitals in Kenya, 35–69% of women did not perceive themselves
* Corresponding author at: University of Washington, 1959 NE Pacific Street, Box
to be at risk for cervical cancer [8,9]. Additionally, in the Limuru
356421, Seattle, WA 98195-6421, USA. Tel.: +1 310 733 7881; fax: +1 216 685 8652. study, women frequently had misconceptions about cervical
E-mail address: [email protected] (J.I. Rosser). cancer risk factors and said that they did not feel the need for

http://dx.doi.org/10.1016/j.pec.2015.03.017
0738-3991/ß 2015 Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: Rosser JI, et al. Changing knowledge, attitudes, and behaviors regarding cervical cancer screening: The
effects of an educational intervention in rural Kenya. Patient Educ Couns (2015), http://dx.doi.org/10.1016/j.pec.2015.03.017
G Model
PEC-5016; No. of Pages 6

2 J.I. Rosser et al. / Patient Education and Counseling xxx (2015) xxx–xxx

screening because they felt healthy [7]. Perceptions of risk are three months later to complete a follow-up survey; cell phone
formed by information, emotional experiences, and cultural reminders were sent a few days prior to scheduled follow-up
frameworks within a community; and studies have suggested times. Participants who did not follow-up were called and those
that perception of personal risk for developing cancer can affect without functioning cell phone numbers were followed-up at their
cancer screening behaviors [10]. homes using the standard method of patient tracing used by the
Stigma is another potential barrier to screening. Due to cervical clinic’s community health workers based on locating information
cancer’s sexual risk factors and association with HIV, stigma is (i.e. nearby landmarks, local nicknames, etc.) obtained at baseline.
being increasingly recognized as a potential deterrent to screening All participants were invited to seek free cervical cancer screening
acceptance [6]. A study exploring the acceptability of different at the health facility either on a study day or another clinic day at
approaches to cervical cancer education in South Africa also their convenience. Screening verification was done through
highlighted the importance of providing a clear, non-stigmatizing interview and clinic records.
message to increase screening uptake [11].
The Information Motivation and Behavior model purports 2.3. Educational intervention
that several factors influence motivation and behavior but that
having information is an essential component of behavior The intervention consisted of a 30-min interactive talk about
change; this model has been used by HIV and family planning cervical cancer. The talk reviewed basic health facts about cervical
health education programs around the world, including pro- cancer, risk factors, how screening is performed, what screening
grams in Kenya [12]. However, there are currently no widely results mean, and treatment options. The talk also included a
used, reproducible educational interventions to promote cervical guided discussion about barriers to screening and fears or stigma
cancer screening in rural Africa. In 2012, the Family AIDS Care associated with screening. For standardization, each session was
and Education Services (FACES) began supporting a Cervical guided by a flip-chart and corresponding script, with content
Cancer Screening and Prevention (CCSP) program in rural FACES- derived from WHO guidelines and other studies of common
supported government health facilities in the Nyanza Province of misconceptions about cervical cancer [3,5,16]. Community health
western Kenya. As part of this program, screening and treatment workers who administered the talk attended a one-day training to
for precancerous lesions were offered free of charge to all women learn and practice the teaching materials. In order to minimize
(regardless of HIV status). The program developed a clinic-based contamination between the intervention and control arms, the
health talk following principles of culturally-sensitive health health talk was given to the group of women in the intervention
education [13], including creating an educational design that is arm (typically 4–6 women) in a private area of each health facility
familiar to participants, training community members to deliver and sessions typically concluded after all participants had
the health message in the local language, and tailoring the health completed the baseline survey and most women in the control
information to the local context. In order to evaluate the arm had left clinic.
effectiveness of this educational intervention, we conducted a
randomized control trial among women who had not yet been 2.4. Survey tool/measurements
screened. The primary outcomes measured were the change in
knowledge scores and the final screening rates at follow-up. The orally administered surveys lasted approximately 20 min.
Secondary outcomes were changes in cervical cancer awareness, Questions were adapted from previously validated questionnaires
perception of personal risk for cervical cancer, stigma, and [3,11,17] and piloted prior to administration. Surveys collected
screening acceptability. demographic data and included yes/no and true/false questions
regarding cervical cancer awareness, knowledge, perception of
2. Materials and methods personal risk, stigma, and screening acceptability. All survey tools
were written in English and translated into Kiswahili and Dholuo.
2.1. Setting and sampling Trained interviewers administered the surveys in the partici-
pant’s preferred language and recorded responses at the time of
This study took place in eleven CCSP-supported rural health the interview into tablets using the Open Data Kit database
facilities in Suba and Mbita, two of the poorest districts in rural program.
Kenya with an estimated HIV-prevalence of 25% [14,15]. The health
facilities included two district hospitals (highest level facility in a 2.5. Analysis
district offering inpatient, outpatient, and surgical services), five
sub-district hospitals (middle level facility with inpatient and Awareness of cervical cancer was assessed by five yes/no
outpatient services), and four local dispensaries (local outpatient questions asking participants if they had ever heard of cervical
facility). Women attending these health facilities were directly cancer, screening, Pap smears, visual inspection with acetic acid
approached while waiting for health services and invited to (VIA), and human papilloma virus (HPV). An Awareness Score (AS)
participate. Women qualified if they were eligible for cervical was generated out of five possible points with one point given for a
cancer screening according to FACES guidelines (i.e. non-pregnant ‘‘yes’’ response. The Knowledge section consisted of 15 true/false
women at least 23 years of age), had not previously been screened, statements that included both facts and common myths about
could speak Kiswahili, Dholuo, or English, and were able and cervical cancer, risk factors, and HPV. A Knowledge Score (KS) was
willing to provide informed consent. Participants were random- then generated, with one point given for each correct answer and
ized to the control or intervention arm. Randomization was done in no points given for incorrect answers and ‘‘I don’t know.’’
computer-generated blocks of eight. Perception of Risk was assessed by asking participants to respond
‘‘yes,’’ ‘‘no,’’ or ‘‘I don’t know’’ to the statement ‘‘I think I am at risk
2.2. Study design for cervical cancer.’’ Stigma was evaluated in two steps. First,
participants answered an HIV stigma questionnaire that had been
All participants completed a baseline survey on the day of previously validated in this region to measure HIV stigma
enrollment. Participants in the intervention arm then participated [17]. Next, participants were asked these same stigma questions
in the health talk, followed by a post-education survey on the same in relation to cervical cancer. A Cervical Cancer Stigma Score and an
day. All participants were invited to return to the health facility HIV Stigma Score were both created out of 9 possible points, with

Please cite this article in press as: Rosser JI, et al. Changing knowledge, attitudes, and behaviors regarding cervical cancer screening: The
effects of an educational intervention in rural Kenya. Patient Educ Couns (2015), http://dx.doi.org/10.1016/j.pec.2015.03.017
G Model
PEC-5016; No. of Pages 6

J.I. Rosser et al. / Patient Education and Counseling xxx (2015) xxx–xxx 3

one point given for each ‘‘yes’’ response. Based on previous studies, followed-up, there were no demographic differences between the
a score 1.8 points would suggest the presence of disease-related intervention and control arm. Of those who followed-up, baseline
stigma [17,18]. Finally, Screening Acceptability was measured by scores across all outcome measures did not significantly differ
asking whether the participant would accept screening. Partici- between the two study arms.
pants were offered screening and uptake was recorded. Women
who did not get screened were asked their reasons for declining. 3.2. Primary outcomes: effects of an educational intervention on
The differences between mean survey scores from the initial knowledge and screening
survey to the three-month follow-up survey for the two study arms
were compared using chi-square or t-tests where appropriate. The Knowledge Scores increased significantly after the educational
screening rates at three-month follow-up for the two arms were intervention. At three months follow-up, mean Knowledge Scores
compared using a chi-square test. Outcome measures for the initial in the intervention arm increased 26.4% (8.7 points increased to
survey and the immediate post-education survey were also 11.0 points) compared to only a 17.6% increase (8.5 points
compared using chi-square or t-tests to measure the immediate increased to 10.0 points) in the control arm, t(326) = 2.64, p < 0.01
impact of the health talk for all women enrolled in the intervention (Table 2).
group. Stigma Scores, which had a skewed distribution, were Over half of all participants in both the intervention arm
analyzed with the Wilcoxon–Mann–Whitney test. Randomization (N = 122/207; 58.9%) and the control arm (N = 129/212; 60.9%) got
and data analysis were done using STATA version 12.0 (College screened during the study period, with no significant difference in
Station, TX). screening rates between the two groups, X2 (1, N = 419) = 0.16,
p = 0.69. Screening rates were significantly higher amongst women
2.6. Ethics who followed-up (N = 228/328; 69.5%) compared to women who did
not follow-up (N = 23/91; 25.3%), X2 (1, N = 419) = 58.04, p = <0.001.
All study participants were informed about the study design, However, amongst only those women who followed-up, there was
objectives, and follow-up protocol and they signed a written again no significant difference in screening rates between the
consent in their preferred language. Ethical approval was obtained intervention (N = 109/161; 67.7%) and control (N = 119/167; 71.2%)
from the Kenya Medical Research Institute Ethical Review Commit- arms (p = 0.49), X2 (1, N = 328) = 0.49, p = 0.48.
tee and the University of California, San Francisco Committee on
Human Research. 3.3. Secondary outcomes: effects of an educational intervention on
awareness, risk perception, stigma, and screening acceptability
3. Results
Awareness Scores did increase significantly more in the
3.1. Demographic characteristics intervention versus control arm at three months follow-up. Mean
Awareness Scores increased by 53.8% (2.6 points increased to
Between March and April 2013, women were recruited from 4.0 points) in the intervention arm compared to only 37.5% (2.4
eleven facilities to participate in the study. 419 women met study points increased to 3.3 points) in the control arm, t(326) = 2.6,
criteria, consented, and were individually randomized to the p < 0.01 (Table 2).
intervention (N = 207) or control (N = 212) arm (Fig. 1). Demo- Perceived cervical cancer risk increased in both groups after
graphic characteristics did not differ significantly between the three months with no significant difference between women who
intervention and control groups, with the exception that women in received education and those who did not, X2 (2, N = 328) = 0.73,
the control group had higher parity (p = 0.02) and gravidity p = 0.69. Cervical Cancer Stigma Scores decreased in both groups
(p = 0.03) (Table 1). with no difference between the intervention and control groups
Three hundred and twenty-eight (78%) women completed the (Z = 0.49, p = 0.62). HIV Stigma Scores also decreased overall, and
three-month follow-up survey, 161 (78%) from the intervention the decrease was significantly larger in the intervention group
and 167 (79%) from the control (p = 0.81). Compared to women (Z = 2.0, p = 0.05). Screening acceptability increased overall at
who did follow-up, women who did not follow-up were younger follow-up but the change in acceptability did not differ signifi-
(31.6 yrs vs 33.8 yrs; p < 0.05), more likely to be interviewed at a cantly between women in the intervention and control arms, X2 (2,
district hospital (36.3% vs 22.9%; p = 0.04), and more likely to have N = 328) = 1.19, p = 0.55. Additionally, screening rates were not
never been tested for HIV (13.2% vs 5.2%; p < 0.01). Of those who significantly associated with final measures of cervical cancer
awareness, knowledge, risk, or stigma (Table 2).
Immediately following the health talk, women in the interven-
tion arm (N = 207) had significantly increased cervical cancer
knowledge (p < 0.001), awareness (p < 0.001), perception of
personal risk (p = 0.001), and had decreased cervical cancer stigma
(p < 0.001) and HIV stigma (p < 0.001). Screening acceptability
also increased but was not significantly different than the initial
survey (p = 0.26). These post-education measures were compara-
ble to those seen at three-month follow-up and therefore are not
reported.

3.4. Reported reasons for declining screening

On the day of enrollment, the most common reasons for

declining screening were ‘‘busy today/no time to wait’’ (N = 92/
168; 54.8%), ‘‘currently menstruating’’ (N = 36/168; 21.4%), and
‘‘undecided/want more time to think about it’’ (N = 34/168; 20.2%).
At follow-up, in addition to ‘‘busy today/no time to wait’’ (N = 15/
Fig. 1. Participant flow chart. 79; 19.0%) and ‘‘currently menstruating’’ (N = 14/79; 17.7%), two

Please cite this article in press as: Rosser JI, et al. Changing knowledge, attitudes, and behaviors regarding cervical cancer screening: The
effects of an educational intervention in rural Kenya. Patient Educ Couns (2015), http://dx.doi.org/10.1016/j.pec.2015.03.017
G Model
PEC-5016; No. of Pages 6

4 J.I. Rosser et al. / Patient Education and Counseling xxx (2015) xxx–xxx

Table 1
Demographic characteristics.

Intervention (N = 207) Control (N = 212) p-Value

a
Age (mean  sd) 32.6  8.9 34.0  9.6 0.13
Relationship status 0.42
Single 24 (11.6%) 16 (7.6%)
Married 134 (64.7%) 148 (69.8%)
Widowed 46 (22.2%) 43 (20.3%)
Separated/divorced 3 (1.5%) 5 (2.4%)
Highest educational level 0.68
Primary school incomplete or none 99 (47.8%) 101 (47.6%)
Primary school complete 57 (27.5%) 52 (24.5%)
Some secondary school or beyond 51 (24.6%) 59 (27.8%)
Occupation 0.32
Professional/technical/managerial 13 (6.3%) 20 (9.4%)
Housewife/farming 81 (39.1%) 89 (42.0%)
Non-professional work outside the home 113 (54.6%) 103 (48.6%)
Facility typea 0.35
District hospital (Level 1 Facility) 56 (27.1%) 52 (24.5%)
Sub-district hospital (Level 2 Facility) 92 (44.4%) 97 (45.8%)
Dispensary (Level 3 Facility) 59 (28.5%) 63 (29.7%)
Transportation to clinic 0.23
Walking 112 (54.1%) 132 (62.3%)
Motorcycle 89 (43.0%) 74 (34.9%)
Other (bicycle, boat, bus) 6 (2.9%) 6 (2.8%)
Travel time to clinic (mean minutes  sd) 45.2  37.3 49.7  42.7 0.25
Primary source of health information 0.14
Health facility or healthcare worker 171 (82.6%) 186 (87.7%)
Outside source (radio, church, school, etc.) 36 (17.4%) 26 (12.3%)
Knows someone with cervical cancer 62 (30.1%) 59 (28.0%) 0.63
Prior health seeking behavior
History of STD testing 36 (17.5%) 45 (21.2%) 0.33
History of breast exam 15 (7.3%) 10 (4.7%) 0.27
History of mammogram 9 (4.4%) 8 (3.8%) 0.77
History of HIV testing* 192 (92.8%) 198 (93.4%) 0.80
Reproductive history
Gravida (mean  sd) 3.2  2.3 3.7  2.5 0.03
Parity (mean  sd) 2.9  2.1 3.4  2.3 0.02
Age of sexual debut (mean  sd) (N = 105) 16.5+ 2.6 16.5  2.6 0.72
# of current sexual partners (mean  sd) 0.9  0.4 0.9  0.5 0.50
# of lifetime sexual partners (mean  sd) 2.2  1.3 2.3  1.2 0.30
Sex workera 3 (1.5%) 5 (2.4%) 0.50
Family planning
None 122 (58.9%) 105 (49.5%) 0.08
Type of family planning (% of those using FP) 0.97
Depo-provera 40 (48.2%) 53 (49.5%)
Long-term (IUCD or Implant) 21 (25.3%) 23 (21.5%)
Condom 9 (10.8%) 12 (11.2%)
Permanent (tubal ligation or vasectomy) 7 (8.4%) 9 (8.4%)
Other (patch, OCP’s, natural, multiple) 6 (7.2%) 10 (9.4%)
HIV statusa 0.73
Positive 115 (55.6%) 118 (55.7%)
Negative 59 (28.5%) 55 (25.9%)
Unknownb 33 (15.9%) 39 (18.4%)
Average duration since tested positive (years) 3.4  2.8 3.1  2.7 0.46
Average duration since last negative HIV test 0.7  0.8 1.0  1.7 0.24
a
Characteristics that differed significantly (p < 0.05) between participants who followed-up and those who did not follow-up. Women who did NOT follow-up were more
likely to be younger, attend a district hospital, be a sex worker, have no history of HIV testing, and have an unknown HIV status.
b
‘‘Unknown’’ = never screened for HIV or no negative test within the last year.

other reasons frequently given at this point were ‘‘do not received the health talk and women who were simply informed
understand enough about screening’’ (N = 15/79; 19.0%) and ‘‘fear that screening was available. Cervical cancer risk perception,
of pain with the speculum exam’’ (N = 13/79; 16.5%). There was no stigma, and screening acceptability also improved at follow-up.
significant difference between reasons given by women in the However, the intervention did not have a significant effect on these
different study arms. measures when compared to the control group.
Despite widespread recognition that cervical cancer knowledge
4. Discussion and conclusion is low in Sub-Saharan Africa, there are few prospective controlled
studies evaluating educational interventions. Trials of cervical
4.1. Discussion cancer education programs in Turkey and Nigeria have shown
increased knowledge post-intervention [19,20], but knowledge
Our single brief health talk significantly increased cervical and attitudes are highly contextual and educational programs may
cancer knowledge and awareness at three months. Screening not be readily transferable. Our study shows the positive impact of
uptake was moderate at three-months follow-up in both groups an educational intervention on knowledge in Eastern Africa, the
and screening rates were no different between women who region with the highest incidence of cervical cancer worldwide [1].

Please cite this article in press as: Rosser JI, et al. Changing knowledge, attitudes, and behaviors regarding cervical cancer screening: The
effects of an educational intervention in rural Kenya. Patient Educ Couns (2015), http://dx.doi.org/10.1016/j.pec.2015.03.017
G Model
PEC-5016; No. of Pages 6

J.I. Rosser et al. / Patient Education and Counseling xxx (2015) xxx–xxx 5

Table 2
Changes in knowledge and attitudes in intervention versus control arms.

Initial survey Follow-up survey Change

Intervention Control p-Value Intervention Control p-Value Intervention Control p-Value

(N = 161) (N = 167) (N = 161) (N = 167)

Knowledge Scorea 8.7 (8.3–9.0) 8.5 (8.1–8.9) 0.17 11.0 (10.7–11.3) 10.0 (9.6–10.3) <0.001 2.3 (1.9–2.7) 1.5 (1.0–2.0) <0.01c
(0–15 points)
Awareness Scorea 2.6 (2.4–2.8) 2.4 (2.2–2.6) 0.44 4.0 (3.8–4.1) 3.3 (3.1–3.4) <0.001 1.4 (1.1–1.6) 0.9 (0.7–1.1) <0.01d
(0–5 points)
Perception of personal 115 (71.4%) 116 (69.5%) 0.15 135 (83.9%) 143 (85.6%) 0.65 12.5% 16.1% 0.69d
cervical cancer riskb
Cervical Cancer Stigma 0.8 (0.5–1.1) 0.6 (0.4–0.9) 0.49 0.2 (0.1–0.3) 0.2 (0.0–0.3) 0.49 0.6 (0.3–0.9) 0.5 (0.2–0.7) 0.62d
Scorea (0–9 points)
HIV Stigma Scorea 1.2 (0.8–1.5) 0.7 (0.5–0.9) 0.13 0.4 (0.2–0.6) 0.4 (0.2–0.5) 0.63 0.8 (0.4–1.1) 0.3 (0.1–0.6) 0.05d
(0–9 points)
Screening Acceptabilityb 127 (78.9%) 133 (79.6%) 0.87 145 (90.1%) 155 (92.8%) 0.37 11.2% 13.2% 0.55d
a
Reported mean (95% CI).
b
Reported N (%).
c
Primary outcome: at 3-month follow-up, Knowledge Score increased significantly more in the intervention arm compared the control arm.
d
Secondary outcomes: at 3-month follow-up, Awareness Score increased significantly more in the intervention arm compared to the control arm. The HIV Stigma Score
also decreased significantly more in the intervention arm, although the final HIV Stigma Score at follow-up did not differ between arms. There was no difference in the change
in perception of personal cervical cancer risk, cervical cancer stigma, or screening acceptability between the two arms.

Data on how increased knowledge affects screening rates is change. Other important factors may include shorter wait times,
even more limited. A retrospective study in East Africa showed a repeated screening opportunities, and addressing women’s fears
positive association between screening history and cervical cancer of the exam, as supported by this study and in prior literature
knowledge, raising the question of whether education can increase [22,23].
screening [21]. However, in our study, screening rates did not differ
between women who attended the health talk and those who were 4.3. Practice implications
simply informed that they were eligible for free screening. It is
possible that information about screening availability may be just This study validated the positive influence of an educational
as effective as more in-depth education. A shorter message may in intervention on knowledge, a commonly cited barrier to screening.
fact leave more time for screening, addressing the long wait times Increasing knowledge about cervical cancer and awareness about
cited by many participants as their reason for declining screening. screening opportunities can be achieved with a brief clinic-based
This study is unique in creating and validating a cervical cancer health talk, but this is just the first step to improving screening
educational intervention in this setting through a randomized trial. rates. Attitudes and motivations are shaped not only by individual
However, it has some limitations. Sampling was limited to clinic knowledge but also cultural shifts within a community; and
attendees and may not be representative of the wider community. decisions are made based on these and other competing interests
Additionally, social desirability bias may have influenced such as finances and time [24]. As health systems increase efforts
responses to questions about general awareness, risk perception, to address the enormous burden of cervical cancer disease,
and stigma. However, the fact-based cervical cancer knowledge educational programs will need to expand beyond the clinic and
questions should not have been subject to this type of bias. into the community and continue striving to provide services in a
Interestingly, all outcome measures improved significantly patient-centered manner.
from baseline to follow-up in both arms. The changes in the control
group could be attributed to several things including: (1) Conflict of interest
contamination bias through peer to peer education after the
educational intervention, (2) increased overall community aware- The authors have no conflicts of interest.
ness during the follow-up period from ongoing community health
campaigns, and (3) an increased awareness and interest in cervical
Role of funding
cancer as a result of simply being informed about screening
availability. A strength of this study is that it compares the
This work was supported by a grant from the Doris Duke
outcome measures to a control group, thereby minimizing the Charitable Foundation to the University of California, San Francisco
potential overestimation of the intervention’s impact.
to fund Clinical Research Fellow Joelle Rosser. While working on
this project, M.J.H. was supported through a National Institutes of
4.2. Conclusion
Health career development award (KL2 RR024130-04).

The educational intervention increased knowledge and aware-

Acknowledgements
ness about cervical cancer screening, but it did not result in higher
screening rates. A minimum level of understanding may be
Elina Serrano for flipchart picture design. Jane Adera for
necessary for a woman to accept screening but perhaps simply
Kiswahili and Dholuo translations.
informing women about screening availability is sufficient. The
lack of difference in screening rates between the two study arms
also raises the question of what factors that remained unchanged
by the health talk could still influence a woman’s screening Appendix A. Supplementary data
behavior. Stigma and perception of personal risk, which did not
differ between the study arms at follow-up, may still be important Supplementary data associated with this article can be found, in
barriers to screening but may require more than a health talk to the online version, at http://dx.doi.org/10.1016/j.pec.2015.03.017.

Please cite this article in press as: Rosser JI, et al. Changing knowledge, attitudes, and behaviors regarding cervical cancer screening: The
effects of an educational intervention in rural Kenya. Patient Educ Couns (2015), http://dx.doi.org/10.1016/j.pec.2015.03.017
G Model
PEC-5016; No. of Pages 6

6 J.I. Rosser et al. / Patient Education and Counseling xxx (2015) xxx–xxx

and tailored approaches. Health Educ Behav: Off Publ Soc Publ Health Educ
References
2003;30:133–46.
[14] Camlin CS, Kwena ZA, Dworkin SL. Jaboya vs. jakambi: status, negotiation, and
[1] Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. HIV risks among female migrants in the ‘‘sex for fish" economy in Nyanza
CA Cancer J Clin 2011;61:69–90. Province, Kenya. AIDS Educ Prev 2013;25:216–31.
[2] Louie KS, de Sanjose S, Mayaud P. Epidemiology and prevention of human [15] Cohen CR, Montandon M, Carrico AW, Shiboski S, Bostrom A, Obure A, et al.
papillomavirus and cervical cancer in sub-Saharan Africa: a comprehensive Association of attitudes and beliefs towards antiretroviral therapy with
review. Trop Med Int Health 2009;14:1287–302. HIV-seroprevalence in the general population of Kisumu, Kenya. PLoS One
[3] Francis SA, Nelson J, Liverpool J, Soogun S, Mofammere N, Thorpe Jr RJ. Examining 2009;4:e4573.
attitudes and knowledge about HPV and cervical cancer risk among female clinic [16] WHO. Cervical cancer screening in developing countries: report of a WHO con-
attendees in Johannesburg, South Africa. Vaccine 2010;28:8026–32. sultation 2002. http://whqlibdoc.who.int/publications/2002/9241545720.pdf.
[4] Hoque M, Hoque E, Kader SB. Evaluation of cervical cancer screening program at a [17] Turan JM, Bukusi EA, Onono M, Holzemer WL, Miller S, Cohen CR. HIV/AIDS
rural community of South Africa. East African J Public Health 2008;5:111–6. stigma and refusal of HIV testing among pregnant women in rural Kenya:
[5] McFarland DM. Cervical cancer and Pap smear screening in Botswana: knowl- results from the MAMAS study. AIDS Behav 2011;15:1111–20.
edge and perceptions. Int Nurs Rev Sep 2003;50:167–75. [18] Weiser SD, Heisler M, Leiter K, Percy-de Korte F, Tlou S, DeMonner S, et al.
[6] White HL, Mulambia C, Sinkala M, Mwanahamuntu MH, Parham GP, Money- Routine HIV testing in Botswana: a population-based study on attitudes,
ham L, et al. ‘Worse than HIV’ or not as serious as other diseases’? Conceptu- practices, and human rights concerns. PLoS Med 2006;3:e261.
alization of cervical cancer among newly screened women in Zambia. Soc Sci [19] Bebis H, Reis N, Yavan T, Bayrak D, Unal A, Bodur S. Effect of health education
Med 2012;74:1486–93. about cervical cancer and papanicolaou testing on the behavior, knowledge,
[7] Gatune JW, Nyamongo IK. An ethnographic study of cervical cancer among and beliefs of Turkish women. Int J Gynecol Cancer 2012;22:1407–12.
women in rural Kenya: is there a folk causal model? Int J Gynecol Cancer [20] Wright KO, Kuyinu YA, Faduyile FA. Community education on cervical cancer
2005;15:1049–59. amongst market women in an urban area of Lagos, Nigeria. Asian Pac J Cancer
[8] Gichangi P, De Vuyst H, Estambale B, Rogo K, Bwayo J, Temmerman M. HIV and Prev 2010;11:137–40.
cervical cancer in Kenya. Int J Gynaecol Obstet 2002;76:55–63. [21] Lyimo FS, Beran TN. Demographic, knowledge, attitudinal, and accessibility
[9] Were E, Nyaberi Z, Buziba N. Perceptions of risk and barriers to cervical cancer factors associated with uptake of cervical cancer screening among women in a
screening at Moi Teaching and Referral Hospital (MTRH), Eldoret, Kenya. Afr rural district of Tanzania: three public policy implications. BMC Public Health
Health Sci 2011;11:58–64. 2012;12:22.
[10] Lee SJ. Uncertain futures: individual risk and social context in decision-making [22] Busingye P, Nakimuli A, Nabunya E, Mutyaba T. Acceptability of cervical cancer
in cancer screening. Health Risk Soc 2010;12:101–17. screening via visual inspection with acetic acid or Lugol’s iodine at Mulago
[11] Maree JE, Wright SC. Cervical cancer: does our message promote screening? A Hospital, Uganda. Int J Gynaecol Obstet 2012;119:262–5.
pilot study in a South African context. Eur J Oncol Nurs 2011;15:118–23. [23] Stormo AR, Altamirano VC, Perez-Castells M, Espey D, Padilla H, Panameno K,
[12] Onono M, Blat C, Miles S, Steinfeld R, Wekesa P, Bukusi E, et al. Impact of family et al. Bolivian health providers’ attitudes toward alternative technologies for
planning health talks by lay health workers on contraceptive knowledge and cervical cancer prevention: a focus on visual inspection with acetic acid and
attitudes among HIV-infected patients in rural Kenya. Patient Educ Couns cryotherapy. J Womens Health (Larchmt) 2012;21:801–8.
2014;94:438–41. [24] Sanders Thompson VL. Cultural context and modification of behavior change
[13] Kreuter MW, Lukwago SN, Bucholtz RD, Clark EM, Sanders-Thompson V. theory. Health Educ Behav: Off Publ Soc Publ Health Educ 2009;36(Suppl. 5):
Achieving cultural appropriateness in health promotion programs: targeted 156S–60S [discussion 167S-171S].

Please cite this article in press as: Rosser JI, et al. Changing knowledge, attitudes, and behaviors regarding cervical cancer screening: The
effects of an educational intervention in rural Kenya. Patient Educ Couns (2015), http://dx.doi.org/10.1016/j.pec.2015.03.017