Ferrer, Ramirez Et Al (2010)

8/4/2019 Ferrer, Ramirez Et Al (2010)

1/9

Psycho-Oncology

Psycho-Oncology (2010)

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/pon.1793

The impact of cultural characteristics on colorectalcancer screening adherence among Filipinos in the

United States: a pilot studyRizaldy R. Ferrer1, Marizen Ramirez2, Linda J. Beckman1, Leda L. Danao3 and Kimlin T. Ashing-Giwa4

1California School of Professional Psychology, Alliant International University, Alhambra, CA, USA2Department of Occupational and Environmental Health, University of Iowa, Iowa City, IA, USA3Division of Cancer Prevention and Control Research, School of Public Health/Jonsson Comprehensive Cancer Center, University of California, Los

Angeles, CA, USA4Center of Community Alliance for Research and Education, Division of Population Sciences, City of Hope National Medical Center, Duarte, CA, USA

Abstract

Objectives: Studies on colorectal cancer screening among specific Asian American groups are

limited despite the fact that Asians are comprised of culturally distinct subgroups. The purposeof this study was to investigate the impact of cultural characteristics on colorectal cancer

screening adherence among Filipinos in the United States.

Methods: One hundred and seventeen Filipino men and women aged 50 years or older

participated in the cross-section research design. Lifetime proportion of immigration, language

preference and cultural beliefs of personal control regarding health outcomes measured cultural

characteristics. Demographic and healthcare variables were also measured to describe the

study sample. Participant recruitment employed culturally responsive sampling methods.

Results: There was no significant association between language preference and screening.

Likewise, perceived personal internal control of health outcome was not related to screening.

However, personal external control revealed a marginally significant association. The percent

of lifetime residence in the United States was significantly greater among those who were

adherent to screening than those who were not adherent. After adjusting for demographic and

healthcare variables, the relationship between length of immigration and screening adherence

was no longer significant. Finally, age and doctors recommendation showed significant impacton colorectal cancer screening adherence.

Discussion: This pilot study adds to the knowledge regarding cultural factors associated

with colorectal cancer screening behaviors among Filipino Americans. Future research is

needed to confirm findings that will be useful in developing culturally appropriate strategies to

increase screening adherence.

Copyright r 2010 John Wiley & Sons, Ltd.

Keywords: colorectal cancer; Filipino Americans; acculturation; Asian Americans; screening

adherence

Introduction

Among ethnic minority populations in the UnitedStates (U.S.), Asians had the second highestincidence rate of colorectal cancer for men andthird highest for women from 20022006. How-ever, they showed lower colorectal cancer mortalityrate compared with other ethnic minorities [1].Regardless, characteristics associated with color-ectal cancer outcomes are not well understood.Meanwhile, a deeper understanding of screeningadherence might bring forth insight about effective

interventions.High-quality screening procedures are effective

in substantially reducing cancer incidence and

preventing many cancer-related deaths [2]. Forcolorectal cancer screening, fecal occult blood test(FOBT) every year, flexible sigmoidoscopy (SIG)every 5 years, and colonoscopy (COL) every 10years are recommended, according to the nationalguidelines for people at average risk [3]. Theseindividuals are men and women aged 50 years orolder with no known personal or family history ofcolorectal cancer. Despite strong evidence thatincidence of colorectal cancer can be reducedthrough screening, adherence rates among Amer-icans remain low [4]. Among Asian Americans

aged 50 years or older, 55.9% reported receivingFOBT within the past year, and/or lower endo-scopy (SIG or COL) in the past 10 years. This rate

* Correspondence to: AlliantInternational University,1000 South Fremont Avenue,

Unit 5, Alhambra, CA91803, USA. E-mail:[email protected]

Received: 1 September 2009

Revised: 14 May 2010

Accepted: 27 May 2010

Copyrightr 2010 John Wiley & Sons, Ltd.


2/9

was lower when compared with non-HispanicWhite and non-Hispanic Black Americans.

Factors associated with colorectal cancerscreening

Sociodemographic characteristics such as increasingage [59] and positive family history of colorectalcancer [8,1014] predicted screening adherence. How-ever, other characteristics such as gender, socio-economic status, and level of education resulted ininconsistent findings [8,1319].

Healthcare factors having health insurance,regular physician, usual source of care, regularmedical evaluation, and receiving a doctor recom-mendation for colorectal screening were associatedwith higher utilization of colorectal screening[5,811,14,2023].

Asian Americans and colorectal cancer screening

The impact of demographic and healthcare factorson screening adherence among Asian Americans ispoorly understood. In addition, culture, an im-portant determinant of cancer screening [24],remains an understudied factor in colorectal cancerscreening [25].

In addition, generation level, language fluency,and social activity preference had no impact onscreening adherence among Chinese participants[26]. However, English proficiency among Koreanswas found significantly associated with FOBTreceipt [27]. In the same sample, Koreans whospent more than 25% of lifetime in the U.S. weretwo times likelier to have had SIG when comparedwith those who spent less than 25% of theirlifetime. In general, knowledge about the relation-ships between length of immigration, languagepreference as well as other cultural factors, such ashealth beliefs and colorectal cancer screening, islimited.

Exploration of the factors associated with color-ectal cancer screening adherence among the Asian

Americans is timely. They are comprised of morethan 11 subgroups that differ in language andculture [28], and their cultural values and behaviorssignificantly vary [29]. Therefore, it is important todisaggregate these groups in order to identify andunderstand unique factors associated with color-ectal cancer screening. For example, the impact oflanguage on adherence needs to be examined foreach Asian subpopulation given that level ofproficiency may vary between groups.

Filipino americans and colorectal cancer screening

The second largest Asian subgroup in the U.S. isthe Filipinos [28]. Knowledge about colorectalcancer outcome is sparse. Filipino men and womenin the U.S. were reportedly at a greater risk of

colorectal cancer compared with their counter-parts living in the Philippines [30]. Additionally,Filipinos were more likely to receive a late-stagediagnosis of colorectal cancer compared to theirCaucasian and other Asian counterparts [31].

Filipinos reportedly ranked low in up-to-date

colorectal cancer screening between 2001 and 2003in comparison with their Japanese, Chinese, SouthAsian, and Vietnamese counterparts [32,33]. Like-wise, conveniently sampled Korean women weremore likely to be adherent to colorectal cancerscreening than Filipino women [34]. In 2005, only46% of Filipinos in California reported up-to-datereceipt of FOBT or lower endoscopy [35].

One factor associated with colorectal cancerscreening in the Filipino population is lifetimeproportion of residency in the U.S. [34,35]. Therole of other cultural characteristics such as

language preference and cultural-related beliefs isstill unknown.

This study addressed a critical research gap. Thegoal was to describe the demographic, healthcareand cultural characteristics of Filipinos living inthe U.S. who were surveyed regarding colorectalcancer screening adherence. The study also ex-plored how the cultural factors impact screeningadherence.

Methods

Participants and proceduresThe institutional review board of Alliant Interna-tional University, Los Angeles, approved the study.The authors obtained written consent from a totalof 117 eligible Filipino participants who met thefollowing criteria of persons at average risk ofcolorectal cancer, defined by the American CancerSociety as: male or female, aged 50 years or older,and no previous personal or family history ofcolorectal cancer.

Participant recruitment and data collection wereexecuted based on a culturally responsive approach

suggested in past studies [25,36,37]. This includesestablishing liaisons between the investigator andprospective participants as well as utilizing purpo-sive, convenience and snowball sampling methods.

In the initial step, the principal investigator (PI)contacted a total of five Filipino community groupleaders as well as seven personal contacts todescribe study purpose and determine their will-ingness to serve as liaisons to identify eligibleparticipants. Those who agreed submitted a list ofnames and contact information using a standardreferral form provided to them. After initial

contact with the persons identified on the list,survey packets were sent to eligible individuals whoagreed to participate. Completed surveys in sealedself-addressed stamped envelopes were mailed backto the PI.

Copyrightr 2010 John Wiley & Sons, Ltd. Psycho-Oncology (2010)

DOI: 10.1002/pon

R. R. Ferrer et al.


3/9

Participants were also recruited from threeinformal social events. The PI described thepurpose of the study and invited volunteers toindependently complete the self-administered sur-vey packet. Participants who chose to respond at alater date were provided with a self-addressed

stamped envelope for ease of returning the surveyvia mail.

The second step utilized a snowball samplingmethod. Participants who were recruited andcompleted surveys through the convenience andpurposive sampling method previously describedwere requested to voluntarily refer other eligibleparticipants following similar procedure as duringthe initial phase. The locale of the study is inCalifornia; however, as a result of the differentrecruitment techniques, a few responses were alsoreceived from several other states in America.

Design and measures

A cross-section survey design was conducted inSpring 2006. Respondents reported their demo-graphic characteristics: age, gender, academiccompletion, employment, and income level. Todescribe healthcare characteristics, participantswere asked if they had medical insurance, oneregular doctor, routine medical examination withinthe past year, and received physician recommenda-tion to undergo colorectal cancer screening.

Cultural characteristics were operationalized inseveral ways, beginning with language preference.The Suinn-Lew Asian Self-Identity AcculturationScale (SL-ASIA) is a 21-item instrument thatmeasures multiple dimensions of acculturationincluding language, behaviors, identity, friendshipchoice, generation level, and attitude [38]. For thisstudy, only four items of the language dimensionwere used. Questions were asked about languagespoken, preferred, used for reading, and used forwriting. The term Filipino languages replacedAsian languages on the response options. Theoptions were (1) Filipino languages only, (2) mostly

Filipino languages, some English, (3) Filipino andEnglish languages about equally well, (4) mostlyEnglish, some Filipino language, (5) only English.

To obtain an average score, item scores weresummed and divided by the number of itemsanswered. Respondents were classified as Westernidentified if their mean score is X4, Filipinoidentified or preference towards native language ifthey score o3, and bicultural identified or equalpreference to native and American English lan-guages if their mean score is X3 but o4.

Overall, SL-ASIA demonstrated acceptable in-

ternal consistency reliability with Cronbachs aranging from 0.68 to 0.91 based on multiple studiesof Asian adults [3841]. For this study using thefour language items, the internal consistencyreliability analysis yielded Cronbachs a5 0.69.

Second, length of immigration was measuredusing the lifetime proportion of U.S. residency.This was computed by dividing the participantsreported number of years in the U.S. sinceimmigration by their age expressed in years.

Third, cultural health beliefs were measured by

adapting two subscales from the Cultural BeliefScales for Mammography Screening (CBSMS) [42]because, to the authors knowledge, no instrumentspecific to colorectal cancer screening existedduring the study period. Items on CBSMS wereoriginally developed through rigorous literaturereview, focus groups, and content validation byidentified experts in order to create three culturallysensitive constructs among African Americanwomen. One of the dimensions, personal control,with two distinct subscales was used in this study.

Internal control is defined as the influence of

oneself to finding health problems early, whereasexternal control is the perception that healthoutcomes are influenced by others and chance.An example item for the internal controlsubscale isFinding health problems early is my responsibil-ity, whereas an example item for the externalcontrolsubscale is There is nothing that I can do tofind health problems early.

All 12 items focused on finding health problemsearly. For one item that asked specifically aboutbreast cancer, the term breast cancer was replacedby colorectal cancer. Responses ranged from 1,endorsing strongly disagree, to 5, strongly agree.To obtain a mean scaled score for each construct,the summed scores were divided by the number ofanswered items. A score of 5 for each constructsuggests high internal control or external control.

The instrument demonstrated acceptable inter-nal consistency reliabilities for internal controlsubscale, Cronbachs a50.76 and for externalcontrol, a5 0.82 [42]. This study also has anacceptable reliability coefficient for internal controland external control subscales (Cronbach a5 0.91and a5 0.76, respectively).

For the outcome measurecolorectal cancer

screening adherence was defined as a self-reportrather than clinical records. Participants were askedwhether they have received any of the followingcolorectal cancer screening tests as described:FOBT, SIG and COL. The FOBT examines bloodin the stool that could come from cancer orpremalignant polyps in the colon or rectum. TheSIG examines up to one-third of the colon througha flexible lighted tube, whereas the COL, anothercolorectal cancer screening test similar to SIG, usesa longer tube that can examine the entire colon [43].

Respondents were classified adherent if they

reportedreceiving any of the three screening tests atthe recommended national guidelines [3]. Respon-dents were classified not-adherent if they had notreceived any tests at the recommended time intervalor had never received any test.

Colorectal cancer screening adherence among Filipinos


DOI: 10.1002/pon


4/9

Data analysis

The SPSS 16.0 and STATA 9.0 software packageswere used for the statistical tests. Variablesincluding education, current employment, andannual income were collapsed into two or three

categories due to small cell sizes. Descriptive testswere used to describe the frequency distribution,central tendency, and dispersion of the independentand dependent variables. Bivariate tests using w2

test of independence for categorical variables andanalysis of variance for continuous variables wereconducted to determine differences between color-ectal cancer screening adherent and not-adherentgroups.

A multivariable logistic regression was per-formed to create models for predicting colorectalcancer screening adherence. The level of signifi-cance was set at a level ofap0.05. First, culturalvariables including proportion of lifetime in theU.S., cultural health beliefs (i.e. internal andexternal control), and language preference (i.e.Western identified, Filipino identified, or biculturalidentified) were entered in a model to determinetheir mutually adjusted effects on colorectal cancerscreening adherence. Then, two models wereconstructed to control for confounding by demo-graphic and healthcare characteristics. In the firstmodel, the analysis included demographic andhealthcare characteristics that had cell sizes greaterthan five and were significantly associated with

screening determined by w2

tests. In the secondmodel, the authors considered the possible mediat-ing effect of doctors recommendation on culturalcharacteristics. It is possible, for example, thatincreased time of residency in the U.S. directlyleads to increased healthcare utilization and thus,receipt of doctors recommendation for screening.Therefore, in model 2, we excluded this possibleintermediate, doctors recommendation, from themultivariable model. For both models, age wasrescaled to reflect a 5-year increase in age, whilelifetime proportion of residency was rescaled toreflect a 10% increase.

Results

Screening rates

Table 1 shows that 29.1% (n5 34) of the sample(N5 117) have received FOBT within the past 12months. In addition, 35.9% (n5 42) were adherentto SIG, whereas 41.9% (n5 49) were adherent toCOL. Overall, 61.5% (n5 72) were adherent to any

of the three tests (FOBT, SIG, or COL), while38.5% (n5 45) were not adherent. Of those whoare adherent, approximately 34% of the partici-pants actually received two or three tests within therecommended time interval.

Demographic characteristics

Forty-two men (35.9%) and 75 women (64.1%)were between 50 and 86 years old with a mean age

of 61.02 years (SD5

7.63) (Table 2). Those whoare adherent were significantly older (meanage5 63.08) than those who were not adherent(mean age5 57.76), po0.001. Majority of thesample completed higher education, in which 50and 20% have college and graduate degrees,respectively.

Among the Filipino sample, about 55% (n5 64)were currently employed, and 45% (n5 52) werenot employed (i.e. retired, homemakers or cur-rently unemployed). The annual income for 30%(n532) of Filipinos waso$30 000, whereas 70%(n576) have an annual income of $30 000. Gender,academic completion, employment, and annualincome were not significantly different betweencolorectal adherence groups.

Healthcare characteristics

In Table 2, 93.1% (n5 108) reported that theywere covered by medical insurance, but only 62%(n570) were adherent to colorectal cancer screen-ing. Additionally, of those who have receivedroutine medical examination within the past 12months (n5 102), only about two-thirds (n5 60)

reported having had any colorectal cancer testwithin the ideal time interval. Approximately, 68%(n579) of the participants were recommended bytheir doctor to undergo colorectal cancer screening,but only 60 individuals reported current adherence.There were statistically significant differences forhaving health insurance (p5 0.005), receivingmedical examination in the past year (p5 0.004)and doctors recommendation (po0.001) betweenadherent and not-adherent groups.

Cultural characteristics and bivariate results

Table 3 illustrates that on average, the participantshave lived in the U.S 42.7% of their lifetime. Thosewho were adherent lived in the U.S. approximately48.6% of their lifetime. In contrast, the proportion

Table 1. Colorectal cancer screening receipt among Filipinosin the United States (N5117)

Total Not

adherent

Adherent

FOBT 117 83 (70.9) 34 (29.1)

SIG 117 75 (64.1) 42 (35.9)COL 117 68 (58.1) 49 (41.9)

Any of the above tests

(FOBT/SIG/COL)

117 45 (38.5) 72 (61.5)

Adherent to one test 32 (27.4)

Adherent to two tests 27 (23.1)

Adherent to three tests 13 (11.1)

Value within parenthesis represents percentage.

R. R. Ferrer et al.


DOI: 10.1002/pon


5/9

of lifetime spent in the U.S among the not-adherentsample (33.2%) was significantly lower, Po0.001.

The language preference average score in thesample population was 3.13. This suggests that theparticipants are able to speak, read and write, andprefer both English and Filipino languages equallywell (bicultural identified). Language preference was

not significantly different between adherent andnot-adherent groups.The participants have a high perceived internal

control over health problems (mean5 4.02), whichsuggests that finding health problems early is ones

responsibility. Expectedly, perceived external con-trol is low (mean5 2.24), which means that theparticipants tended to disagree that finding healthproblems early is controlled by others or chance.

For internal control, there was no significantdifference between adherent (mean5 4.08) and not-adherent (mean5 3.89) groups (p5 0.361). Thedifference in external control between adherent(mean5 2.14) and not-adherent (mean5 2.43)

groups was marginally significant (p5 0.085).

Predictors of screening adherence andmultivariate results

Table 4 shows two multivariable models, one isadjusted for doctors recommendation and theother is not. In both the models, we found someconsistent results. For every 5-year increase in age,the odds of being adherent to any colorectal cancerscreening significantly doubled (Model 1:

OR5

2.10, 95% CI5

1.26-3.51, p5

0.005; Model2: OR52.1, 95% CI5 1.293.41, p5 0.003). Theseeffects were adjusted for demographic, healthcareand cultural variables. Perceived internal controlwas not significantly associated with colorectal

Table 3. Cultural characteristics and colorectal cancer screen-ing adherence

Total Not-

adherent

Adherent p

Cultural factorsa

Lifetime proportion

residency in the U.S.

0.43 (0.23) 0.33 (0.22) 0.49 (0.22) o0.001

Language preference 3.13 (0.43) 3.06 (0.45) 3.17 (0.42) 0.359

Internal control 4.01 (1.14) 3.89 (1.16) 4.08 (1.12) 0.361

External control 2.26 (0.88) 2.43 (0.96) 2.14 (0.82) 0.085

Values are represented as mean (SD).aAnalysis of variance test.

Table 2. Demographic and healthcare characteristics

Total Not-adherent Adherent p


Agea o0.001

Mean (years) 61.02 57.73 63.08

SD 7.63 7.23 7.18Gender 0.648

Male 42 (35.90) 15 (12.8) 27 (23.1)

Female 75 (64.10) 30 (25.6) 45 (38.5)

Academic completion 0.120

High School/associate/some college 35 (29.9) 11 (9.4) 24 (20.5)

College degree 59 (50.4) 28 (23.9) 31 (26.5)

Graduate degree 23 (19.7) 6 (5.1) 17 (14.5)

Employment status 0.224

Not employed 52 (44.8) 17 (14.7) 35 (30.2)

Employed 64 (55.2) 28 (24.1) 36 (31.0)

Annual income 0.471

o$30 K 32 (27.3) 14 (12.0) 18 (15.4)

X$30 K 85 (72.6) 31 (26.5) 54 (46.2)

Healthcare characteristics

Health insuranceb 0.005

Yes 108 (93.1) 38 (32.8) 70 (61.90)

No 8 (6.9) 7 (6.0) 1 (0.90)

One regular doctor 0.833

Yes 64 (54.7) 25 (21.4) 39 (33.3)

No 53 (45.3) 20 (17.1) 33 (28.2)

Routine medical examinationb 0.004

Yes 102 (88.7) 33 (28.7) 69 (60.0)

No 13 (11.3) 10 (8.7) 3 (2.6)

Doctor recommended screening o0.001

Yes 79 (67.50) 19 (16.2) 60 (51.3)

No 38 (32.50) 26 (22.2) 12 (10.3)

Some variables do not total to 100% (N5 117) due to missing responses. Value within parenthesis represents percentage.aAnalysis of variance test.bFishers exact test, all other variables examined using w2test.



DOI: 10.1002/pon


6/9

cancer screening adherence after adjusting forcovariates. However, as perceived external controlincreased, the odds of current adherence decreased(Model 1: OR5 0.53, 95% CI5 0.271.03, p50.06; Model 2: OR50.58, 95% CI5 0.311.08,

p5 0.08).In model 1, the effect of doctors recommenda-

tion on screening was strong and positive. Com-pared with those who have not receivedrecommendation, those who have were over fivetimes more likely to report current adherence(OR5 5.54, 95% CI5 1.8216.87).

There were inconsistent results in the effect oflifetime proportion in the U.S. In model 1,

increased lifetime proportion in the U.S.(OR5 1.02, 95% CI5 1.001.05) was not asso-ciated with screening. In terms of languagepreference, being bicultural identified (OR5 0.54,95% CI5 0.122.43) or western identified(OR5 0.67, 95% CI5 0.123.71) was not asso-ciated with adherence to colorectal cancer screen-ing. Similarly in model 2, bicultural and westernidentification had no effect on screening.

Discussion


This pilot study adds to the knowledge regardingfactors associated with colorectal cancer screeningamong Filipino Americans. The findings show thatincreasing age is an important factor for currentadherence, consistent with a comprehensive reviewof literature [14]. However, the impact of age oncolorectal cancer screening was only consistentwith some Asian subgroups [5,33,44] but not withothers [19,27]. Despite this inconsistency, it is stillimportant to increase efforts to recruit Filipino

Americans beginning at age 50 for colorectalcancer screening studies.There are useful methods to recruit and retain

Filipino American participants, which is a hard-to-reach population. This pilot study began to explore

some of the culturally sensitive approach, whichincluded inviting community liaisons and personalcontacts to assist with subject recruitment. How-ever, other strategies may also be employed infuture studies such as performing key informantinterviews as a process of identifying communityleaders for recruitment leads; and targeting com-munity based organizations, churches, and Filipinogrocery stores as recruitment sites [36,45].

Having insurance, routine medical exam and/orreceived doctors recommendation predicted ad-herence to colorectal cancer screening for someAsian American groups [26,27,46]. In this study,Filipinos who received doctors recommendation

were more than five times likelier to reportadherence. This further validates the importantrole of physicians or other appropriate health careproviders to encourage people at the right age toscreen for colorectal cancer.

Having regular doctor and health insurance wereinsignificant predictors of colorectal cancer screen-ing adherence [34,35]. However, this study isunable to present any findings regarding theseassociations because some variations in responseswere insufficient to allow for an accurate estima-tion. Additional studies are, therefore, needed.

Cultural characteristics

The main purpose of the study is to explorecultural factors associated with colorectal cancerscreening. The lack of significant relationshipbetween language preference and screening receiptamong Filipinos is supported by previous literature[34]. Perhaps, proficiency in English language doesnot pose a barrier to Filipino Americans, giventhat the Philippines considers English as an officiallanguage.

It is important to emphasize that language isonly one dimension of acculturation explored inthis study. Therefore, additional research is stillconsidered necessary to further describe the func-tion of other dimensions, such as behaviors,

Table 4. The impact of cultural factors on colorectal cancer screening among Filipinos, N5116

Model 1 Model 2

OR SEa p 95% CI OR SEa

p 95% CI

Age (5-year increase) 2.10 0.55 0.005 1.253.51 2.10 0.52 0.003 1.293.41

Doctors recommendation 5.54 3.14 0.003 1.8216.88

Lifetime proportion in the U.S. (10% increase) 1.02 0.02 0.174 1.001.05 1.04 0.01 0.011 1.011.06

Cultural health beliefs

Internal control 1.08 0.27 0.759 0.661.77 1.19 0.28 0.464 0.751.87

External control 0.53 0.18 0.063 0.271.03 0.58 0.18 0.088 0.311.08

Language proficiency

Filipino identified (Reference)

Bicultural identified 0.54 0.41 0.421 0.122.43 0.74 0.53 0.676 0.183.00

Western identified 0.67 0.59 0.652 0.123.71 0.76 0.62 0.733 0.153.78

Adjusted for gender, education, employment status, income, and having one regular doctor. One case was dropped out of the analysis due to missing responses.aSE, standard error.

R. R. Ferrer et al.


DOI: 10.1002/pon


7/9

identity, friendship choice, generation level andattitude [38,40].

Lifetime proportion in the U.S. among Koreanand Filipino women predicted cervical, breast andcolorectal cancer screening procedures [34]. Simi-larly, a 10% increase of lifetime spent in the U.S.

increased the odds of current SIG or COL screen-ing receipt by 1.3 times among Filipino Americanimmigrants [35]. Interestingly, this study revealedthat there is no relationship between length ofimmigration and screening adherence after adjust-ing for other variables, particularly doctorsrecommendation. However, when the effect ofdoctors recommendation was removed in themultivariate model, a significant relationship wasfound.

These interesting yet conflicting findings suggesttwo possible mechanisms through which doctors

recommendation affect the relationship betweenimmigration and screening behaviors. On onehand, previous studies [34,35] did not measuredoctors recommendation to undergo colorectalscreening as a determinant to adherence, anduncontrolled confounding may have led to spur-ious results. On the other hand, the strong positiveassociation between doctors recommendation andscreening may reflect a mediating effect. It isinteresting that, in the bivariate analyses, lifetimeproportion of residency in the U.S. has a strongsignificant association with screening; yet, thisrelationship completely disappears in the regres-sion model. Therefore, it is possible that length ofresidency in the U.S. predicts healthcare utilizationpatterns and increased physician visits, which thenincreases the probability of receiving doctorsrecommendation for screening. This complexrelationship then may predict actual screeningbehaviors. This relationship should be investigatedfurther in future studies.

Cultural perceptions of health outcomes wereexplored because these factors tend to influencehealth behaviors among ethnic minority groups[42]. For example, fatalism is a belief that is shared

by different cultures like Asians and AfricanAmericans. In the context of health-related beha-viors, this is described as the belief that powerfulothers (e.g. God) and/or chance take control overones health outcome [4749].

To the authors knowledge, this study is first toexplore and report on the role of culturallyinformed health beliefs on colorectal cancer screen-ing among Filipino Americans. Although theadapted measure was initially validated for breastcancer screening, its use based on the items ofinternal control and external control subscales is

justified. The reliability of these two subscales forthe current sample was acceptable.The marginally significant association found

between external control and screening adherencesuggests that higher perception of external control

decreases the likelihood to be currently screenedfor colorectal cancer. Perhaps, Filipinos do notneed to rely on the influence of others in order toundergo colorectal cancer testing based on recom-mended guidelines.

Moreover, it is reasonable that since the sample

reported low external control, their internal controlwould be high. Yet, in the multivariate analysisinternal control had no significant impact onscreening outcome. Perhaps, the two constructsare independent of each other. In fact, the resultssuggest that perceptions of external control mayhave a greater influence on screening behaviors.

The authors found no previous literature thatoffered a suitable explanation for this trend.Instead, two studies using an identical measure ofhealth locus of control actually reported incon-sistent results. In one study, participants highly

endorsed internal locus and external locus ofcontrol [50]. Yet, in another study of Filipinomigrant workers, only the external locus of controlperception was high [51].

The perception of control regarding healthamong the Filipinos may depend on the type andseverity of health problems. For certain, less severehealth issues, it may be perceived that it is up tothem to reduce the burden. While, for other typesof diseases such as terminal illnesses of advancedstages, powerful others or chance dictate theirhealth outcomes.

In the interest of colorectal cancer screeningresearch, a qualitative research method should beconsidered in future studies. For example, a focusgroup aimed at investigating the underlying per-ceptions surrounding cultural health beliefs, parti-cularly perceived control may clarify what it meansfor Filipinos to endorse internal control or externalcontrol.

Limitations

This study has limitations that necessitate prudencein interpretation of results. For example, self-

report surveys are prone to response bias. Inaddition, changes in screening behaviors over timecannot be determined using cross-sectional designs.Therefore, factors associated with adherence needto be confirmed using longitudinal data.

Convenience, purposive, and snowball samplingmethods limit generalizability to the Filipinopopulation. The present sample had a somewhathomogeneous grouping in demographic andhealthcare characteristics. As a result, severalvariables had small cell sizes and, therefore, wereexcluded in the multivariate analysis. Nevertheless,

the sampling methods appear to be appropriatewhen recruiting and maintaining hidden, hard-to-reach participants.

The authors acknowledge potential issuesregarding cross-cultural validity of the measures



DOI: 10.1002/pon


8/9

adapted in this study. However, the reliabilitycoefficients for the cultural characteristics subscaleswere reasonable. Sample size also limited the powerin detecting effects. This was observed in themarginally statistically significant association be-tween external control and screening. With a larger

sample, it could be expected that this relationshipwould reach statistical significance.

Finally, using several sampling strategies resultedin different sources of response. For example, onexamining mailed-in surveys, there were unexpectedresponses from several U.S. states, outside theinitial targeted region of Southern California. Forthis study, however, source of response (e.g.informal social gathering) and state of residencewere not monitored and compared across groups.Future research with larger Filipino sample sizeshould consider investigating possible differences

between source of data and colorectal cancerscreening.

References

1. Horner M, Ries L, Krapcho M et al. Seer CancerStatistics Review, 19752006. National Cancer Institute,Bethesda, MD. Available from: http://seer.cancer.gov/csr/1975_2006/, based on November 2008 SEER datasubmission, posted to the SEER web site, 2009.

2. Levin B, Lieberman DA, McFarland B et al. Screeningand surveillance for the early detection of colorectal

cancer and adenomatous polyps, 2008: A joint guidelinefrom the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and theAmerican College of Radiology. Gastroenterology2008;134:15701595.

3. Winawer S, Fletcher R, Rex D et al . Colorectalcancer screening and surveillance: clinical guidelinesand rationale-update based on new evidence. Gastro-enterology 2003;124:544560.

4. Shapiro JA, Seeff LC, Thompson TD et al. Colorectalcancer test use from the 2005 national health interviewsurvey. Cancer Epidemiol Biomarkers Prev 2008;17:16231630.

5. Walsh JM, Kaplan CP, Nguyen B, Gildengorin G,McPhee SJ, Perez-Stable EJ. Barriers to colorectal

cancer screening in Latino and Vietnamese Americanscompared with non-Latino White Americans. J GenIntern Med 2004;19:156166.

6. Sirovich BE, Schwartz LM, Woloshin S. Screening menfor prostate and colorectal cancer in the United States:does practice reflect the evidence? J Am Med Assoc2003;289:14141420.

7. Codori AM, Petersen GM, Miglioretti DL, Boyd P.Health beliefs and endoscopic screening for colorectalcancer: potential for cancer prevention. Prev Med 2001;33:128136.

8. Lemon S, Zapka J, Puleo E, Luckmann R, Chasan-Taber L. Colorectal cancer screening participation:comparisons with mammography and prostate-specificantigen screening. Am J Public Health 2001;91:

12641272.9. Jeanette MT, Bruce SL, Joel LW, Robert ES, Mutlu H.

Using the transtheoretical model to stage screeningbehavior for colorectal cancer. Health Educ Behav 2003;30:322336.

10. Costanza ME, Luckmann R, Stoddard AM et al.Applying a stage model of behavior change to coloncancer screening. Prev Med 2005;41:707719.

11. Manne S, Markowitz A, Winawer S et al. Correlates ofcolorectal cancer screening compliance and stage ofadoption among siblings of individuals with early onsetcolorectal cancer. Health Psychol 2002;21:315.

12. Myers RE, Vernon SW, Tilley BC, Lu M, Watts BG.Intention to screen for colorectal cancer among Whitemale employees. Prev Med 1998;27:279287.

13. Slattery ML, Kinney AY, Levin TR. Factors associatedwith colorectal cancer screening in a population-basedstudy: the impact of gender, health care source, andtime. Prev Med 2004;38:276283.

14. Subramanian S, Klosterman M, Amonkar MM, HuntTL. Adherence with colorectal cancer screening guide-lines: a review. Prev Med 2004;38:536550.

15. Brenes GA, Paskett ED. Predictors of stage of adoptionfor colorectal cancer screening. Prev Med 2000;31:410416.

16. Cibula DA, Morrow CB. Determining local colorectalcancer screening utilization patterns. J Public HealthManag Pract 2003;9:315321.

17. Greiner KA, James AS, Born W et al. Predictors offecal occult blood test (FOBT) completion amonglow-income adults. Prev Med 2005;41:676684.

18. Hay JL, Ford JS, Klein D et al. Adherence to colorectalcancer screening in mammography-adherent olderwomen. J Behav Med 2003;26:553576.

19. Sun WY, Basch CE, Wolf RL, Li XJ. Factors associatedwith colorectal cancer screening among Chinese-Amer-icans. Prev Med 2004;39:323329.

20. Brawarsky P, Brooks DR, Mucci LA. Correlates ofcolorectal cancer testing in Massachusetts men andwomen. Prev Med 2003;36:659668.

21. Holmes-Rovner M, Williams GA, Hoppough S,

Quillan L, Butler R, Given CW. Colorectal cancerscreening barriers in persons with low income. CancerPract 2002;10:240247.

22. Swan J, Breen N, Coates RJ, Rimer BK, Lee NC.Progress in cancer screening practices in the UnitedStates: results from the 2000 National Health InterviewSurvey. Cancer 2003;97:15281540.

23. Weinberg DS, Turner BJ, Wang H, Myers RE, Miller S.A survey of women regarding factors affecting color-ectal cancer screening compliance. Prev Med 2004;38:669675.

24. Theisen C. In different cultures, cancer screeningpresents challenges. J Natl Cancer Inst 2004;96:1012.

25. Ashing-Giwa K, Kagawa-Singer M. Infusing cultureinto oncology research on quality of life. Oncol Nurs

Forum 2006;33:3136.26. Teng EJ, Friedman LC, Green CE. Determinants of

colorectal cancer screening behavior among ChineseAmericans. Psycho-Oncology 2006;15:374381.

27. Juon HS, Han W, Shin H, Kim KB, Kim MT. Predictorsof older Korean Americans participation in colorectalcancer screening. J Cancer Educ 2003;18:3742.

28. Reeves TJ, Bennett CE. We the people: Asians in theUnited States. Census 2000 Special Report 2004, UnitedStates Census Bureau.

29. Sue DW, Sue D (eds). Counseling the Cultural Diverse:Theory and Practice (4th edn). Wiley: New York, NY,2003.

30. Flood DM, Weiss NS, Cook LS, Emerson JC,Schwartz SM, Potter JD. Colorectal cancer incidence

in Asian migrants to the United States and theirdescendants. Cancer Cause Control 2000;11:403411.

31. Lin SS, Clarke CA, Prehn AW, Glaser SL, West DW,OMalley CD. Survival differences among Asiansubpopulations in the United States after prostate,

R. R. Ferrer et al.


DOI: 10.1002/pon


9/9

colorectal, breast, and cervical carcinomas. Cancer2002;94:11751182.

32. Maxwell AE, Crespi CM. Trends in colorectal cancerscreening utilization among ethnic groups in California:are we closing the gap? Cancer Epidemiol BiomarkersPrev 2009;18:752759.

33. Wong ST, Gildengorin G, Nguyen T, Mock J. Disparities

in colorectal cancer screening rates among Asian Ameri-cans and non-Latino Whites. Cancer 2005;104:29402947.

34. Maxwell AE, Bastani R, Warda US. Demographicpredictors of cancer screening among Filipino andKorean immigrants in the United States. Am J PrevMed 2000;18:6268.

35. Maxwell AE, Danao LL, Crespi CM, Antonio C,Garcia GM, Bastani R. Disparities in the receipt offecal occult blood test versus endoscopy among FilipinoAmerican immigrants. Cancer Epidemiol BiomarkersPrev 2008;17:19631967.

36. Maxwell AE, Bastani R, Vida P. Strategies to recruit andretain older Filipino-American immigrants for a cancerscreening study. J Commun Health 2005;30:167179.

37. Suinn RM, Rickard-Figueroa K, Lew S, Vigil P. The

Suinn-Lew Asian Self-Identity Acculturation Scale: aninitial report. Educ Psychol Meas 1987;47:401407.

38. Ownbey SF, Horridge PE. The Suinn-Lew Asian Self-identity Acculturation Scale: test with a non-student,Asian-American sample. Soc Behav Pers 1998;26:5768.

39. Suinn RM, Ahuna C, Khoo G. The Suinn-Lew AsianSelf-Identity Acculturation Scale: concurrent and factor-ial validation. Educ Psychol Meas 1992;52:10411046.

40. Ponterotto JG, Baluch S, Carielli D. The Suinn-LewAsian Self-Identity Acculturation Scale (SL-ASIA).Meas Eval Couns Dev 1998;32:109124.

41. Russell KM, Champion VL, Perkins SM. Developmentof cultural belief scales for mammography screening.Oncol Nurs Forum 2003;30:633640.

42. American Cancer Society. Colorectal Cancer Facts andFigures (Special edition). American Cancer Society:Atlanta, GA, 2005. (Available from: http://www.cancer.org/downloads/STT/CAFF2005CR4PWSecured.pdf).

43. American Cancer Society. Colorectal Cancer Facts andFigures (Special edition), 2005. Retrieved 10 April 2006.Available from http://www.cancer.org/downloads/STT/

CAFF2005CR4PWSecured.pdf.44. Honda K. Factors associated with colorectal cancer

screening among the US urban Japanese population.Am J Public Health 2004;94:815822.

45. Ko CM et al. Filipina American womens breast cancerknowledge, attitudes, and screening behaviors. BMCPublic Health 2003;3:2732.

46. Honda K. A model of perceived risk for colorectalcancer among Japanese Americans. J Cancer Educ2004;19:251257.

47. Ashing-Giwa K. Health behavior change models andtheir socio-cultural relevance for breast cancer screeningin African American women. Women Health 1999;28:5371.

48. Chng CL, Collins JR. Providing culturally competent

HIV prevention programs. Am J Health Studies 2000;16:2432.

49. Liang W, Yuan E, Mandelblatt JS, Pasick RJ. How doolder Chinese women view health and cancer screening?Results from focus groups and implications for inter-ventions. Ethnic Health 2004;9:283304.

50. Malcarne VL, Fernandez S, Flores L. Factorial validityof the multidimensional health locus of control scalesfor three American ethnic groups. J Health Psychol2005;10:657667.

51. Holroyd EA, Molassiotis A, Taylor-Pilliae RE. Filipinodomestic workers in Hong Kong: health relatedbehaviors, health locus of control and social support.Women Health 2001;33:181205.



DOI: 10 1002/pon

Documents

Ferrer, Ramirez Et Al (2010)