Determinants of colorectal cancer screening behavior among Chinese Americans

PSYCHO-ONCOLOGY

Psycho-Oncology 15: 374–381 (2006)Published online 5 September 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/pon.958

DETERMINANTS OF COLORECTAL CANCERSCREENING BEHAVIOR AMONG CHINESE

AMERICANS

ELLEN J. TENGa,b,c,*, LOIS C. FRIEDMANc and CHARLES E. GREENd

aMichael E. DeBakey VA Medical Center, Houston, USAbVeterans Affairs South Central Mental Illness Research, Education, and Clinical Center, 116 MHCL,

2002 Holcombe Blvd, Houston, Tx 77030 USAcThe Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine,

One Baylor Plaza, MS 350, Houston, TX 77030 USAdUniversity of Texas Medical School, 6431 Fannin Street, Houston, Tx 77030 USA

SUMMARY

Colorectal cancer (CRC) is the most commonly diagnosed cancer among Chinese Americans and is the third leadingcause of cancer death in this population. The objectives of this study were to determine the rates of CRC screening(via fecal occult blood test (FOBT), flexible sigmoidoscopy (FSIG), and colonoscopy) among Chinese Americansand predictors of utilizing these screening procedures. Participants ðN ¼ 206Þ completed a self-administeredquestionnaire assessing cancer screening behaviors and beliefs about perceived risk of developing cancer andtreatment efficacy. A series of logistic regressions indicated that physician recommendation to obtain CRC screeningsignificantly predicted whether Chinese Americans undergo FOBT, FSIG, or colonoscopy screening ðp50:001Þ.Acculturation and perceived risk of developing CRC did not predict obtaining any of the screening procedures.FOBT was the most commonly reported screening method used by respondents (65%), followed by FSIG (54%)and colonoscopy (49%). These findings highlight the need to make physicians more aware of the impact theirrecommendations have in determining CRC screening behavior among Chinese Americans. Copyright# 2005 JohnWiley & Sons, Ltd.

KEY WORDS: oncology; cancer; Asian Americans; Chinese; colorectal screening

INTRODUCTION

Cancer continues to be the leading cause of deathamong Asian American women (Parker et al.,1998) and is a growing problem for Asian Americanmen (Miller et al., 1996). Despite the fact that earlycancer detection can save lives, research consistentlyhas shown that Asian Americans utilize cancerscreening less than other ethnic/racial subgroups,including African Americans and Caucasians(Couglin and Uhler, 2000; Centers for DiseaseControl, 1999; Parker et al., 1998). Current color-ectal cancer (CRC) screening guidelines (AmericanCancer Society, 2005) specify that men and women

should use one of the following screening proce-dures beginning at age 50: Fecal occult bloodtesting (FOBT) every year, flexible sigmoidoscopy(FSIG), double-contrast barium enema (DCBE)every 5 years, or colonoscopy every 10 years. Theefficacy of these various procedures has beenexamined through case–control (Muller and Son-nenberg, 1995; Selby et al., 1992; Winawer et al.,1993) and randomized controlled studies (Hard-castle et al., 1996; Mandel et al., 1993; Kronborget al., 1996). FOBT and FSIG have been shown toreduce CRC mortality by up to 33% (Mandelet al., 1993) and 80% (Muller and Sonnenberg,1995; Selby et al., 1992), respectively.

In contrast, little data are available regardingthe efficacy of DCBE. In addition to the discom-fort typically associated with the procedure (Kahiand Rex, 2004), Winawer et al. (2000) found it tobe less sensitive in detecting polyps compared to

Received 25 August 2004Copyright # 2005 John Wiley & Sons, Ltd. Accepted 10 June 2005

*Correspondence to: Department of Psychology, SouthernIllinois University Carbondale, Mail Code 6502, Carbondale,IL 62901, USA. E-mail: [email protected]

other procedures including colonoscopy. Althoughcurrently there is no evidence for the efficacy ofcolonoscopy from randomized controlled trials, arecent study indicates that a single colonoscopy ismore effective at detecting pre-malignancies thanFOBT and FSIG combined (Lieberman andWeiss, 2001).

Despite the availability of effective CRC screen-ing methods, Ioannou et al. (2003) reported thatless than half (43%) of the United States popula-tion used these procedures (i.e. FOBT in past year,and/or colonoscopy or FSIG in past 5 years).Utilization is less among specific subgroups, withAsian/Pacific Islanders reporting a CRC screeningrate of 34.8% (Ioannou et al., 2003). Furthermore,much of the existing research treats Asian Amer-icans as a homogenous group, when in fact thisgroup is highly diverse and has cancer incidencerates that vary considerably from one subgroup toanother. Relatively few studies have examinedcancer screening practices among specific AsianAmerican subgroups, and even less attention hasbeen directed at CRC screening despite the factthat CRC is the most commonly diagnosed canceramong Chinese Americans and is the third leadingcause of cancer death in this population (Milleret al., 1996).

Despite the high incidence of this potentiallycurable cancer, little is known about CRC screen-ing behaviors among Chinese Americans. Studiesof the general population of the United States havefound that certain demographic variables (e.g.education, income and marital status) predictCRC screening utilization (Anderson and May,1995; Lemon et al., 2001; Polednak, 1990;Richardson et al., 1995; Ruffin et al., 2000;Weinrich, 1990). With the exception of a fewstudies, relatively little data are available regardingCRC screening practices of Chinese Americans.Recent evidence (Sun et al., 2004) suggests thatChinese individuals are more likely to obtainFOBT screening after immigrating to the UnitedStates. Sun et al. postulate that this increase inscreening behavior may be related to the recentimprovements in China with regard to educatingthe general public and making screening (e.g.FOBT) more accessible. Although this is anencouraging finding, the majority of the extantliterature indicates that CRC screening practicesamong Chinese Americans are significantly lowerthan that of the general US population. Forexample, Yu et al. (2001) found that only 15%of the Chinese American men and women

surveyed in Chicago had ever been screened witha FOBT. Slightly higher screening rates werereported by Tang et al. (2001), who found that25% of the Chinese American women theysampled had ever obtained a FOBT and 31%had ever obtained a FSIG. Tang et al. also foundthat acculturation and physician recommendationpredicted use of FOBT and FSIG. However, thesedata are limited in their generalizability due to thesmall sample size and the exclusion of men fromthe study. Also, the sample was composed of low-income women generally low in acculturation,further limiting its applicability to the entireChinese–American population.

In summary, CRC often can be treated success-fully if detected early. Because of the highincidence and mortality rates of CRC amongChinese Americans (Miller et al., 1996; Parkeret al., 1998), it is important that they obtainregular screening. With the exception of the fewstudies that have been conducted with thispopulation (e.g. Tang et al., 2001; Yu et al.,2001) relatively few studies have examined the useof various CRC screening methods among ChineseAmericans, and no study to date has investigatedthe use of colonoscopy among this population.

The purpose of the present study was (1) toexamine CRC screening behaviors among ChineseAmericans, using a larger and more generalizablesample than past studies, and (2) to determinepredictors of CRC screening (including FOBT,FSIG, and colonoscopy) among Chinese Amer-icans. Due to the declining use of DCBE (Kahiand Rex, 2004), this screening procedure wasnot evaluated in the current study. Based onpast findings (Tang et al., 2001), we hypothesizeda positive relationship between acculturationand utilization of CRC screening, and alsothat physician recommendation would be thebest predictor of utilization of these screeningprocedures.

METHOD

Participants

Participants were 206 Chinese Americans (menand women) recruited from senior centers andchurches in San Francisco ðn ¼ 123Þ and Houstonðn ¼ 83Þ. Recruitment occurred at these twolocations because they are among the top eightcities in the country that have the largest Asian

COLORECTAL CANCER SCREENING BEHAVIOR AMONG CHINESE AMERICANS 375

Copyright # 2005 John Wiley & Sons, Ltd. Psycho-Oncology 15: 374–381 (2006)

population (US Census Bureau, 2000), and re-search personnel were available at both locationsto assist with the study. As the current study waspart of a larger project designed to evaluate ratesof breast and uterus screening, persons completingthe survey were 18 years or older. The responserate of those asked to complete the survey was89%. Of those who completed the survey, 12individuals reported a history of either breast,cervical, colon, or prostate cancer and wereexcluded from the analyses, yielding a final sampleof 194 (86 men and 108 women) individuals. Themean age of this final sample was 57.55 yearsðS:D: ¼ 14:01Þ. Seventy-five percent of those whoreported their marital status were married and25% were either single or widowed. The majority(82.9%) of participants reported having been bornoutside the United States. Average level ofacculturation, according to the Brief AcculturationScale (Burnam et al., 1987), was medium(M ¼ 2:26, S.D.=0.51). The majority of thesample (60.2%) reported their language preferenceto be both English and Chinese. Twenty-sixpercent of participants preferred primarily Englishand 13% preferred primarily Chinese. Fifteenpercent of those reporting had completed fewerthan 12 years of education, 34% reported havingcompleted high school and 51% had obtained acollege degree or more. The present study wasapproved by the Baylor College of MedicineInstitutional Review Board. Participants wererecruited from Fall, 2002 through Summer, 2003.

Materials

Health questionnaire. Participants in the presentstudy responded to a three-page self-administeredquestionnaire assessing cancer-screening proce-dures recommended by the American CancerSociety (2005). This questionnaire was modeledafter surveys used in past studies (e.g. Friedmanet al., 1995) assessing compliance with health-related behaviors. The questionnaire in the currentstudy was written in both English and Chinese,with the English version printed on one side of apage and the Chinese version on the back side of apage. Participants were instructed to respond tothe survey using the language format with whichthey were most comfortable. Using a ‘yes’ or ‘no’format, respondents indicated if they ever had anFOBT, FSIG or colonoscopy. To ensure thatparticipants understood the procedures associated

with these names, each procedure was describedusing everyday language. For example, questionsrelated to FSIG were prefaced with, ‘A flexiblesigmoidoscopy is a procedure during which adoctor checks inside the rectum and lower bowelwith a flexible lighted tube to look for cancer orpolyps’. Questions related to colonoscopy wereprefaced with, ‘A colonoscopy is a proceduresimilar to the flexible sigmoidoscopy. In a colono-scopy, a longer lighted tube is used that allows thedoctor to see the entire colon. If polyps are foundthey can be removed’. Perceived risk of developingCRC was assessed with the question, ‘What doyou think your chances are of getting colon cancersome day’? Response options ranged from 1 (nochance) to 4 (high). Physician recommendation ofthe three CRC screening procedures was assessedusing a response scale ranging from 0 (stronglydisagree) to 4 (strongly agree). For example,participants responded to the statement, ‘Mydoctor strongly encourages me to get acolonoscopy’. A 5-point scale was used for thisseries of questions in the event that participantsheld a neutral position.

Health history items asked whether participantshad a medical examination within the precedingyear and whether they had a regular doctor formedical care. Questions regarding personal andfamily histories of colorectal cancer also wereincluded. Information on age, marital status,ethnicity, education and acculturation level wereobtained from the questionnaire.

Brief Acculturation Scale. Shortened from theoriginal 26-item inventory (Burnam et al., 1987),this abbreviated measure of acculturation (alsopresented in Chinese format in the current study),estimates acculturation to Western culture fromgeneration, language and social activity prefer-ences. Generation was assigned a numerical value(e.g. first generation received a score of 1; secondgeneration received a score of 2). Languagepreference was reported on a 5 point Likert scale,ranging from 1 (only English) to 5 (only Chinese).The same scale was used to indicate preference forsocial activities, with scores ranging from 1(Chinese only) to 5 (Only non-Chinese). Afterreversing scale scores for the language preferenceitem, each person’s total score was summed acrossthe three items and divided by the number of itemschecked. In accordance with the scoring guidelinesrecommended by Burnam et al. (1987), scores

E.J. TENG ET AL.376


ranging from 1 to 1.75, 1.76 to 3.25 and 3.26 to 5reflected low, medium and high levels of accultura-tion, respectively. The scale has demonstratedgood internal reliability (alpha=0.97; Burnamet al., 1987).

Design/Procedure

Permission was requested by the research staff ateach recruitment site before distributing surveys.Participants were accessed via senior centers andchurch organizations in locations largely popu-lated by Chinese Americans. In these large groupsettings, a brief introduction was made to informpotential participants that the study was an effortto examine health-related behaviors among Chi-nese Americans and that participation was volun-tary and anonymous. Interested individuals werethen handed a survey to complete, which were thencollected by research staff.

Descriptive statistics for the sample are pre-sented in Table 1. A series of logistic regressionsexamined whether participants over the age of 50had ever obtained an FOBT, FSIG or colono-scopy. Analyses were conducted according tocurrent CRC screening recommendations outlinedby the American Cancer Society (2005). Since we

were interested in determining if respondents hadever obtained a specific screening procedure withinthe recommended time frame (e.g. FSIG every 5years), we used the following age cutoffs inanalyzing each screening method: the FOBTanalysis included persons 50 years or more; theFSIG analysis included persons 55 years or more,and the colonoscopy analysis included persons 60years or more. Hierarchical entry of predictorvariables included acculturation, risk perception,and physician recommendation of these screeningprocedures. These variables were dichotomizeddue to insufficient cell sizes. Risk perception wasdivided into two groups consisting of ‘no chance/low’ and ‘moderate/high’. Responses to itemspertaining to physician recommendation of var-ious procedures were separated into two groupsconsisting of ‘strongly agree/agree’ and ‘neutral/disagree/strongly disagree’.

RESULTS

CRC screening rates by location

Chi-square analyses were conducted on CRCscreening rates of Chinese American men andwomen according to their geographical location

Table 1. Descriptive information of sample

Variable n %

Family history of colorectal cancer 25 13.2

Regular physician for medical care (‘yes’) 172 90.1

Medical examination (past year) 168 88.9

Perceived risk of developing colorectal cancer

Moderate/high 31 18.9

No chance/low 133 81.1

Benefits of early detection

Strongly agree/agree 66 86.9

Neutral/disagree/strongly disagree 10 13.1

Physician recommendation

Fecal occult blood test}strongly agree/agree 94 55

Flexible sigmoidoscopy}strongly agree/agree 86 51.5

Colonoscopy}strongly agree/agree 97 57.4

Screening behaviors (at least once)

Fecal occult blood testa 80 65.0

Flexible sigmoidoscopyb 47 54.0

Colonoscopyc 29 49.2

aParticipants over 50 years of age.bParticipants over 55 years of age.cParticipants over 60 years of age.



(i.e. San Francisco vs Houston). As shown inTable 2, no significant differences were found inthe rate of FOBT, FSIG, and colonoscopy screen-ing among these individuals.

FOBT screening

FOBT screening was evaluated for men andwomen over the age of 50 who did not report ahistory of colon cancer ðn ¼ 123Þ. Seventy-oneparticipants were excluded from the analysis basedon these inclusion criteria. Sixty-four percent ofthis sample (29% men; 35% women) reportedhaving undergone at least one FOBT screening. Asshown in Table 3, a logistic regression analysisfound that acculturation (Wald w2 (1)=0.3, n.s.)did not predict FOBT. The addition of riskperception (Dw2 (1)=11.24, p50:001) and physi-cian recommendation (Dw2 (1)=54.69, p50:001)incremented the fit of the model. This final model,using acculturation, risk perception, and physician

recommendation as predictors, fit the data (Hos-mer-Lemshow w2 (6)=2.54, n.s.), accounting for9% of the variability in utilization of FOBT.Exponentiation of the beta weight for physicianrecommendation indicated that it was associatedwith increased odds of obtaining an FOBT.

FSIG screening

FSIG screening was examined for men andwomen over the age of 55 who did not report ahistory of colon cancer ðn ¼ 87Þ. Based on theseinclusion criteria, 107 participants were excludedfrom the analysis. Fifty-three percent of thissample (31% men; 22% women) reported havinghad at least one FSIG screening. A logisticregression analysis (see Table 3) found thatacculturation did not predict FSIG (Wald w2

(1)=2.02, n.s.). The addition of risk perception(Dw2 (1)=10.36, p50:005) and physician recom-mendation (Dw2 (1)=42.76, p50:001) to the model

Table 2. CRC screening rates of participants in San Francisco vs Houston

Location

Screening method S F participants ðn ¼ 123Þ Houston participants ðn ¼ 83Þ p-value

FOBT n ¼ 43 n ¼ 37 0.782

(35.0%) (30.1%)

FSIG n ¼ 22 n ¼ 25 0.138

(25.3%) (28.7%)

Colonoscopy n ¼ 9 n ¼ 20 0.049

(15.3%) (33.0%)

p-values were obtained from the chi-square test using Yates’ continuity correction.

Table 3. Relationship between selected demographic, attitudinal, and CRC recommendations and colorectal cancer screening

practices

FOBT FSIG Colonoscopy

ðN ¼ 121Þ ðN ¼ 85Þ ðN ¼ 56ÞOdds ratio Odds ratio Odds ratio

Variable (95% CI) (95% CI) (95% CI)

Acculturation 1.09 0.35 0.40

(0.43–2.75) (0.10–1.28) (0.12–1.31)

Risk perception 1.80 0.68 0.90

(0.57–5.70)�� (0.13–3.51)� (0.16–5.09)�

Physician recommendation 3.71 9.10 9.52

(1.11–12.46)�� (1.87–44.21)�� (1.56–58.82)��

�Statistically significant at 0.005 level.��Statistically significant at 0.001 level.

E.J. TENG ET AL.378


resulted in incremental improvements in the modelfit, which accounted for 25% of the variability inutilization of FSIG. Goodness of fit of the fullmodel, which included acculturation, risk percep-tion, and physician recommendation as predictors,was acceptable (Hosmer-Lemshow w2 (7)=5.61,n.s.). Physician recommendation was associatedwith a substantial increase in utilization of FSIG.

Colonoscopy screening

Colonoscopy was examined for men and womenover the age of 60 who did not report a history ofcolon cancer ðn ¼ 59Þ. Based on these inclusioncriteria, 135 participants were excluded from theanalysis. Fifty-one percent of this sample (22%men; 29% women) reported having undergone atleast one colonoscopy screening. A logistic regres-sion analysis (see Table 3) found that acculturationdid not predict colonoscopy (Wald w2 (1)=1.40,n.s.). The addition of risk perception (Dw2 (1)=10.695, p50:005) and physician recommendation(Dw2 (1)=12.87, p50:001) both incremented the fitof the model. Using acculturation, risk perception,and physician recommendation as predictor vari-ables, this full model fit the data (Hosmer-Lemshow w2 (6)=7.50, n.s.), accounting for 27%of the variability in utilization of colonoscopy.Physician recommendation was associated withincreased odds of obtaining a colonoscopy.

DISCUSSION

Although regular screening can effectively reduceCRC mortality, screening rates are dismally lowfor the general population in the United States(Ioannou et al., 2003) and even lower for therapidly growing Asian–American population(Centers for Disease Control, 1999). Few research-ers have examined screening rates and factorsinfluencing screening behavior in different sub-groups of Asian–Americans despite the variabilityin cancer rates among these different subgroups.The aims of this study were to determine (1) ratesof FOBT, FSIG and colonoscopy screeningamong Chinese Americans and (2) predictors ofutilizing these screening procedures.

The current study indicated that Chinese Amer-icans report engaging in a higher rate of CRCscreening than that of previous studies with

samples restricted to younger women (Tanget al., 2001) or with combined Asian subcultures(Atkin, 2003). Of the respondents over 50 years ofage, 65% reported having had at least one FOBT,54% at least one FSIG and 49% at least onecolonoscopy. One possible reason for these rela-tively high rates of CRC screening as compared tothat of the general population may be because thecurrent sample was relatively well-educated (i.e.51% reported having at least a college education).Another explanation may be that the sample is notrepresentative of the general population, asrecruitment took place in senior centers andchurches, where community educational forumsare often made available. Thus, this sample mayhave been more informed about the importance ofpreventive health practices. Although the reportedCRC screening rates in the current study arerelatively high, a substantial number of people stilldo not obtain screening.

Some researchers studying Asian–Americanwomen have found that certain cultural factors(e.g. modesty, language and acculturation) predictwhether they obtain breast and cervical cancerscreening (Centers for Disease Control and Pre-vention, 1992; Tang et al., 1999, 2000). However,the data are equivocal regarding the influence ofcultural factors on screening behavior. For exam-ple, in examining a number of cultural factors suchas acculturation, modesty, use of Eastern medicineand degree of family support for cancer screening,Tang et al. (2001) found acculturation to be theonly significant cultural factor to predict obtainingan FOBT and FSIG at least once. Physicianrecommendation was an additional predictor ofscreening, but only of FSIG.

The current study, using a larger sample ofChinese Americans (men and women), indicatedthat physician recommendation of CRC screeningpredicted utilization of FOBT, FSIG and colono-scopy. Acculturation and perceived risk of devel-oping CRC did not predict obtaining any of thescreening methods, suggesting that these variablesare not enough to motivate Chinese Americans toavail themselves of CRC screening. Rather, thekey factor in determining whether Chinese Amer-icans obtain CRC screening is physician recom-mendation. The importance of physicianrecommendation has been recognized in a numberof other studies (Kelly and Shank, 1992; Lee et al.,1999; Lewis and Jensen, 1996), and its absencetraditionally has been one of the reasons mostcommonly given by patients (regardless of ethnicity)



for not obtaining cancer screening tests (NationalCancer Institute Breast Cancer Consortium, 1990).Research has shown that physician recommenda-tion of screening increases patient compliance(Rimer et al., 1988). Because respect for authorityis highly valued in Chinese culture, physicianencouragement of regular cancer screening mightbe especially effective with this patient population(Lee et al., 1999).

As previously mentioned, while the recruitmentmethod used in this study resulted in a higherresponse rate than other survey methodologiesthat are not conducted in person, it should benoted that the Chinese individuals sampled in thisstudy belonged to community centers andchurches, which may offer them more exposureto positive health behaviors than persons who areless involved with the community. This may be analternative explanation for the apparent increasein CRC screening among older Chinese indivi-duals. Another possibility to consider is thepresence of regional variations in cancer screeningrates among the Chinese. For example, Tang et al.(2001) found lower CRC screening rates (25% hadever obtained FOBT; 31% had ever obtainedFSIG) among Chinese American women on theEast Coast. FOBT screening rates were even lower(15%) among Chinese American men and womensurveyed in the mid-west (Yu et al., 2001). Resultsfrom the current study did not reveal differences inscreening rates between persons from San Fran-cisco compared to Houston, but this should beexplored further in future studies. Due to feasi-bility issues, cancer-screening data were notcollected directly from patient health records butrelied on self-report. It should be acknowledgedthat the data are subject to inaccuracies inherent inthis type of methodology (e.g. accurate recall andreporting, item comprehension).

Although these limitations should be consideredwhen interpreting the data, findings from thisstudy have several potentially important implica-tions. It is essential that physicians (1) be aware ofthe impact their recommendations can have and(2) recommend cancer screening more often toChinese Americans over the age of 50. Based onthe findings from this study, physicians arerecommending screening to this population onlyabout 50 percent of the time. Although more thanhalf the respondents reported having had at leastone FOBT, approximately half of this sample alsohas never obtained an FSIG or colonoscopy.Future efforts should be aimed at increasing

utilization of the latter two procedures, as endo-scopic screening has demonstrated efficacy inreducing the incidence of distal CRC (Atkin,2003). More effort is needed to enhancephysicians’ awareness of their potential impactupon screening utilization among Chinese Amer-icans. At a more systemic level, clinical remindersystems to prompt physicians to recommendregular screening also might lead to improvedscreening utilization. More efforts also are neededwithin the Chinese–American community to pro-mote healthy behaviors through education and theprovision of community resources for preventivehealthcare. Such measures could greatly reduceincidence and mortality rates of CRC amongChinese Americans by conveying the importanceof adhering to regular cancer screening guidelinesand making the recommended procedures moreaccessible.

ACKNOWLEDGEMENTS

This material is based in part upon work supported bythe Office of Academic Affiliations, VA Special MIR-ECC Fellowship Program in Advanced Psychiatry andPsychology, Department of Veterans Affairs.

REFERENCES

American Cancer Society. 2005. Cancer Prevention andEarly Detection Facts and Figures 2005. AmericanCancer Society: Atlanta.

Anderson LM, May DS. 1995. Has the use of cervical,breast, and colorectal cancer screening increased inthe United States? Am J Public Health 85: 840–842.

Atkin W. 2003. Options for screening for colorectalcancer. Scand J Gastroenterol 237: 13–16.

Burnam MA, Hough RL, Karno M et al. 1987.Acculturation and lifetime prevalence of psychiatricdisorders among Mexican Americans in Los Angeles.J Health Soc Behav 28: 89–102.

Centers for Disease Control and Prevention. 1992.Behavioral risk factor survey of Chinese: California1989. MMWR Morb Mortal Wkly Rep 41: 266–271.

Centers for Disease Control and Prevention. 1999.Screening for colorectal cancer: United States, 1997.MMWR Morb Mortal Wkly Rep 45: 116–121.

Coughlin SS, Uhler RJ. 2000. Breast and cervical cancerscreening practices among Asian and Pacific IslanderWomen in the United States, 1994–1997. CancerEpidem Biomar 9: 597–603.

Friedman LC, Woodruff A, Lane M et al. 1995. Breastcancer screening behaviors and intentions among

E.J. TENG ET AL.380


asymptomatic women 50 years of age and. Am J PrevMed 11(4): 218–223.

Hardcastle JD, Chamberlain JO, Robinson MH et al.1996. Randomized controlled trial of faecal-occultblood screening for colorectal cancer. Lancet 348:1472–1477.

Ioannou GN, Chapko MK, Dominitz JA. 2003.Predictors of colorectal cancer screening participationin the United States. Am J Gastroenterol 98(9):2082–2091.

Kahi CJ, Rex DK. 2004. Current and future trends incolorectal cancer screening. Cancer Metastasis Rev 23:137–144.

Kelly RB, Shank JC. 1992. Adherence to screeningflexible sigmoidoscopy in asymptomatic patients. MedCare 30: 1029–1042.

Kronborg O, Fenger C, Olsen J, Jorgensen OD,Sondergaard O. 1996. Randomised study of screeningfor colorectal cancer with faecal occult-blood test.Lancet 348: 1467–1471.

Lee MM, Lee F, Stewart S, McPhee S. 1999. Cancerscreening practices among primary care physiciansserving Chinese Americans in San Francisco. WJM170(3): 148–155.

Lemon S, Zapka J, Puleo E et al. 2001. Colorectalcancer screening participation: Comparisons withmammography and prostate-specific antigen screen-ing. Am J Public Health 91: 1264–1272.

Lewis SF, Jensen NM. 1996. Screening sigmoidoscopy:Factors associated with utilization. J Gen Intern Med11: 542–544.

Lieberman DA, Weiss DG. 2001. One-time screening forcolorectal cancer with combined fecal occult-bloodtesting and examination of the distal colon. New EngJ Med 345: 555–560.

Mandel JS, Bond JH, Church TR et al. 1993. Reducingmortality from colorectal cancer screening for fecaloccult blood test. New Eng J Med 328: 1365–1371.

Miller BA, Kolonel LN, Bernstein L et al (eds). 1996.Racial/Ethnic Patterns of Cancer in the United States1988–1992. National Institutes of Health PublicationNo. 96 4101, National Cancer Institute: Bethesda,MD.

Muller A, Sonnenberg A. 1995. Protection by endoscopyagainst death from colorectal cancer: An case–controlstudy among veterans. Arch Intern Med 155:1741–1748.

National Cancer Institute Breast Cancer Consortium.1990. Screening mammography: A missed clinicalopportunity? The results of the NCI Breast ScreeningConsortium and National Health Interview Studies.JAMA 264: 54–58.

Parker SL, Davis KJ, Wingo PA, Ries LA, Heath CW.1998. Cancer statistics by race and ethnicity.CA-Cancer J Clin 48(1): 31–48.

Polednak AP. 1990. Knowledge of colorectal cancer anduse of screening test in persons 40–74 years of age.Prev Med 19: 213–226.

Richardson JL, Danley K, Mondrus GT, Deapen D,Mack T. 1995. Adherence to screening examinationsfor colorectal cancer after diagnosis in a first degreerelative. Prev Med 24: 166–170.

Rimer BK, Davis SW, Engstrom PF, Myers RE, RosanJR. 1988. Some reasons for compliance and non-compliance in a health maintenance organizationbreast screening program. J Comp Health Care 3:103–114.

Ruffin MT, Gorenflo DW, Woodman B. 2000. Pre-dictors of screening for breast, cervical, colorectal,and prostatic cancer among community-based pri-mary care practices. J Am Board Fam Pract 13:1–10.

Selby JV, Friedman GD, Quesenberry CP, Weiss NS.1992. A case–control study of screening sigmoido-scopy and mortality from colorectal cancer. New EngJ Med 326: 653–657.

Sun WY, Basch CE, Wolf RL, Li XJ. 2004. Factorsassociated with colorectal cancer screening amongChinese–Americans. Prev Med 39: 323–329.

Tang TS, Solomon LJ, McCracken LM. 2001. Barriersto fecal occult blood testing and sigmoidoscopyamong Chinese–American Women. Cancer Pract9(6): 277–282.

Tang TS, Solomon LJ, McCracken LM. 2000. Culturalbarriers to mammography, clinical breast exam, andbreast self-exam among Chinese–American women 60and older. Prev Med 31: 575–583.

Tang TS, Solomon LJ, Yeh CJ, Worden JK. 1999. Therole of cultural variables in breast self-examinationand cervical cancer screening behavior in young Asianwomen living in the United States. J Behav Med 22:410–436.

US Census Bureau. 2000. The Asian Population: 2000.Census 2000 Brief, C2KBR/01-16, February 2002.

Weinrich S. 1990. Predictors of adults’ participation infecal occult blood screening. Oncol Nurs Forum 17:715–720.

Winawer SJ, Felhinger BJ, Schottenfeld D, Miller DG.1993. Screening for colorectal cancer with fecal occultblood testing and sigmoidoscopy. J Nat Cancer Inst85: 1311–1318.

Winawer SJ, Stewart ET, Zauber AG, National PolypStudy Work Group et al. 2000. A comparison ofcolonoscopy and double-contrast barium enema forsurveillance after polypectomy. N Engl J Med 342:1766–1772.

Yu ES, Kim KK, Chen EH, Brintnall RA, Liu WT.2001. Colorectal cancer screening among ChineseAmericans: A community-based study of knowledgeand practice. J Psychosoc Oncol 19 (3/4): 97–112.



Documents

Determinants of colorectal cancer screening behavior among Chinese Americans