Navigating cancer care in a country not your own: the immigrant experience. 

Phyllis ButowUniversity of Sydney

Australia is multicultural

24% of Australians were born overseas

44% had at least one parent born overseas

Over 560,000 people (2.8% of the total population) speak English poorly or not at all

Australian census

Ethnic disparities in outcomes

Minorities diagnosed with cancer have poorer outcomes than majority groups: lower screening rates lower survival rates poorer quality of life higher anxiety and depression

Du XL, et al. Cancer 109(11):2161-70, 2007

Chu KC, et al. J Natl Med Assoc 99(10):1092-100, 2007

Krupski TL. Et al. Ethnicity & Disease 15(3):461-8, 2005

Gotay CC. et al. Psycho-Oncology 11(2):103-13, 2002

Butow et al, European J of Cancer, In Press.

Psychological wellbeing in minority

versus mainstream cancer patients

Ethnic group

Asian/API

Std. Mean Difference

IV, Random, 95% CI

-1 -0.5 0 0.5 1

Favors mainstream Favors minority

US Hispanic

All combined

US minorities

Outcome

Distress

Depression

Anxiety

Asian/API

Asian/API

US Hispanic

US Hispanic

Romanian Hung.

All combined

All combined

Luckett T et al. Lancet Oncology

2011; 12(13): 1240-1248.

How is the system

failing minorities???

Disparities – MechanismsBarriers to Access and Quality of Care

PreventionPreventionEarly DetectionEarly Detection Diagnosis/

Incidence

Diagnosis/

IncidenceTreatmentTreatment

Post-treatment QoLPost-treatment QoL Survival and

Mortality

Survival and

Mortality

Economic

Social Cultural

Ward et al. (2004): adapted from Freeman (1989) and Institute of Medicine (2003)

Language proficiency,

Health literacy

Beliefs and attitudes

(eg decision-making)

Income

Education

Insurance

Poor health High

stress

Low health literacy disproportionately affects racial and ethnic minorities

US Dept of Health and Human Services 2010

Ngui et al, J Health Care 2010;18:931-49

Explored unmet needs in 1040 cancer survivors recruited through cancer registries

Non-White ethnicity only factor which predicted higher information needs

Lindau et al, Amer J Obstets and Gynae 2002; 186: 938

529 English-speaking patients in women's clinics Minority women half as likely to know purpose of Pap

test (9% vs 21%; P <.03) & significantly more likely to have low literacy levels compared with white women (46% vs 15%; P <.05)

Literacy the only factor independently associated with knowledge related to cervical cancer screening (adjusted odds ratio, 2.25; 95% CI, 1.05-4.80)

Physicians detected only 20% of the lowest readers

Program of research on CALD

Phyllis Butow David Goldstein Maurice Eisenbruch Michael Jefford Penelope Schofield Afaf Girgis Madeleine King Rick Iedema

Lynley Aldridge (Co-ordinator)

Ming Sze (PhD student)

Skye Dong

Sara Fagir

Sarah Abdo

Suzanne Loway-Aziz

Evi Politi

Takis Katsampanis

Kathy Lee

Icie Wan

Investigators Research staff

Statistician

Melanie Bell

Dr Rina Hui Dr Amanda Goldrick Dr Ray Ashgari Dr Jodi Lynch Dr Janette Vardy Dr. Winston Liauw Dr. Craig Lewis Dr Fran Boyle Dr Nicholas Wilken Dr Liz Hovey

Collaborating oncologists

Funded by NHMRC

Cancer Australia

Beyond Blue

Cure Cancer Australia

SESAHS Multicultural Unit

Community partners Greek: Oncologist, GP, Psychologist, Cancer

Council Regional program coordinator, Priest, Consumer Representative, Member of Helenic Centre for Language and culture

Arabic: GP, Social worker, Cancer Council Regional program coordinator, Consumer Representative, Christian Priest, Muslim Religious leader

Chinese: Oncologist (palliative care), Retired GP, Psychologist, Christian Religious leader, Buddhist Religious leader, Consumer Representative

Study designs Population-based cross-sectional study through

state cancer registries survivors diagnosed in past 6 years

Clinic-based cross-sectional study through hospitals patients in active treatment, diagnosed in past 6-12 months

Focus groups and interviews patients, carers and interpreters

Audio-taping of first two consultations after diagnosis with poor prognosis followed by interview with patient and family

SURVIVAL SAMPLE (n=596)

Demographics% Immigrant

N=277% AngloN=319

Cultural background: English-speaking Arabic-speaking Greek-speaking Chinese-speaking

91324

54

Gender: Male 54 49

Marital status: Married or partnered 80 74

Place of residence: # Rural (versus urban) 1 12

Education: # <High school High school/ tech college University

214237

57123

p<.0001

p<.0001

DemographicsM (SD)

ImmigrantM (SD)Anglo

Years in Australia: 30 (14.6 years)

Literacy % %

Understanding of the Australian Health System: Not well: 27% 11%

P<0.0001

Confidence speaking English (immigrant only) : Not confident 30%

Difficulty communicating with doctor Often 24%

SAMPLE (n=596)

Top 10 unmet needs of immigrants Needs % of Immigrants

endorsing

• Managing FCR 30

• Written information in own language 27

• Information about cancer and its treatment 27

• Unable to do usual things 24

• Not sleeping well 24

• Medical Guidance 24

• Specialist who speaks my language 24

• Information about CAM 24

• Help asking questions 23

• Other health professionals who speak my language 22

Top 10 unmet needs of Anglos Needs % of Anglos

endorsing

• Unable to do usual things 19

• Managing FCR 17

• Changes in sexual relationships 17

• Lack of energy 16

• Not sleeping well 15

• Changes in sexual feelings 15

• Information about sexual relationships 14

• Support services 11

• Unable to work 11

• Moving on in life 10

Information

Only 21% of immigrant patients reported receiving information about their cancer and its treatment in their own language

Factors associated with severity of unmet needs (Immigrants only)

Immigrants who did not understand the health system had 2 times higher unmet needs (p<0.0003)

Participants who needed an interpreter had 2 to 4 times higher unmet needs (p<0.0001)

NB, getting an interpreter did not modify this!

Focus groups and interviews; The patient experience

22 Greek, 49 Chinese, 20 Arabic patients

Any type / stage of cancer diagnosed in last three years Recruited from support groups, multicultural services

and hospitals 68% poor English

Data collection: Focus groups and interviews audio-taped, transcribed

and translated into English

Results

Three sources of distress: Cancer

A death sentence Stigma

Family issues Protecting the family

Support and Communication

Distance and Isolation …

Communication barriers

“You have no way to communicate you know … You can’t hear, you can’t speak. It’s very frightening.”

– Mandarin speaker

Communication barriersInformation inaccessible

“They gave me some books where I did the chemotherapy, but I couldn’t read them, they were in English.” Greek patient

“He speaks, but what we hear is only 20%, understand 20%.”

“I just sit there and say yes, yes, but I left his consultation and I didn’t understand most things.” Arabic patient

“You know it is like the chicken talking to the duck. It’s different voices.” Cantonese patient

Information needs “Those doctors are very

obliged by the law, … (to provide adequate information) but when they are dealing with Chinese patients they don’t do it.”

Mandarin patient

“I don’t know if there is racism in this country. If it is someone who doesn’t speak English then they won’t give you a straight answer.”

Arabic patient

Communication barriers: Impact on physical and psych health

His surgeon is a Caucasian, so is his dentist…he (head & neck patient) couldn’t understand what they said…so he skipped the dentist’s appointment. He didn’t see a dentist at all. Cantonese patient

Sometimes we want to express things but we can’t. So it’s very difficult (psychologically) to relate to them (hospital staff). Mandarin patient

Interpreter challengesWorry about accuracy (both family and professionals) “I talk … for a few sentences and he interprets those few

sentences. But we don’t know whether the interpretation is correct….” Cantonese Patient

Speak the right language! “I saw once a person who speaks Cantonese. They

found him a Vietnamese interpreter. They didn’t realize we have Cantonese and Mandarin amongst Chinese”. Cantonese patient

Audiotapes of consultations Arabic, Greek and Chinese patients, and Anglo

controls recruited through Oncology clinics in 9 Sydney hospitals

First 2 consultations after diagnosis of metastatic disease audio-taped: Interpreter usually present Professional or family

Bi-lingual RAs interviewed patient and family after second consultation

144 consultations audiotaped

Consultation analysis

Audiotapes transcribed

All speech in Arabic, Chinese or Greek, translated into English Of patient, carer or interpreter

Transcript divided into units of speech

Two bi-lingual RAs coded transcripts: inter-rater reliability 85%

Coding 1: Overall picture Each unit coded * for:

Source who is speaking and to whom doctor, patient, family, nurse, interpreter

Content diagnosis, prognosis, socio-emotional etc

Function ask question, inform, reassure etc

* CanCode (Dent et al Patient Education and Counseling. 2005;56(1):35-44.

Identical / Not Identical interpretation If not identical, what has changed?

Medical terms omitted Missing information Added information

Outcomes of alteration Softening More authoritative and directive Clarifying Misinformation Emotional tone changed

Coding 2: Interpreter exchanges

Length of consultations

Mean number of words

(excluding translations)

Anglo-Australians 2246

Immigrants (with interpreter) 1443 P = 0.005

Immigrants (without interpreter) 2093

What does the doctor discuss?

• Doctors spoke proportionally less with immigrants with interpreters

than with Anglo-Australians

• about cancer related issues (p = 0.005),

• summarising and informing (p ≤ 0.003)

• Doctors spent proportionally more time

• on other medical issues (p = 0.0008)

• directly advising (p = 0.0008)

• More paternalistic style? Or responding to consumer preferences?

Results

On average, interpretations were equivalent 60% of the time

95% CI: 52, 68%

Professionals : 65% Family / friends 50%

CI for the difference = 3, 28%, p = .02

Impact

68% of non-equivalent interpretations were judged to have had no or a positive effect

32% were judged to have had a potentially negative effect

Impact? Potentially negative

Outcome Mean proportion

Range

Increased certainty 3% 2-5%

Dr: We think there is a 40% chance that the treatment will prolong your life

 

Int: The treatment will prolong your life

Family interpreter

Impact? Potentially negative

Outcome Mean proportion

Range

Reduced certainty 1% 0.7-2.2%

Dr: You must take the tablets every day for them to work

 

Int: You should take the tablets every day

Professional interpreter

Impact? Potentially negative

Outcome Mean proportion

Range

Euphemistic 1.4% 0.01-2.8%

Dr: Your cancer has grown despite the chemotherapy we have given you.

 

Intr: The growth in your lungs has got a little bigger, although you have had

chemotherapy

Professional interpreter

Impact? Potentially negative

Outcome Mean proportion

Range

Misinformation 10% 7-12%

Dr: From the information I got from you regarding your tumour in China I don’t

think that you had that particular receptor looked at.

 

Int: From the information you gave me about treatment in China, I felt you hadn’t

received this treatment

 

Professional Interpreter

Impact? Potentially negative

Outcome Mean proportion

Range

Misinformation 10% 7-12%

Dr: I think it is better that she knows. Especially when she is asking the question, for

her to know and to be very honest with her. She then knows that she can trust me to

be honest and I will answer her questions.

 

Int: Hmm, yeh, yeh

Dr: [Went on explaining the extent of the disease - 3A lung cancer]

 

Int: She said...it should be in its initial stage.

Family interpreter

Prognosis

Non-interpreted speech often about prognosis

50% of doctor and 59% of patient prognostic speech units not interpreted or interpreted non-equivalently

Conclusions?

CALD patients report communication and navigation challenges, and a high unmet need for information

Doctors explain less to immigrants than Anglo-Australians

Mis-interpretation common

Interpreters

We cannot under-estimate the impact of the interpreter. Further research is need to understand their: Motivations Assumptions Dilemmas Role boundaries Training needs

Interpreters

Presence of professional interpreter is critical

Ideally interpreter and Dr meet before consultation to:

brief interpreter ensure medical terms and concepts

understood discuss cultural issues and optimal ways to

convey messages

Cultural advocates / navigators may be necessary

Bilingual nurses who contact patient at diagnosis and critical transitions?

Cost-effective strategy may be centralised phone contact

CALD CONNECT 5 phone calls at key times in illness trajectory, incorporating:

needs and understanding assessment care co-ordination link to written materials in patient’s language referral and liaison back to treatment team as needed

Patients and families contact at additional times if needed Focus groups supported this model

Shaw J et al, Supportive Care in Cancer. 2013 epub

Development of CALD care-co-ordination questionnaire underway (Young et al: CINSW funding)

Systematic review of colorectal cancer interventions for CALD 32 included studies Patient education (phone or in-person) +

navigation increased colorectal cancer screening rates by about 15 percentage points, in minority populations.

Provider-directed multi-modal interventions (education sessions and reminders, + pure education) increased colorectal cancer screening rates, also by about 10 to 15 percentage points

Naylor K et al, J Gen Intern Med 2012; 27: 1033-46

Interventions

Translated materials

Access resources from countries of origin?

Survey of resources and adaptation of existing resources for survivors underway:

Jefford et al, Cancer Australia funding

Interventions Audiotaped Oncology consultations

To allow review at home, and with family

Question prompt lists in patient’s language To foster patient empowerment

Information booklets in patient’s language

Phase II study underway Schofield et al (NHMRC funded)

Interventions Cross-cultural training for health professionals:

Awareness of vulnerability in some CALD patients: Refugee status, lack of acculturation, low SES, isolation, cultural

differences in views of illness and treatment

Principles and schemas to assist HPs in interacting with CALD patients recognising one’s own cultural bias separating clinicians’ values and beliefs from their patients’ accessing and acquiring culturally relevant knowledge learning how to elicit patients’ illness explanatory models how to use relevant resources (e.g. interpreters and translated

resources) effectively

Cross-cultural training

Phase II feasibility study underway online training only versus online

training plus face-to-face workshop

Meiser et al, ARC funded.

Summary

We need to better understand health literacy issues in CALD patients

Interventions needed!