Developing a discrete choice experiment in Malawi: eliciting preferences for breast cancer early detection services

Figures & data

Table 1 Potential attributes and levels for cancer services compiled from the literature

Conceptual domain	Attribute
Psychosocial^Citation25,Citation27,Citation28	Risk of dying from breast cancer
	Physical discomfort
	Embarrassment
Invitation^Citation25,Citation26,Citation31	Invitation/recruitment method
	How much information is shared
Convenience^Citation25,Citation26,Citation29,Citation30	Scheduling delay
	Availability of services, open hours
	Distance to facility
	Transportation costs
	Travel time
	Waiting time
Facility infrastructure^Citation29,Citation30,Citation33^–^Citation36	Setting (health facility, community, pharmacy, mobile unit)
	Availability of drugs and medical equipment
	Cleanliness of facility
Privacy^Citation26,Citation32	Demonstration, instruction on own breast
	Size of education class
	Changing area (private or open)
Provider^Citation26,Citation33^–^Citation36	Provider type
	Technical skills of staff
	Sex
	Attitude, personal manner of staff
Detection strategy	Mammography
	Digital mammography
	MRI and nuclear evaluation
	Breast self-exam education
Accuracy^Citation25^–^Citation27	Accuracy of detecting cancer
	How good the test is at saying you do not have cancer
	Risk of follow-up
Frequency^Citation26,Citation27	Number of tests performed over next 25 years
Results notification^Citation31	Time to results
	Mode of results
Costs^Citation25,Citation32,Citation34,Citation35	Cost of test

Note: Citations provided for previous applications in choice experiments or preference studies.

Table 2 Sample characteristics of health care worker and community interview participants

	Number	Percentage
Health care workers	10
Sex
Male	3	30
Female	7	70
Profession
Nurse	8	80
Cliniciana	2	20
Total mean years of experience	12
Mean years at current facility	3
Community women	20
Religion
Christian	18	90
Muslim	2	10
Marital status
Married	11	55
Widowed	3	15
Single/divorced	6	30
Education
No formal schooling	6	30
Attend some primary school	9	45
Attended some secondary school	5	25
Economic status
Source of regular income	7	35
Family owned house (compared to renting)	13	65
Electricity	5	25
Iron sheet roofing (compared to grass thatched)	11	55

Note:

a Includes clinical officers and resident physicians.

Table 3 Potential attributes and levels emerging from the interviews

Potential attribute	Lay terminology	Illustrative quote from health care worker and community interviews	Level included in DCE
Health information source, invitation to participate	Radio Posters Health surveillance assistant (HSA) Community health meeting Health talks in waiting area	“And many people are not literate, so we can’t assume that they will be reading…but many have radios. So they can listen from the radio”. “Most of the time I hear about health issues on the radio”. “Because there are some posters that you can put at a health facility. Some posters that you can put even in the streets…There are some posters that just carry the message without the pictures and there are some posters that would carry sensitive pictures. If there is a poster of sensitive pictures, you put it on a place within the hospital setup”. “In the communities we have health surveillance assistants […] they are given the right information to disseminate […] the basic information we want them to give to the community”. “For home based care […] those meetings they talk about health issues especially on HIV”. “Sometimes, we also learn at church, doctors came to teach us”. “When they are giving the health talk they explain that maybe because the disease starts differently [in] others…they just have a lump”.	Health talk in waiting area
Type of facility	Private Government	“Because at private when you go you pay. When you receive treatment you get cured. But also when you go to government hospital you get cured, and it’s free treatment. So the best one is government hospital”. “When I go to a government hospital I receive proper treatment. When you go to a private hospital, they just tell you to go after they have given you medication. They just want money”.
Health infrastructure	Medication availability Proper resources Staffing shortages	“Sometimes, when you go to the hospital they tell you that they have no medication. So, you then have to buy the medication. You buy it at the pharmacy”. “If the resources are not there: the theatre may be out of gauze, the drips are not well sterilized or they are not sterilized at all because the sterilizer, is not functioning. They go back [home]”. “It depends if there is staff, then that’s fine. But you know with this crisis of human resource, our setup, it also becomes a challenge”.
Perceived quality	How they greet me, provider interactions Queues and wait time	“Like, the attitudes of the medical personnel. Because if patients come to our hospitals, the way we handle them matters. And then these patients go to the village and then they tell the community there. You know, I visited such a hospital. And then the way they handled me whether it was bad or good. Then that you know, will make these other friends of theirs to either visit this hospital or not to visit this hospital”. “Because when I go to a government hospital, it is difficult for me to get treatment. For them to treat us, it takes long”. “So all those patients who are referred from private clinics, they come here and discover how long the waiting list is, and then kind of get demotivated about it […] when they get into the hospital most times they find a very long queue”.
Distance to facility	How long it takes to get there	“Aah, on minibus, it doesn’t take very long. Because a minibus is a minibus. Really. But from my place, to where I take the minibus (where she gets on the minibus) it is a little bit far. It takes me about 45 minutes from my place to where I take the minibus”. “It’s one and a half hours, it’s not very far, it’s not far because even a child walk to get there”. “Eee…the district one, to travel by foot, you can start at as early as 5 o’clock if you walk very quickly, at around 8 you have reached at the hospital”.	<1 hour 1–2 hours 2+ hours
Transportation	Transport	“Most of the times when am going to the hospital, I walk because of little [financial] assistance. Maybe if am lucky that means a bicycle. If it can be possible that I can be taken on a bicycle (hiring someone to let her ride along)”. “We also walk to [health clinic] but because the distance is long, we reach there late. When we have money for transport, we take a minibus”.
Services provided	Common things Specialty care Combined services	“I like the big government hospital – it’s where I like to go. To me it’s where I see that I find assistance. And all the specialists are there”. “Bigger hospitals are important, but it depends on the sickness. If my sickness…is serious, I can go to receive treatment at the bigger hospitals…they give us better treatment”. “I think it’s the same as we say, “Come for baby checkup”…Then I think it would be better if we included the breast [exam]. We should not do only one. We should combine”.
Health encounter; point of entry	Under five (child health checkups) Family planning Antenatal care Cervical cancer screening	“We go lots for our children to be weighed because they have just started…it’s daily at [health clinic] so we don’t like to delay we will like our child to continue and receive injections”. “They are told at an earlier stage to be examining their breasts. Because in under five or family planning they give talks – health talks about self-breast examination”. “If I had a breast problem that means maybe I would go to family planning”. “What I know through learning some other times when we go for antenatal visits when you are pregnant, they were just telling us most of the time when you wake up, maybe you should be laying on your back and be rubbing your breasts, going round. If you feel a lump or pain you should rush to the hospital”. “To do two screenings, for the breast and for the VIA. Then, it will be much easier for the patient, I mean for the client, because she comes for VIA but at the same time her breasts are examined for breast cancer. Rather than her going home and coming back again another day for this other examination”.	Under 5 child checkup Family planning Cervical cancer screening
Type of provider	Health surveillance assistants (HSA) Medical assistants Doctors Nurses	“Sometimes she (HSA) reminds me: You should not forget to go to the hospital. I feel very free with this woman”. “The health centers are managed mostly by a medical assistant and then, so they have a lot of work to do”. “So sometimes the medical assistants don’t know much about the epidemiology of some of these conditions. They are able to pick out that this is a breast lump or what but then sometimes they don’t just see they just don’t just see the urgency in certain situations”. “Because we meet with doctors who are experts in everything”. “There are other nurses who give proper treatment”.	HSA Doctor
Sex of provider	Female Male	“I would choose a woman because we are of the same nature”. “I can choose a male health worker […] For example…my first child, I was assisted by a male health worker to deliver it. I did not see any difference. The only difference was that the female health worker was just yelling. So the male health worker just said that he would assist me…he did not yell at me at all”.	Female Male
Privacy	Feeling shy	“It’s not a shameful thing. I don’t feel shy to take off my clothes for a doctor. Because I say that they should see me”. “You have to hide them with a screen, so they are not seen by all the other people. There are no rooms; it is just open. So sometimes there are screens in between the beds, so we put another screen just to hide the front of it”.
Detection method	Get checked (CBE) Teach/learn how to check (BSE instruction) combination	“It would catch as many as possible so we would miss less of these breast lumps and breast problems”. “You can improve on health education. That means you can give talks starting from the community up to a health care facility. In the communities […] we can teach women in the villages to do self-breast examination”. “Yeah since we are in a setting where people are on queues waiting for so long sitting there so bored someone could just. Someone like a nurse could just give them a health talk when they are in the sitting area…they will not be resisting when we suggest to them to do the breast exam at opportunity time like those, for example, if they come for the family planning clinic”.	Clinical breast exam (CBE) Breast health awareness education and BSE Both CBE and breast health awareness with BSE

Abbreviations: DCE, discrete choice experiment; VIA, visual inspection with acetic acid; HSA, health surveillance assistant; BSE, breast self-exam.

Figure 1 Example choice scenario in discrete choice experiment.

LiangWLawrenceWFBurnettCBAcceptability of diagnostic tests for breast cancerBreast Cancer Res Treat200379219920612825854

 PubMed Web of Science ®Google Scholar

Gyrd-HansenDCost-benefit analysis of mammography screening in Denmark based on discrete ranking dataInt J Technol Assess Health Care2000160381182111028136

 PubMed Web of Science ®Google Scholar

Gyrd-HansenDSøgaardJAnalysing public preferences for cancer screening programmesHealth Econ200110761763411747045

 PubMed Web of Science ®Google Scholar

GerardKShanahanMLouviereJUsing stated preference discrete choice modelling to inform health care decision-making: a pilot study of breast screening participationAppl Econ200335910731085

 Web of Science ®Google Scholar

MarcusENDrummondDDietzNUrban women’s preferences for learning of their mammogram result: a qualitative studyJ Cancer Educ201227115616422072125

 PubMed Web of Science ®Google Scholar

LinsellLForbesLJPatnickJWardleJAustokerJRamirezAJWomen’s preferences for the delivery of the National Health Service Breast Screening Programme: a cross-sectional surveyJ Med Screen201017417618021258127

 PubMed Web of Science ®Google Scholar

GuptaSVijACafieroERetail venue based screening mammography: assessment of women’s preferencesAcad Radiol201219101268127222958721

 PubMed Web of Science ®Google Scholar

HansonKMcPakeBNakambaPArchardLPreferences for hospital quality in Zambia: results from a discrete choice experimentHealth Econ200514768770115619273

 PubMed Web of Science ®Google Scholar

ChristofidesNJMuirheadDJewkesRKPenn-KekanaLConcoDWomen’s experiences of and preferences for services after rape in South Africa: interview studyBMJ2006332753520921316330476

 PubMed Web of Science ®Google Scholar

BechMSørensenJLauridsenJEliciting women’s preferences for a training program in breast self-examination: a conjoint ranking experimentValue Health20058447948716091025

 PubMed Web of Science ®Google Scholar

KrukMEPaczkowskiMMbarukuGde PinhoHGaleaSWomen’s preferences for place of delivery in rural Tanzania: a population-based discrete choice experimentAm J Public Health20099991666167219608959

 PubMed Web of Science ®Google Scholar

KrukMEPaczkowskiMMTegegnAWomen’s preferences for obstetric care in rural ethiopia: a population-based discrete choice experiment in a region with low rates of facility deliveryJ Epidemiol Community Health2010641198498819822558

 PubMed Web of Science ®Google Scholar