A Discrete Choice Experiment to Examine the Preferences of Patients With Cancer and Their Willingness to Pay for Different Types of Health Care Appointments

Background: This study sought to understand the preferences of patients with cancer and the trade-offs between appointment attributes using discrete choice experiment (DCE). Methods and Study Design: Patients with cancer at 3 hospitals completed a self-administered DCE. Each scenario described 6 attributes: expertise of health care professionals (HCPs), familiarity of doctors with patients' medical history, waiting time, accompaniment by family/friends, travel time, and out-of-pocket costs. Patient preferences were estimated using logistic regression. Willingness to pay (WTP) estimates were derived from regression coefficients. Results: Of 512 patients contacted, 185 returned the questionnaire. The mean age was 61 years, and 60% of respondents were female. The mean time since cancer diagnosis was 34 months, 90% had received treatment; and 61% had early-stage disease. The most important attributes were expertise and familiarity of doctors with patients' medical history; distance traveled was least likely to influence patient preferences. The WTP analysis estimated that patients were willing to pay $680 (95% CI, 470–891) for an appointment with a specialist, $571 (95% CI, 388–754) for doctors familiar with their history, $422 (95% CI, 262–582) for shorter waiting times, $399 (95% CI, 249–549) to be accompanied by family/friends, and $301 (95% CI, 162–441) for shorter traveling times. Male patients had a stronger preference for accompaniment by family/friends. The expertise of HCP was the most important attribute for patients regardless of geographic remoteness. Conclusions: Our study can assist the development of patient-centered health care models that improve patient access to experienced HCPs, support the role of primary care providers during the cancer journey, and educate patients about the roles of non-oncology HCPs to cope with increasing demand for cancer care.

Abstract

Background: This study sought to understand the preferences of patients with cancer and the trade-offs between appointment attributes using discrete choice experiment (DCE). Methods and Study Design: Patients with cancer at 3 hospitals completed a self-administered DCE. Each scenario described 6 attributes: expertise of health care professionals (HCPs), familiarity of doctors with patients' medical history, waiting time, accompaniment by family/friends, travel time, and out-of-pocket costs. Patient preferences were estimated using logistic regression. Willingness to pay (WTP) estimates were derived from regression coefficients. Results: Of 512 patients contacted, 185 returned the questionnaire. The mean age was 61 years, and 60% of respondents were female. The mean time since cancer diagnosis was 34 months, 90% had received treatment; and 61% had early-stage disease. The most important attributes were expertise and familiarity of doctors with patients' medical history; distance traveled was least likely to influence patient preferences. The WTP analysis estimated that patients were willing to pay $680 (95% CI, 470–891) for an appointment with a specialist, $571 (95% CI, 388–754) for doctors familiar with their history, $422 (95% CI, 262–582) for shorter waiting times, $399 (95% CI, 249–549) to be accompanied by family/friends, and $301 (95% CI, 162–441) for shorter traveling times. Male patients had a stronger preference for accompaniment by family/friends. The expertise of HCP was the most important attribute for patients regardless of geographic remoteness. Conclusions: Our study can assist the development of patient-centered health care models that improve patient access to experienced HCPs, support the role of primary care providers during the cancer journey, and educate patients about the roles of non-oncology HCPs to cope with increasing demand for cancer care.

Background

Health care utilization is influenced by the predisposition of patients and their ability and need to use health care services.1 Andersen's Behavioral Model of Health Services Use1 takes into account factors that could impact health care utilization inclusive of predisposing (age, gender), enabling (financial income, insurance), and need factors (perceived health status). Farmer et al2 further suggests that patients' relationship with their primary care providers and ease of access to attend appointments were particularly important to patients who lived in more remote regions.

Ultimately, patients' health-seeking behaviors are partially determined by what they value and are measured by their willingness to accept available options after considering the potential trade-offs between these options. Patients' preferences also vary depending on when they participate in research relative to their care pathway; as noted by Brown et al,3 43% of patients (N=683) with early-stage breast cancer changed their preferences after a medical consultation.

Choice-based techniques such as discrete choice experiments (DCE)4 have been successfully used to evaluate genetic counseling for cancer,5 screening programs,6 and cancer treatments.7 Few DCE studies have been performed on how patients value different aspects of health care appointments, how these preferences may differ, and how this might affect health care utilization.8,9

We hypothesized that different patient- and health care–related characteristics will influence patient preferences about attending health care appointments to varying degrees.10 Prior assumptions were made about the behavior of included characteristics and were described in detail in an earlier publication.10 For example, we hypothesize that in an ideal world, patients will prefer to travel shorter versus longer distances to health care appointments. We also included key variables described in a systematic analysis of studies using Andersen's Behavioral Model.11 These variables included age, marital status, gender, education, employment status, financial situation, ethnicity, health insurance, and perceived health status (an example of the DCE questionnaire is provided in supplemental eAppendix 1, available online with this article at JNCCN.org).

A DCE was chosen as the preference-elicitation method in our study to

  • Quantify and establish the relative importance of different health care appointment–related attributes,

  • Identify similarities and differences in preferences among patient subgroups, and

  • Determine willingness to pay (WTP) estimates for health care appointment–related attributes.

Methods and Study Design

A full description of the methods has been described previously.10 The most salient methodological aspects of the study are summarized in the following sections.

Participants and Setting

Our study sample consisted of English-speaking patients who were diagnosed with cancer from January 2009 and attended any 1 of 3 adult oncology services across southwestern Victoria, Australia. Eligible patients who attended an appointment at the oncology services during the period July to November 2014 were given a questionnaire by the administrative or medical staff members. Patients were able to provide informed consent; there were no exclusion criteria.

A letter introducing the aim of the study, the DCE questionnaire, and a reply paid envelope were distributed to 512 eligible participants. Additional information about sociodemographic, disease-related, and treatment-related characteristics were collected. The questionnaire was designed to be self-completed and participants did not receive assistance from research staff.

The Barwon Health Human Research Ethics Committee approved the study. A returned questionnaire was assumed to represent implied consent.

The DCE Questionnaire

The DCE enabled the comparison of multiple appointment-related features in hypothetical scenarios to simulate decision-making in reality; choice responses were then analyzed to identify patients' preferences for different types of health care appointments. Respondents were asked to choose from several options, wherein each option was defined by its appointment attributes. Each attribute takes one of a number of possible levels. The relative importance of each attribute-level combination was estimated to reflect patient preferences and trade-offs.

Attributes and Levels: Attributes and levels were derived from the published literature and an analysis of qualitative data collected in interviews (N=21) from both patients with cancer and health care professionals (HCPs) caring for them.10 This approach generated a comprehensive range of attributes that are potentially of importance to patients when they attend health care appointments (Table 1).

Experimental Design of DCE:The combination of 6 attributes with 3 levels yielded 729 (=36) possible appointment profiles. A smaller fractional factorial design (FFD) was selected according to the D-efficiency criteria.12 The final FFD design randomly blocked 128 choice sets into 16 blocks with 8 choice tasks (supplemental eAppendix 1).

Table 1.

Attributes and Levels Used in the Discrete Choice Experiment

Table 1.

For each choice task, respondents were asked to choose between Appointment A and Appointment B (Figure 1). There was no opt-out option, because most patients in this situation were likely to attend health care appointments due to the factors included in the design. We expected that all of the included attributes were important to the respondents,10 rather than to assume, a priori, that parameters were zero.1315

Eleven patients who had participated in the qualitative interviews were invited to pilot test the DCE; one did not return the questionnaire. Respondents were provided with descriptions of the attributes and levels, and were given the opportunity to comment on the DCE design and layout. Five respondents found the DCE questionnaire complex and were unable to complete the questions. Subsequently, changes were made to include an example choice task that demonstrated how to choose between the 2 appointment options.

Statistical Analysis

The analysis was undertaken using a range of logistic regression models, which are commonly used to explore situations in which the dependent variable (ie, appointment choice) is binary. A conditional logit function (CL)16 is widely used in choice modeling, but it assumes all responses come from a common decision-maker, thus limiting consideration of response heterogeneity. The mixed logistic regression (ML) model17 overcomes some of these limitations, because it (1) allows regression coefficients to be drawn from a distribution and vary across respondents, implying that different respondents may have different preferences, and (2) accounts for different scenarios that the same respondent may face.

These regression models are based on the random utility model (RUM) framework, in which each respondent n is assumed to be rational and maximize utility (U),18 wherein they must choose an

Figure 1.
Figure 1.

An example of a choice task used in the discrete choice experiment.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 14, 3; 10.6004/jnccn.2016.0036

alternative j (observed) over a number of s choice scenarios that leads to a “superior” level of utility. The RUM framework consists of a dichotomous outcome variable that is dependent (Appointment A vs Appointment B) and of independent predictor variables (observed or unobserved).17 The observed variables include the attributes of the alternatives j or the observed characteristics of the respondents, such as sociodemographic data. The unobserved variables are treated as random components (ε) within the RUM.

The respondents are assumed to choose the appointment option that maximizes utility; hence we estimated the following model:

Uappointment= ß0 + ß1Expertise1 + ß2Expertise2 + ß3Drknows1 + ß4Drknows2 + ß5WaitTimes1 + ß6WaitTimes2 + ß7Support1 + ß8Support2 + ß9TravelTimes1 + ß10TravelTimes2 + ß11Cost_Cont + ε where Uappointment represents the utility of a hypothetical appointment scenario, ß0 is a constant that indicates the general respondents' preferences for Appointment A versus Appointment B, and ß1 through ß10 are the coefficients that reflect the relative importance of each attribute level compared with its reference level. All variables other than cost were treated as binary variables, whereas the cost parameter was modeled as a continuous variable (Cost_Cont) to allow estimation of WTP estimates around each attribute-level combination.

A positive (negative) ß coefficient indicates whether the attribute-level combination was preferred (not preferred) relative to the reference level. The higher the absolute value of the ß coefficient, the stronger the preference for that level relative to the reference level.

Subgroup analysis was undertaken using CL to explore differences in preferences among sociodemographic, disease-related, and treatment-related characteristics. Where possible, the subgroups were prespecified to ensure adequate sample size for estimation, thus enabling greater confidence in the comparison of results among subgroups (Table 2).

Results from the regression analyses are presented as parameter estimates, with 95% CIs and P values.

Table 2.

Subgroups and Their Descriptions

Table 2.
A 2-sided P value less than 0.05 was considered to be statistically significant. The goodness of fit for each model was estimated using the likelihood ratio chisquare statistic and McFadden's pseudo R2. All analyses were conducted using STATA Statistical Software: Release 13.1 (STATA-13; StataCorp LP, College Station, TX).

Results

Respondent Characteristics

A total of 512 questionnaires were distributed to patients across 3 hospital sites; the participation rate was 36% (N=185). There were 128 completed questionnaires and 57 partially completed questionnaires. The total number of responses resulted in 2,960 discrete-choice observations (sample sizes and sampling probabilities are provided in supplemental eAppendix 2). Respondents took an average of 17 minutes to complete the questionnaire (SD, 10; 2–90).

Patient characteristics are summarized in Table 3. Mean age for all respondents was 61 years (SD, 14; 22–92) with mean time since diagnoses of 34 months (SD, 46; 1–300).

Respondents' Preferences

Both CL (chi-square, 396; 11 df; P< .001; pseudo R2=0.249) and ML (chi-square=76; 11 df; P<.001) models indicated that all the attributes of the appointments were highly statistically significant relative to the reference level (Table 4). These attributes performed in line with prior assumptions,10 which confirms the theoretical validity of the study.

The most important attributes for determining patient preferences were expertise of the HCP (cancer specialist vs nurse/general practitioner [GP]: ß=1.98; P<.001) and familiarity of the doctors with their medical history (usual GP vs new practice: ß=1.61; P<.001); travel time (30 minutes vs 2–3 hours: ß=0.96; P<.001) was least likely to influence patient preferences.

Additional results from the ML model relate to the SD of the independent parameters. The last column of Table 4 shows that the derived SD for waiting time (3 weeks), unavailability of family/friends to stay overnight if required, and travel times were not statistically significant; these indicated that the distributional information for these parameters was contained within the mean with minimal heterogeneity. The statistically significant SD for expertise of HCP, familiarity of doctors with medical history, waiting time (6 weeks), and nonaccompaniment by family/friends suggests there is considerable heterogeneity around the mean for these parameters.

WTP Estimates Between Attributes: WTP estimates were derived from the CL model by comparing the coefficients for each attribute-level combination with the coefficient of cost. Figure 2A shows that, as expected, they reflected the magnitude of the coefficients in the CL regressions. Respondents' WTP for Expertise2 was $680 (CI, $470–$891), which indicated that patients were on average willing to pay $680 to consult cancer specialists rather than

Table 3.

Characteristics of Respondents in the Discrete Choice Experiment

Table 3.
nurse/GPs, and WTP for TravelTimes2 was $301 (CI, $162–$441), which meant that patients were willing to pay $301 for shorter traveling times of 30 minutes versus 2 to 3 hours.

Variation in Patient Preferences Across Subgroups: The WTP estimates for all appointment-related attributes were analyzed across major cities, inner regional and outer regional residences (Figure 2B). Expertise of the HCP and familiarity of the doctors with patients were the most important attributes for respondents from major cities (cancer specialist vs nurse/GP: ß=0.98; P<.001; usual GP vs new practice: ß=1.09; P<.001) and inner regional areas (cancer specialist vs nurse/GP: ß=1.41; P<.001; usual GP vs new practice: ß=1.04; P<.001). The degree of specialist care (cancer specialist vs nurse/GP: ß=1.14; P<.001) and waiting times (1 vs 6 weeks: ß=1.19; P<.001) strongly influenced the preferences of patients in outer regional areas.

When preferences were analyzed by gender, men strongly favored specialist care (cancer specialist vs nurse/GP: ß=1.56; P<.001) and being accompanied by family/friends (accompaniment vs nonaccompaniment: ß=1.11; P<.001). Being accompanied by family/friends (accompaniment vs nonaccompaniment: ß=0.50; P<.001) was much less important to women than the familiarity (usual GP vs new practice: ß=1.08; P<.001) and expertise (cancer specialist vs nurse/GP: ß=1.03; P<.001) of the HCP.

Discussion

All of the attributes explored in this study were important to patients when choosing between hypothetical health care appointments. The relative importance (from most important to least) of these appointment attributes was expertise of the HCP, familiarity of the doctor with their medical history, waiting time for appointments, availability of social support, and travel times to appointments. Consistent with our results, a DCE9 indicated that general practice patients strongly preferred seeing a doctor of their choice and were willing to trade off speed of access for continuity of care.

By the year 2020, it is estimated that the provision of oncology services will not meet the demand for care from patients with cancer.19 Hence, our patients' preferences for HCP with higher levels of cancer expertise and those familiar with their medical history

Table 4.

Conditional Logistic and Mixed Logistic Regression Models of Patient Preferences

Table 4.
could pose a problem in the future, particularly in rural areas where there are shortages of oncology services.20 Currently, more than 50% of primary care physicians in the United States already assist patients with cancer with their decisions, manage patients' comorbid illnesses, and refer to other specialties.21 The role of oncology nurse practitioners is also evolving, and they provide similar patient-related health outcomes compared with doctors.22 A complementary, co-management approach to cancer care among nurse practitioners, GPs, or primary care physicians and oncologists may alleviate some of the supply/demand burden. However, this approach may require patient education about the roles of nononcology HCPs, because patients in our study appeared to greatly value the expertise of these individuals.

The presence of a social support network was also very important to patients with cancer. These social networks may provide patients with both practical and emotional support that could improve their adherence and compliance to treatment.23 A meta-analysis24 suggested that higher levels of perceived social support, a larger social network, and being married were associated with relative risk reductions in cancer mortality of 25%, 20%, and 12%, respectively. A subgroup analysis of our population also revealed that male patients placed more importance on the presence of social networks compared with female patients. This observation is in keeping with a population-based study25 that examined the relationship between perceived social support and delay in seeking medical attention among patients with cancer. The authors found that the presence of partner support significantly decreased the time from symptoms to medical presentation for male patients. Future campaigns could harness these sex-based differences to improve health care services and service delivery through encouraging men to discuss their health with partners, and to empower their partners to be more attentive to men's health.25

Distance traveled was the least important factor that influenced our population's preferences, especially for respondents from major cities who preferred higher levels of specialist care, and those from outer regional areas who presumably had limited access to specialist health care locally. Previous research has estimated that the median travel time for patients in urban areas to access health care was 12.6 minutes, whereas those from rural areas could travel up to 42 minutes.26 However, those with complex illnesses such as cancer may be “forced” to travel longer distances for their specialist care. Hence, our study reaffirms that patients' willingness to travel27 may not shaped by distance

Figure 2.
Figure 2.

(A) Willingness to pay (WTP) for appointment-related attributes, mean WTP (AUD$) and 95% CI. (B) WTP for appointment-related attributes, mean WTP (AUD$) and 95% CI across major cities, inner regional, and outer regional residences.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 14, 3; 10.6004/jnccn.2016.0036

per se; rather it is associated with the acceptability, access, and affordability of health care and patients' sociodemographic characteristics.

Our study has several limitations. First, DCEs may not represent complex real-life situations given the limited number of attributes and levels tested. Second, misinterpretation of the information and the use of cognitive shortcuts by respondents could oversimplify the tasks and may result in choices that are not considered rational in the economic sense. Third, the DCE results are not representative of all patients with cancer, because we only explored English-speaking adult patients who attended adult oncology services.

However, our study provides valuable information about how patients weigh their care options and trade off different health care appointment features, demonstrating preference heterogeneity among respondents. We studied patient preferences about any health care appointments and did not stratify these visits into initial, follow-up, or treatment appointments. We believe this article provides novel evidence and information with respect to the “average” patient with cancer. We also examined multiple characteristics that influence patient preferences toward health care appointments, and derived WTP estimates in hypothetical appointment settings.

The WTP estimates in our study could guide the cost evaluation of health care models to cope with the increasing costs of medical care.28 The Health and Hospitals Fund (HHF)29 was established in Australia in 2009 to provide funding for developing regional cancer services, upgrading regional health infrastructure, and supporting clinical training in regional settings. The allocated HHF funding for Australian health care models emphasizes some of the attributes that were found to be important to patients in our study, including ease of access to specialist care.29

A patient-centered approach to developing current health care models is likely to improve patient and HCP satisfaction while ensuring that health care funds are allocated in the patients' best interests. A better understanding of patient preferences will guide future investment opportunities as policymakers aim to bridge the gap between the optimal health care models for patient-centered service delivery and the current medical infrastructure.

To further understand the relationship between stated (those elicited in the DCE) and revealed preferences (actual health seeking behavior),30 future research plans include understanding if and how much patients' preferences in relation to their treatment and management plans impact their survival outcomes. Potential confounding factors will be accounted for by analyzing information from the Evaluation of Cancer Outcomes31 database that collects information about patients with cancer presenting to southwestern Victoria, Australia.

Acknowledgments

The authors wish to thank all of the patients and health care professionals who participated in the interviews and discrete choice experiment study, as well as the medical oncologists and reception staff who assisted with participant recruitment across southwestern Victoria, Australia.

See JNCCN.org for supplemental online content.

The authors have disclosed that they have no financial interests, arrangements, affiliations, or commercial interests with the manufacturers of any products discussed in this article or their competitors. Dr. Wong is supported by clinical postgraduate research scholarships from the National Health and Medical Research Council (APP1074400), Rotary Bowel Scan, and Barwon South Western Regional Integrated Cancer Services.

References

  • 1.

    AndersenRM. Revisiting the behavioral model and access to medical care: does it matter?J Health Social Behav1995;36:110.

  • 2.

    FarmerJIversenLCampbellNC. Rural/urban differences in accounts of patients' initial decisions to consult primary care. Health Place2006;12:210221.

    • Search Google Scholar
    • Export Citation
  • 3.

    BrownRButowPWilson-GendersonMMeeting the decision-making preferences of patients with breast cancer in oncology consultations: impact on decision-related outcomes. J Clin Oncol2012;30:857862.

    • Search Google Scholar
    • Export Citation
  • 4.

    BlinmanPKingMNormanR. Preferences for cancer treatments: an overview of methods and applications in oncology. Ann Onco2012;23:11041110.

    • Search Google Scholar
    • Export Citation
  • 5.

    PeacockSApicellaCAndrewsL. A discrete choice experiment of preferences for genetic counselling among Jewish women seeking cancer genetics services. Br J Cancer2006;95:14481453.

    • Search Google Scholar
    • Export Citation
  • 6.

    HolLde Bekker-GrobEWvan DamL. Preferences for colorectal cancer screening strategies: a discrete choice experiment. Br J Cancer2010;102:972980.

    • Search Google Scholar
    • Export Citation
  • 7.

    MuhlbacherACBethgeS. Patients' preferences: a discrete-choice experiment for treatment of non-small-cell lung cancer. Eur J Health Econ2015;16:657670.

    • Search Google Scholar
    • Export Citation
  • 8.

    Cheraghi-SohiSHoleARMeadN. What patients want from primary care consultations: a discrete choice experiment to identify patients' priorities. Ann Fam Med2008;6:107115.

    • Search Google Scholar
    • Export Citation
  • 9.

    GerardKSalisburyCStreetD. Is fast access to general practice all that should matter? A discrete choice experiment of patients' preferences. J Health Serv Res Policy2008;13(Suppl 2):310.

    • Search Google Scholar
    • Export Citation
  • 10.

    WongSFNormanRDunningTL. A protocol for a discrete choice experiment: understanding preferences of patients with cancer towards their cancer care across metropolitan and rural regions in Australia. BMJ Open2014;4:e006661.

    • Search Google Scholar
    • Export Citation
  • 11.

    BabitschBGohlDvon LengerkeT. Re-revisiting Andersen's Behavioral Model of Health Services Use: a systematic review of studies from 1998-2011. Psychosoc Med2012;9:Doc11.

    • Search Google Scholar
    • Export Citation
  • 12.

    ZwerinaKHuberJKuhfeldWF. A general method for constructing efficient choice designs. Available at: https://faculty.fuqua.duke.edu/~jch8/bio/Papers/Zwerina%20Kuhfeld%20Huber.pdf. Accessed February 8 2016.

    • Search Google Scholar
    • Export Citation
  • 13.

    KesselsRGoosPVandebroekM. A comparison of criteria to design efficient choice experiments. J Mark Res2006;43:409419.

  • 14.

    RoseJMBliemerMC. Sample optimality in the design of stated choice experiments. Available at: http://bit.ly/1SGeouV. Accessed February 8 2016.

    • Search Google Scholar
    • Export Citation
  • 15.

    SandorZWedelM. Profile construction in experimental choice designs for mixed logit models. Marketing Science2002;21:455475.

  • 16.

    GuYHoleARKnoxS. Estimating the generalized multinomial logit model in STATA. Stata J2013;13:382397.

  • 17.

    HoleAR. Mixed logit modelling in Stata—an overview. 2013. Available at: http://www.stata.com/meeting/uk13/abstracts/materials/uk13_hole.pdf. Accessed February 8 2016.

    • Search Google Scholar
    • Export Citation
  • 18.

    LongJSFreeseJ. Regression Models for Categorical Dependent Variables using Stata3rd ed.College Station, TX: Stata Press; 2014.

  • 19.

    EriksonCSalsbergEForteG. Future supply and demand for oncologists: challenges to assuring access to oncology services. J Oncol Pract2007;3:7986.

    • Search Google Scholar
    • Export Citation
  • 20.

    UnderhillCBartelRGoldsteinD. Mapping oncology services in regional and rural Australia. Aust J Rural Health2009;17:321329.

  • 21.

    KlabundeCNAmbsAKeatingNL. The role of primary care physicians in cancer care. J Gen Intern Med2009;24:10291036.

  • 22.

    HorrocksSAndersonESalisburyC. Systematic review of whether nurse practitioners working in primary care can provide equivalent care to doctors. BMJ2002;324:819823.

    • Search Google Scholar
    • Export Citation
  • 23.

    DiMatteoMR. Social support and patient adherence to medical treatment: a meta-analysis. Health Psychol2004;23:207218.

  • 24.

    PinquartMDubersteinPR. Associations of social networks with cancer mortality: a meta-analysis. Crit Rev Oncol Hematol2010;75:122137.

    • Search Google Scholar
    • Export Citation
  • 25.

    PedersenAFOlesenFHansenRP. Social support, gender and patient delay. Br J Cancer2011;104:12491255.

  • 26.

    ChanLHartLGGoodmanDC. Geographic access to health care for rural medicare beneficiaries. J Rural Health2006;22:140146.

  • 27.

    ExworthyMPeckhamS. Access, choice and travel: implications for health policy. Soc Policy Adm2006;40:267287.

  • 28.

    CurrieGRDonaldsonCO'BrienBJ. Willingness to pay for what? A note on alternative definitions of health care program benefits for contingent valuation studies. Med Decis Making2002;22:493497.

    • Search Google Scholar
    • Export Citation
  • 29.

    The Department of Health. Evaluating the Health & Hospitals Fund (HHF): Summary of Evaluation Findings. Available at: http://bit.ly/1SGgmvb. Accessed February 8 2016.

    • Search Google Scholar
    • Export Citation
  • 30.

    AliSRonaldsonS. Ordinal preference elicitation methods in health economics and health services research: using discrete choice experiments and ranking methods. Br Med Bull2012;103:2144.

    • Search Google Scholar
    • Export Citation
  • 31.

    BanksPMathesonLMorrissyK. Characteristics of cancer diagnoses and staging in South Western Victoria: a rural perspective. Aust J Rural Health2014;22:257263.

    • Search Google Scholar
    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.

Correspondence: Shu Fen Wong, MBBS, B Med Sci, FRACP, Andrew Love Cancer Centre, University Hospital Geelong, 70 Swanston Street, Geelong 3220, Australia. E-mail: shufen.wong81@gmail.com

Downloadable materials

Article Sections

Figures

  • View in gallery

    An example of a choice task used in the discrete choice experiment.

  • View in gallery

    (A) Willingness to pay (WTP) for appointment-related attributes, mean WTP (AUD$) and 95% CI. (B) WTP for appointment-related attributes, mean WTP (AUD$) and 95% CI across major cities, inner regional, and outer regional residences.

Article Information

PubMed

Google Scholar

Related Articles

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 80 79 10
PDF Downloads 14 14 4
EPUB Downloads 0 0 0