Decisions about Pap tests: What influences women and providers?

https://doi.org/10.1016/j.socscimed.2009.03.002Get rights and content

Abstract

Despite the success internationally of cervical screening programs debate continues about optimal program design. This includes increasing participation rates among under-screened women, reducing unnecessary early re-screening, improving accuracy of and confidence in screening tests, and determining the cost-effectiveness of program parameters, such as type of screening test, screening interval and target group. For all these issues, information about consumer and provider preferences and insight into the potential impact of any change to program design on consumer and provider behaviour are essential inputs into evidence-based health policy decision making. This paper reports the results of discrete choice experiments to investigate women's choices and providers' recommendations in relation to cervical screening in Australia. Separate experiments were conducted with women and general practitioners, with attributes selected to allow for investigation of how women and general practitioners differ in their preferences for attributes of screening programs. Our results indicate a considerable commonality in preferences but the alignment was not complete. Women put relatively more weight on cost, chance of a false positive and if the recommended screening interval were changed to one year.

Introduction

While population based screening programs such as the National Cervical Screening Program (NCSP) in Australia have successfully reduced morbidity and mortality from cervical cancer there remain many policy challenges in cervical screening, including how to improve participation of under-screened women, avoid early re-screening and maintain and improve accuracy of and confidence in cervical screening tests. Other policy decisions relevant to the cost-effectiveness of cervical screening include decisions about the type of screening test, screening interval, and age to commence and cease screening.

The impact on costs and outcomes of changes to screening programs depends on how consumers and providers behave in response to the policy change. This depends on characteristics of the woman and the provider, as well as characteristics of tests and other program parameters. Thus, information about women's and providers' preferences and insight into the potential impact of any change to the current program on the behaviour of women and providers are essential inputs into evidence-based health policy decision making.

Existing data sources provide limited information for analysis of consumer and provider preferences for screening, or behavioural responses to potential policy changes. For example, in Australia most information about screening is at the aggregate level and limited to estimates of participation rates and outcomes of screening. The Australian Bureau of Statistics (ABS) National Health Survey provides some individual level data that allows analysis of characteristics that influence screening choice (Belkar, Fiebig, Haas, & Viney, 2006). This comprises a single self-report observation of a screening decision, with no information about the decision context. This has limited value in predicting behaviour in relation to policy parameters, or in modelling the consumer–provider relationship.

This paper uses stated preference data from a discrete choice experiment (DCE) to evaluate consumers' and providers' preferences for attributes of alternative cervical screening tests, proposed changes to policy recommendations and potential new technologies. Stated preference data are useful when market (revealed preference) data are not available, or have limited variability. Both situations frequently arise for health programs, because of innovation or policy changes. Further, to predict demand for screening in different policy settings accurately, it is most valid to model both consumers' and providers' preferences. Revealed preference data provide information about the ultimate choice, but not about the consumer–provider context leading to this choice. A choice experiment provides the capacity to investigate these relationships.

Separate experiments were conducted with women and General Practitioners (GPs) to determine the impact of a common set of attributes on their choices and recommendations, respectively. Characteristics of women were incorporated in the choice situations presented to GPs and characteristics of GPs, including their recommendations, were incorporated in the choice situations presented to women. This allows comparison of the preferences of each group. The approach has been applied elsewhere in modelling consumer demand where agency relationships are important (Bartels, Fiebig, & Soest, 2006).

Section snippets

Screening for cervical cancer

In Australia, the NCSP aims to ensure that all women aged 20–69 receive a biennial cervical screening test. The test is most commonly provided at a primary care encounter. There is information asymmetry, and the incentives facing the provider may not be the same as those facing the consumer – for example, providers may be influenced by risk of litigation. Factors relevant to women's decisions about cervical screening include doctor's recommendation, previous experience with cervical

Survey design and data collection

DCE methods have been described elsewhere (Burgess et al., 2006, Viney et al., 2002). Separate choice experiments were undertaken by two samples of women in the target population (previously screened and never-screened women) and a sample of GPs. The relationship between the provider and consumer is addressed by inclusion of consumer characteristics in the providers' experiment and provider characteristics in the women's experiment. The current paper focuses on previously screened women and GPs.

Estimation methods

Statistical analysis of choice data relies on the random utility model (McFadden, 1973, McFadden and Train, 2000, Thurstone, 1927) where each respondent faces a choice amongst J alternatives repeated under S scenarios or choice situations. The utility individual i derives from alternative j in scenario s is composed of systematic and random components denoted byUisj=X′isjβi+ɛisjwhere Xisj is a K × 1 vector of explanatory variables and βi is a conformable vector of coefficients.

Conditional on βi,

Estimation results

Log-likelihood values for each of the MXL specifications are compared with the standard multinomial logit model in Table 5. The MXL models with intercepts as random coefficients provide a dramatic improvement in fit over the MNL models. While these models are nested, the hypothesis tests are non-standard because the parameter space is restricted under the alternative. In such situations the Likelihood Ratio (LR) test statistic does not have the usual chi-square asymptotic distribution (Andrews,

Discussion

To our knowledge this is the first choice experiment in health to examine the relationship between preferences of experts and consumers in a setting where the expert (provider) is in an agency relationship with the consumer. Other studies have examined issues of agency by considering patients' preferences for characteristics of providers (Vick & Scott, 1998) but not compared provider and patient preferences. A previous paper has examined providers' and women's preferences for cervical screening

Acknowledgements

A preliminary version of this paper was presented at the Australian Health Economics Society Conference held in Perth, September 2006. The research was supported by the National Health and Medical Research Council through a Program Grant (Grant No: 254202). We thank anonymous reviewers for their helpful comments. All remaining errors are the responsibility of the authors.

References (35)

  • R.M. Austin et al.

    Increased detection of epithelial cell abnormalities by liquid-based gynecologic cytology preparations. A review of accumulated data

    Acta Cytologica

    (1998)
  • Australian Health Technology Advisory Committee

    Review of automated and semi-automated cervical screening devices

    (1998)
  • R. Bartels et al.

    Consumers and experts: an econometric analysis of the demand for water heaters

    Empirical Economics

    (2006)
  • R. Belkar et al.

    Why worry about awareness in choice problems? Econometric analysis of screening for cervical cancer

    Health Economics

    (2006)
  • J.G. Boyce et al.

    The fallacy of the screening interval for cervical smears

    Obstetrics & Gynecology

    (1990)
  • L. Burgess et al.

    Design of choice experiments in health economics

  • R.T. Carson

    Contingent valuation: a user's guide

    Environmental Science & Technology

    (2000)

    • Cited by (29)

    • Measuring High-Risk Patients’ Preferences for Pharmacogenetic Testing to Reduce Severe Adverse Drug Reaction: A Discrete Choice Experiment

      2016, Value in Health
      Citation Excerpt :

      Eight percent of respondents always preferred the physician-recommended option, even when the recommended option had high risk and/or costs. This is consistent with other studies that found strong impact of physician recommendations on patients’ choices [21–27]. Information on physician recommendations is influential in both cost-conscious and risk-averse individuals.

    • Hospital sector choice and support for public hospital care in New Zealand: Results from a labeled discrete choice survey

      2015, Journal of Health Economics
      Citation Excerpt :

      The resulting factors (including waiting time for the surgery, contact before and after surgery with their surgeon, the risk of complications from the surgery, and the amenities of the hospital room) were varied in the DCE across decision scenarios, with the participants always facing a choice between having a procedure done in a ‘Private Hospital’ or a ‘Public Hospital.’ The use of labels, which is common in marketing research, has been used to examine healthcare choices, including the use of test labels for colorectal screening (de Bekker-Grob et al., 2010), esophagus cancer screening (Kruijshaar et al., 2009), and cervical cancer screening (Fiebig et al., 2009). In this study, the label was intended to estimate the strength of preferences for the hard-to-measure or intangible aspects of provider care that might be associated with a private or public hospital.

    • Recommendations on screening for cervical cancer

      2013, CMAJ. Canadian Medical Association Journal
      Citation Excerpt :

      Other factors affecting the willingness to undergo screening include fear, fatalistic attitudes, embarrassment, fear of pain or discomfort, anxiety and stress related to diagnosis, distrust of the health care system and the belief that screening is not necessary without illness.11 The limited evidence regarding patient preferences for screening intervals82,83 suggests that women who are used to undergoing frequent screening prefer the feeling of security provided by shorter rather than longer screening intervals, and that they are concerned that recommendations for longer intervals between screens are primarily a means to save costs.84 Therefore, the potential harms and benefits should be discussed between patient and provider for informed decision-making.

    • What Breast Cancer Screening Program do Rural Women Prefer? A Discrete Choice Experiment in Jiangsu, China

      2024, Patient

    • Cervical cancer (over-)screening in Europe: Balancing organised and opportunistic programmes

      2023, Scandinavian Journal of Public Health

    • Attributes Used for Cancer Screening Discrete Choice Experiments: A Systematic Review

      2022, Patient
    View all citing articles on Scopus
    View full text
    Copyright © 2009 Elsevier Ltd. All rights reserved.