Understanding why decision aids work: linking process with outcome

https://doi.org/10.1016/S0738-3991(03)00056-9Get rights and content

Abstract

Decision aids help patients make treatment choices. There is little empirical evidence to explain how they work. The results from this randomised controlled trial comparing routine with decision-aided consultations in the prenatal diagnosis for Down’s syndrome context are used to describe the strategies employed during decision making, to assess the impact of a decision aid on decision processes, and to investigate decision process and outcome associations. Data were elicited from two content analyses of consultation transcripts and questionnaires assessing knowledge, anxiety, decisional conflict, reasons, and information usefulness. 68/106 women completed measures at consultation and follow-up. Decision-aided women employed more cognitive and emotional strategies during decision making. More negative evaluations during decision making were associated with better outcomes. Decision-aided consultations facilitated the employment of strategies associated with more effective choices. These consultations take longer and elicit greater expressions of negative affect, so may be less rewarding encounters for health professionals.

Introduction

Decision aids are interventions that help individuals focus on a deliberative choice between two or more treatment options [1], [2], [3], [4]. A decision aid has a minimum of two components (a) a visual representation of the risks, benefits and consequences of all decision options relevant to the individual’s health, and (b) an explicit discussion of the individual’s values or attitudes about the decision options and consequences. The theoretical grounding of decision aids varies along a continuum from intuitive choice models to classical decision theory [2], [3]. Those aids underpinned by expected utility theory usually include a third component, an activity that enables the individual to integrate values with full information and encourage trade-offs between options during the decision making process [5], [6].

Decision aids are more effective than routine information in enabling patients to make difficult treatment choices [1]. Their effectiveness can be explained in terms of either the facilitation of cognitive strategies or changes to emotional processes. The mechanisms through which decision aids impact on cognitive strategies are addressed briefly. The visual representations or decision diagrams provide a memory prompt that summarises all the relevant information during decision making [2]. These prompts reduce the cognitive load during decision making [6], [7] and ensure patients’ judgments are made on complete information rather than memory-accessed and/or biased details [8], [9]. The elicitation of normally unarticulated cognitive mechanisms helps patients generate more reasons for and against the decision options by encouraging them to integrate verbally the decision information with their beliefs [2], [5], [6]. It is likely that together these techniques (a) enable patients to justify these difficult choices to themselves and others [8], and (b) ensures patients explore fully the reasons associated with the options, so reducing the gap between decision making and decision experience [10]. As this more systematic evaluation of the decision information leads patients to develop more stable cognitions, they are less likely to change their beliefs over time [9], [11]. These more robust cognitions are associated with less decisional conflict and/or greater decisional satisfaction post-decision making [2], [12].

The mechanisms by which decision aids impact on emotions are less clearly understood. It is accepted that difficult choices create feelings of conflict [8], that these emotions impact adversely on decision making [6], [8] and that evaluating emotions about a decision in retrospect results in greater discontent [9]. The use of decision aids is associated with lower decisional conflict [1] and their effectiveness attributed to the introduction of a ‘dispassionate arbitrator into a decision setting that is distorted by emotion’ [6]. However, there is little explanation as to how or why the ‘dispassionate arbitrator’ diffuses these emotions during decision making. One explanation may be that the decision aid encourages patients to evaluate the decision-relevant information rather than focus on emotions or feelings [2], [6], [8], [9], [11]. An alternative explanation is that as decision aids explicitly elicit patient values about decision options and consequences, this activity increases patients’ expression of emotion during decision making [5]. There is evidence that expression of affect about stressful events is associated with better long-term health outcomes [13], [14]. It is feasible, then that a decision aid may facilitate patient decision making by increasing the expression of affect during decision making.

Despite the increase in studies evaluating decision aid effectiveness [1], there is little evidence to suggest which of the above arguments explain how or why decision aids facilitate patient decision making. Most studies evaluate decision aid effectiveness by measuring outcomes such as choice, utility, knowledge, satisfaction, decisional conflict and anxiety [1], [7] and a handful measure changes in the patient–professional interaction [1], [15]. However, few, if any, studies have assessed patients’ cognitive and emotional processes during decision making, evaluated how these differed by routine or decision-aided consultations, and investigated how such decision process were associated with decision outcome. Until these factors are understood, the decision aid intervention will remain a ‘black box’ [3] with little evidence to explain what and why components of the decision aid are effective.

This paper aims (a) to describe differences in the cognitive and emotional processes of women receiving either routine or a decision-aided consultation about prenatal diagnosis options, and (b) to investigate the relationship between decision processes and outcomes.

Section snippets

Design

A randomised controlled trial with two consultation modes.

Decision context

Upon receipt of a screen positive result for Down’s syndrome, women choose to have or not have a prenatal diagnostic test. The screening test in this study was an analysis of maternal blood. The blood was analysed for hormones that are known to differ slightly between women having and not having Down’s affected pregnancies. Calibrating the blood test result with individual factors generates a risk figure for Down’s syndrome likelihood.

Results

Eighty-nine percent (117/132) of women agreed to take part in the study. Eleven women were lost post-randomisation either because they screened negative following identification of an error on their result sheet (n=6) or failed to complete the post-consultation questionnaire (n=5) (56/58 routine: 50/59 decision aid). Follow-up questionnaires were not sent to five women receiving a positive diagnostic result and one that miscarried. 68/100 posted, follow-up questionnaires were returned (38/53

Discussion

This study provides evidence that decision aids do impact on both the cognitive strategies and emotional mechanisms women employ when making difficult healthcare choices. During decision making, decision-aided women (a) used more cognitive terms, raised more metaphysical terms and evaluated more information about the decision alternatives both positively and negatively, and (b) expressed more negative emotion words and tended to express fewer positive emotion words than women receiving routine

Acknowledgements

We thank Dr. Anna Middleton for her time and insight when carrying out the inter-rater checks, Professor Jamie Pennebaker for his assistance in analysing these data, and the Medical Research Counsel for supporting Hilary Bekker’s studentship.

References (27)

  • P.A. Ubel et al.

    The role of decision analysis in informed consent: choosing between intuition and systematicity

    Soc. Sci. Med.

    (1997)
  • G. Elwyn et al.

    Measuring the involvement of patients in shared decision-making: a systematic review of instruments

    Patient Educ. Counsel.

    (2001)
  • A.M. O’Connor et al.

    Decision aids for patients facing health treatment or screening decisions: a systematic review

    Br. Med. J.

    (1999)
  • A.M. O’Connor

    Validation of a decision conflict scale

    Med. Decision Making

    (1995)
  • H.A. Llewellyn-Thomas

    Patients’ health-care decision making: a framework for descriptive and experimental investigations

    Med. Decision Making

    (1995)
  • O’Connor AM, Stacey D, Rovner D, Holmes-Rovner M, Tetroe J, Llewellyn-Thomas H, et al. Decision aids for people facing...
  • S.P. Pauker et al.

    Prenatal diagnosis: a directive approach to genetic counseling using decision analysis

    Yale J. Biol. Med.

    (1977)
  • Bekker H, Thornton JG, Airey M, Connelly J, Hewison J, Lilleyman J. et al. Informed patient decision making: an...
  • Shafir E, Simonson I, Tversky A. Reason-based choice. In: Johnson-Laird PN, Shafir E, editors. Reasoning and decision...
  • T.D. Wilson et al.

    Introspecting about reasons can reduce post-choice satisfaction

    Person Soc. Psychol. Bull.

    (1993)
  • D. Frisch et al.

    Assessing the accuracy of decisions

    Theor. Psychol.

    (1993)
  • S. Chaiken

    Heuristic versus systematic information processing and the use of source versus message cues in persuasion

    J. Person Soc. Psychol.

    (1980)
  • M. Holmes-Rovner et al.

    Patient satisfaction with health care decisions: the satisfaction with decision scale

    Med. Decision Making

    (1996)
  • Cited by (104)

    • Evaluation of a prenatal screening decision aid: A mixed methods pilot study

      2022, Patient Education and Counseling
      Citation Excerpt :

      In pregnancy care, the use of DAs showed positive effects on informed decision making [8]. For prenatal testing decisions in particular, they seem to reduce decisional conflict and help women and their partners to make choices informed by evidence and by what matters most to them [12–15]. Despite the efficacy of DAs, however, they are rarely used for prenatal screening [16] or indeed in any context [17,18].

    View all citing articles on Scopus
    View full text