Medical Decision Making

Medical Decision Making http://mdm.sagepub.com/

Patient Decision Aids to Support Clinical Decision Making: Evaluating the Decision or the Outcomes of the Decision Kirsten McCaffery, Les Irwig and Patrick Bossuyt Med Decis Making published online 14 September 2007 DOI: 10.1177/0272989X07306787 The online version of this article can be found at: http://mdm.sagepub.com/content/early/2007/09/14/0272989X07306787 A more recent version of this article was published on - Oct 5, 2007

Published by: http://www.sagepublications.com

On behalf of:

Society for Medical Decision Making

Additional services and information for Medical Decision Making can be found at: Email Alerts: http://mdm.sagepub.com/cgi/alerts Subscriptions: http://mdm.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

Version of Record - Oct 5, 2007 >> OnlineFirst Version of Record - Sep 14, 2007 What is This?

Downloaded from mdm.sagepub.com by guest on June 4, 2013

Med Decis Making OnlineFirst, published on September 14, 2007 as doi:10.1177/0272989X07306787

Patient Decision Aids to Support Clinical Decision Making: Evaluating the Decision or the Outcomes of the Decision Kirsten McCaffery, PhD, Les Irwig, MBBCh, PhD, Patrick Bossuyt, PhD

Decision aids (DAs) are tools to support patients make informed health decisions with their practitioner. They aim to improve patient knowledge of options, incorporate patient preferences and values, and increase patient involvement in health decision making. Increasingly, the debate about DAs concerns how they should be implemented in practice, with the view that DAs are superior to usual clinical care in facilitating health decisions. The authors challenge this view and suggest that DA research has focused on measures of decision process, leaving the effects on the outcome of the decision relatively unknown. It is still unclear in which conditions DAs are better for patient health and well-being than clinician-led decisions. The authors present a new

randomized design to examine the effects of DA-supported patient choice on patient-centered outcomes to identify where DAs are best implemented in clinical practice. In this design, patients are randomized to 1 of 4 arms: intervention A, intervention B, choice of either intervention supported by a clinician, or choice of either intervention supported by a decision aid. Health and quality of life measured over the long term are presented as the primary outcomes. The authors propose that this design will allow the proper assessment of different modes of decision making. Keywords: decision aids; shared decision making; randomized controlled trial design; patient choice; methodology. (Med Decis Making 2007;XX:xx–xx)

SUPPORTING SHARED DECISION MAKING

practitioner/provider. They aim to improve patient knowledge of options, incorporate the patients’ preferences and values, and increase patient involvement in the health decision-making process. This approach fits within the model of shared decision making1–3 and is often seen as most relevant in preference-sensitive situations, in which the patient desires to be involved in the decision process.4 Increasingly, the debate about DAs concerns how they should be implemented in practice, with the view that DAs are superior to usual clinical care in facilitating health decision making.5 In this article, we challenge that view and present a new randomized design to assess the long-term effects of DA-supported patient choice on patient-centered outcomes. A patient DA is defined as an intervention that provides information on the clinical options and outcomes relevant to a person’s health and is designed to help people make specific and deliberative choices in their health care.6,7 DAs are explicit about choices and encourage patients to express their preference in clinical situations. A systematic review of 24 DAs6,7 suggested that DAs increase knowledge and involvement in the decision process,

Decision aids (DAs) are tools to support consumers make informed health decisions with their Received 6 December 2006 from the School of Public Health, University of Sydney, Sydney, Australia (KM, LI), and the Department of Clinical Epidemiology & Biostatistics, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands (PB). KM, LI, and PB are jointly responsible for the conception of the article. KM wrote the 1st draft, and all authors made a substantial intellectual contribution and assisted in reviewing and editing the manuscript. All authors are guarantors. We would like to thank Professor Chris Del Mar and Dr Lyndal Trevena for commenting on earlier drafts of the article. This work was supported in part by an Australian National Health and Medical Research Council (NHMRC) grant 402764 to the Screening and Test Evaluation Program. KM is supported by an NHMRC Career Development Award (402836). The NHMRC played no role in the writing of this article. None of the authors have any conflict of interest. Ethics approval was not required for the material contained in this article. Revision accepted for publication 30 May 2007. Address correspondence to Kirsten McCaffery, PhD, School of Public Health, Edward Ford Building, A27, University of Sydney NSW 2006, Australia; e-mail: [email protected]. DOI: 10.1177/0272989X07306787

MEDICAL DECISION MAKING/MON–MON 2007 Downloaded from mdm.sagepub.com by guest on June 4, 2013

Copyright 2007 by Society for Medical Decision Making.

619

MCCAFFERY, IRWIG, BOSSUYT

and that they create more realistic expectations and reduce uncertainty in decision making (decisional conflict) without increasing anxiety among consumers. In light of these findings, some researchers have suggested that the evidence for the use of DAs in practice is now so strong that it may be considered a class effect.8 Others suggest that DA research should now focus on how to implement DAs rather than whether it is appropriate to use them.7

dissonance and decisional regret. The authors argued that DAs should be assessed according to their impact on decisional regret measures, which may be taken several months after decision making and therapy. DAs should therefore aim to reduce decisional conflict after the decision is made and the treatment experienced. A NEW APPROACH TO EVALUATING DECISION AIDS

EVALUATING DECISION AIDS: SHARED DECISION MAKING Recent debate has raised questions about the evaluation of DAs.9 Among trials to date, evaluation has focused on process measures that reflect the patient’s short-term experience of the decision-making process.7 The consistent effect shown by DAs has been to enhance knowledge and decision-making outcomes such as decisional conflict and involvement in decision making. According to this literature, the implicit goal of the DA is to improve patient knowledge and experience of the decision-making process7; however, this is rarely overtly stated. More recently, it has been argued that DAs should aim to achieve consistency between patient values and choice,10,11 with the view that DAs should be evaluated according to the consistency achieved between these 2 constructs. Similarly, Charles and others12 asserted that the evaluation of DAs should be tied directly to the goals the DA is setting out to achieve. They suggested that the goal of each DA should be made conceptually explicit with a clear view of the mechanisms that the DA is intended to impact. According to the views outlined above, evaluation has remained focused on the short-term effects of the decision process, a perspective that is in part driven by concerns that decision making should be evaluated before the outcome of the decision is known. This is due to concerns about the phenomena known as outcome bias: the tendency to reevaluate events in light of their outcome (i.e., the bias toward rating decisions with poor outcomes poorly).13,14 We suggest that although no individual decision alone should be evaluated after the outcome, the assessment of multiple decisions across a group of patients after the outcome is known should produce an appropriate assessment. An alternative approach is put forward by Feldman-Stewart and others15 based on the DiffCon theory of decision making.16 The theory posits that the goal of decision making is to reduce cognitive

Although Charles, Feldman-Stewart, and others raised an important debate about the aim and evaluation of DAs, their commentary remains framed around the short-term effects on the decision process (process measures) rather than the state of the patient after decision making (outcome measures). We suggest that patient outcomes extending into the long term are of primary importance and take priority over the process measures of decision making. We define patient outcomes in terms of quality of life (QoL) with respect to patient survival, function, and wellbeing. We suggest that the primary goal of a DA should be to maximize the long-term trajectory of QoL outcomes for the patient across the short, medium, and long term rather than to improve the quality of the decision-making process for the patient at the time of the decision, as is currently emphasized in the literature. A DA may plausibly make decision making a longer and more complex process. However, if it results in patients making decisions that ultimately give them better outcomes across the short, medium, and long term, the DA may be viewed as successful. In contrast, if a DA makes the decision process quicker, simpler, and more satisfying at the time of decision making but results in patients making decisions that lead to decreased longevity, poorer functional health, and poorer QoL, we may argue that it has failed. Our view is consistent with Entwistle and others,17 who suggest that decision-making behavior, health status, and well-being should be prioritized over measures such as knowledge and perceptions of the decision process. Importantly, we shall not assume that optimizing patients’ perception of the decision process will lead to better health and QoL outcomes later and represent an appropriate surrogate for decision-making quality. We suggest that the goal of decision making should be to help patients choose the management that would give them the greatest chance of benefit v. harm tradeoff for QoL as defined by the patient over the short, medium, and long term. Evaluation of

620 • MEDICAL DECISION MAKING/MON–MON 2007 Downloaded from mdm.sagepub.com by guest on June 4, 2013

PATIENT DECISION AIDS TO SUPPORT CLINICAL DECISION MAKING

DAs should therefore be tied directly to this goal. With the current focus on the decision process measures, we suggest that the long-term effects of DAs on patient outcomes have been overlooked. In the 24 DA trials included in the 2003 Cochrane Review, all studies used decision process measures (e.g., decisional conflict) as the primary trial outcome. Only 5 studies included any measures of health or QoL following decision making.18–22 Of those, only 3 of the trials were sufficiently powered to detect significant differences between the groups.18,21,22 COMPARING DECISION AIDS-SUPPORTED CHOICE WITH CLINICIAN-LED CHOICE Although DAs are designed to support and enhance health decision making, there are convincing reasons why they may not always achieve these aims. Health decision making is complex. Most health decisions are novel to the patient, emotional, and require the individual to predict how he or she will feel and cope with an unknown health state and understand the chance/probability of its occurring. This may be simplified to 3 key tasks: 1) to predict the utility (how the patient feels and values) of each outcome, 2) to understand the probability of the various outcomes, and 3) to combine the probability with the utility for each outcome. This is not a simple task, and we know that patients have problems with each of the 3 stages. First, research consistently shows that patients have difficulty understanding probabilities.23 Second, patient utilities for many health states are highly unstable (e.g., women’s preferences for anesthesia before labor differ markedly from their preferences during labor).24 Third, combining a poorly understood probability with an unstable utility is likely to produce variable outcomes for the patient (Figure 1). Despite awareness of the complexity of health decision making, there has been a tendency to assume that patients themselves are always best equipped to decide which therapy/option will give them the best outcome over time and realistically predict what their posttreatment/therapy utility will be. Ubel25 catalogued a series of situations in which patients make decisions inconsistent with their values and preferences because of commonly occurring biases and features of health decision making. These include cognitive bias, information overload, use of minimization and avoidant coping strategies,26,27 and high states of anxiety and emotion that hinder information processing and decision making.28 For further discussion of the difficulty associated with the

measurement of preferences and decision making, see Nelson and others29 in this issue of the journal. At present, we do not know in which situations the process of supporting decision making is improved by the use of a DA and in which situations a traditional clinician-led or -recommended approach may produce better long-term outcomes for the patient. The decision process involves making a careful prediction of the patient’s state posttherapy and selecting the option that will give the patient the best outcomes according to his or her preferences. DAs function by encouraging patients to understand the options and the probability of outcomes and to identify their values to select their preferred option in collaboration with their clinician. In contrast, in the traditional decision-making approach, the clinician selects and recommends the option he or she thinks is best for the patient based on his or her knowledge of the patient and his or her clinical experience. We suggest that it is plausible that in some clinical scenarios (including some preference-sensitive scenarios), an experienced clinician may be better placed to predict how well patients will cope with and adapt to a health state and direct a patient to an option that will optimize patient outcomes. This may be superior to that produced by the use of a DA and a shared decision-making approach. For example, an experienced clinician may be better placed to advise a reluctant diabetic patient with peripheral artery stenosis to have an early amputation before further complications are experienced that reduce the patient’s function and well-being. Such a directive approach may produce better outcomes for the patient than a shared decision-making approach. We suggest that DAs should be compared to clinicians in supporting decision making across long-term patient outcomes in randomized trials before we assume that DAs are a blanket solution to the challenges of health decision making. Although we emphasize the primacy of outcomes across the short, medium, and long term, we recognize that increasing the quality of the decisionmaking process at the time of decision making is important and may enhance QoL outcomes for the patient. There are several pathways through which enhancing patient decisions by involving patients in decision making and supporting patient preferences may affect health outcomes.30,31 In the preference trial literature, patient preferences are seen to potentially affect health outcomes by increasing motivation to adhere to the preferred treatment, increasing effort toward making the preferred treatment work, such as increased tolerance for inconvenience/difficulties

DECISION AIDS: PAST, PRESENT, AND FUTURE Downloaded from mdm.sagepub.com by guest on June 4, 2013

621


Point of decision making

Clinician

Patient

Prediction of future outcomes for the patient 1. Understanding the probability of each outcome 2. Estimating utility of each outcome. 3. Combining probability with utility for each outcome

Process measures: Decision making quality

Quality of DM Knowledge Decisional conflict Involvement in DM Realistic expectations Attitudes to DM Choice Satisfaction with DM Decision regret

Outcome measures: QoL trajectory

QoL [short, medium, long term] Survival Function Well-being

TIME Figure 1

The decision process and evaluation of outcomes.

that arise during the treatment.32 There may also be positive psychological reactions to the preferred intervention. These effects of choice may plausibly affect health and QoL outcomes and suggest mechanisms by which patient choice may enhance longer term patient well-being, function, and survival. Importantly, DA research has mostly focused on the decision process and has failed adequately to investigate DAs’ impact on longer-term patient outcomes. Evaluation in well-conducted randomized trials to determine the long-term outcomes for patients is needed, with potential mediators (e.g., adherence, psychological response) also investigated at stages along the pathway to identify any impact on well-being. Such research is necessary before DAs are more universally implemented, as is standard practice for many health interventions. HOW SHOULD WE EVALUATE THE LONG-TERM HEALTH IMPACT OF DECISION AIDS? To answer this question, we propose a randomized design in which patients are randomly allocated to 1 of 4 arms (see Figure 2). Arm 1 and arm 2 represent a

standard randomized controlled trial (RCT), in which patients are allocated to the clinical intervention A or intervention B. These arms represent 2 possible interventions, both of which are currently considered acceptable practice. They may be considered acceptable because they have been shown to be similarly effective, because their relative effectiveness is not known, or because, whatever the evidence, both are currently in use, perhaps with different interventions being used by different practitioners. Arm 3 represents clinician-led decision making, in which both options A and B are on offer and the clinician guides the patient to the management he or she thinks is best. This arm represents the situation in which there are advantages and disadvantages for each option, and patient preferences or characteristics may make one approach more favorable for a particular patient than another. In arm 4, patients are given a DA, and they choose in collaboration with their clinician between intervention A and B. The design allows the optimum approach to decision making to be determined, comparing outcomes where there is no choice (arms 1 and 2) to the additional effects of a clinician-led



Participants

R Arm 1: Rx A

Arm 2: Rx B

Arm 3: Clinicianled choice A or B

NO CHOICE A

Arm 4: DA-supported choice A or B

CHOICE B

A

B

A

B

Process measures: decision process and quality Outcome measures: Survival, function and well-being [short, medium and long term]

Figure 2 Four-arm, decision-support, randomized controlled trial design.

choice (arm 3) and DA-supported choice (arm 4). If arms 1 and 2 have already been carried out in prior trials, we still recommend their inclusion in the proposed design because this will allow outcomes to be directly compared across all 4 arms. Estimates of effect from prior trials may not be generalizable to the spectrum of patients now considering the alternative therapies. Therefore, randomizing to the 4 groups seems appropriate and ethical, as the new trial design will be used when it is unclear whether fixed policy A or B, clinician-led choice, or DAs have the better outcome. Comparing the 2 choice arms against estimates of the effect of policies A or B derived from prior trials is an indirect comparison, which will be less satisfactory. RCTs involving several arms require a larger sample size than RCTs with only 2 arms. However, they are necessary to address the question about the effects of DAS and are feasible, as shown by the fact that they have been successfully run.33–37 Janevic and others35 compared the impact of choice in the comparison of 2 health education programs (group

v. self-directed) with participants randomized to a choice arm (in which they select their preferred education program), to a no-choice arm (in which they were randomly allocated to 1 of the 2 programs on offer), or to a control group. Noel and others36 used a similar design to investigate the impact of choice on a diabetes education program. Outcome measures should be chosen to assess patient survival, function (disability), and psychological well-being over the long term. Obviously, not all conditions are amenable to these outcome measures, and conditions in which survival and function can be measured over the long term are necessary for the design to work. The proposed design builds on that put forward by Rucker,38 which combines a preference trial design and a standard RCT. Participants are randomized to either the preference arm, in which those with a strong preference are matched to their preferred management and those without preference are randomized to A or B, or to the standard RCT arm, in which patients are randomized to either intervention A or B. Our addition to this


623


design is to add the DA as a means to support decision making, to compare this to both a standard RCT design and clinician-led choice, and to emphasize the importance of long-term follow-up in the domains described. We describe a prototypical example of the comparison between a DA-led choice and a clinicianled choice. There are numerous alternative methods of supporting decision making and choice between patient and clinician. These can be investigated using the proposed trial design either by adding further trial arms or by replacing existing ones. For example, there may be tools that support clinicians in eliciting patient preferences and lead the patient to the most suitable choice. Such an intervention might substitute or may be added to the trial design or replace either arm 3 or arm 4, depending on the research question. It is also important to measure intermediate outcomes along the pathway to obtain a thorough as possible understanding of why one intervention might perform better than another. This might include measuring adherence to medication or to health behavioral advice and attitudes to therapy, all of which may plausibly affect long-term outcomes of one clinical option over another. To illustrate our approach, we have selected a DA trial18 from the Cochrane Review. The study was a 2-arm randomized trial, and although it does not illustrate the full 4-arm design, it provides an example in which health and QoL outcomes have been comprehensively assessed in the short, medium, and long term in addition to standard measures of decision process. The trial examined a multimedia shared decision-making program for men facing a treatment decision for benign prostatic hyperplasia. In total, 227 men were randomized to the shared decision-making intervention or to the control group. Participants were followed for 1 year. The study measured the patient’s treatment choice (prostatectomy, medication, or watchful waiting) as the primary outcome. Quality of the decision was measured by assessing knowledge 2 weeks after decision making and satisfaction at 3 and 12 months. Disease-specific health measures (severity of urinary symptoms and their impact on the patient’s daily life) and QoL (perceptions of general health, physical function, and social function subscales of the SF36) were recorded at 3, 6, and 12 months postintervention. Overall, the trial illustrates an impressive attempt to assess the impact of a DA on the decision made and its consequences for the patient over the long term.

CONCLUSION Decision aid trials have the potential to mislead if the primary outcomes for evaluation remain fixed on the decision process. We need to ensure DAs bring long-term benefit to patients and understand in what situations they are most usefully applied and where they are not. This requires randomized trials to examine the effects of DAs on health and QoL outcomes over the long term. Decision making is more than the effect on the decision process; the result of the decision matters. Helping patients make decisions with which they can best live and function is what counts and is therefore what we should be evaluating. REFERENCES 1. Charles C, Gafni A, Whelan T. Shared decision-making in the medical encounter: what does it mean? (or it takes at least two to tango). Soc Sci Med. 1997;44(5):681–92. 2. Charles C, Gafni A, Whelan T. Decision-making in the physician-patient encounter: revisiting the shared treatment decision-making model. Soc Sci Med. 1999;49(5):651–61. 3. Godolphin W. The role of risk communication in shared decision making. BMJ. 2003;327(7417):692–3. 4. Whitney SN, McGuire AL, McCullough LB. A typology of shared decision making, informed consent, and simple consent. Ann Intern Med. 2004;140(1):54–9. 5. O’Connor AM, Graham ID, Visser A. Implementing shared decision making in diverse health care systems: the role of patient decision aids. Patient Educ Couns. 2005;57(3):247–9. 6. O’Connor AM, Rostom A, Fiset V, et al. Decision aids for patients facing health treatment or screening decisions: systematic review. BMJ. 1999;319(7212):731–4. 7. O’Connor AM, Stacey D, Entwistle V, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2003;(2):CD001431. 8. Thornton H, Edwards E, Baum M. Not to act is to act. BMJ Rapid Response. 14 October 2003. 9. Coulter A. Patient information and shared decision-making in cancer care. Br J Cancer. 2003;89(Suppl 1):S15–6. 10. Kennedy AD. On what basis should the effectiveness of decision aids be judged? Health Expect. 2003;6(3):255–68. 11. Sepucha KR, Mulley AG. Extending decision support: preparation and implementation. Patient Educ Couns. 2003;50(3): 269–71. 12. Charles C, Gafni A, Whelan T, O’Brien MA. Treatment decision aids: conceptual issues and future directions. Health Expect. 2005;8(2):114–25. 13. Baron J, Hershey JC. Outcome bias in decision evaluation. J Pers Soc Psychol. 1988;54(4):569–79. 14. Caplan RA, Posner KL, Cheney FW. Effect of outcome on physician judgments of appropriateness of care. JAMA. 1991;265(15): 1957–60.



15. Feldman-Stewart D, Brundage MD. Challenges for designing and implementing decision aids. Patient Educ Couns. 2004;54(3): 265–73. 16. Svenson O. Differentiation and consolidation theory of human decision making: a frame of reference for the study of pre- and post decision processes. Acta Psychol (Amst) 1992;80:143–68.

27. Lazarus RS, Folkman S. Stress, Appraisal and Coping. New York: Springer; 1984.

17. Entwistle VA, Sowden AJ, Watt IS. Evaluating interventions to promote patient involvement in decision-making: by what criteria should effectiveness be judged? J Health Serv Res Policy. 1998;3(2):100–7.

29. Nelson WL, Han P, Fagerlin A, Stefanek M, Ubel P. Rethinking the objectives of decision aids: a call for conceptual clarity. Med Decis Making. 2007;27:000–000.

18. Barry MJ, Cherkin DC, YuChiao C, Fowler FJ, Skates S. A randomized trial of a multimedia shared decision-making program for men facing a treatment decision for benign prostatic hyperplasia. Dis Manag Clin Outcomes. 1997;1(1):5–14. 19. Bernstein SJ, Skarupski KA, Grayson CE, Starling MR, Bates ER, Eagle KA. A randomized controlled trial of informationgiving to patients referred for coronary angiography: effects on outcomes of care. Health Expect. 1998;1(1):50–61. 20. Davison BJ, Degner LF. Empowerment of men newly diagnosed with prostate cancer. Cancer Nurs. 1997;20(3):187–96. 21. Deyo RA, Cherkin DC, Weinstein J, Howe J, Ciol M, Mulley AG Jr. Involving patients in clinical decisions: impact of an interactive video program on use of back surgery. Med Care. 2000; 38(9):959–69. 22. Morgan MW, Deber RB, Llewellyn-Thomas HA, et al. Randomized, controlled trial of an interactive videodisc decision aid for patients with ischemic heart disease. J Gen Intern Med. 2000; 15(10):685–93. 23. Schwartz LM, Woloshin S, Black WC, Welch HG. The role of numeracy in understanding the benefit of screening mammography. Ann Intern Med. 1997;127(11):966–72. 24. Christensen-Szalanski JJ. Discount functions and the measurement of patients’ values: women’s decisions during childbirth. Med Decis Making. 1984;4(1):47–58.

28. Shanteau J, Dino G. Environmental stressor effects on creativity and decision making. In: Svenson O, Maule A, eds. Time Pressure and Stress in Human Judgment and Decision Making. New York: Plenum; 1993. p 293–308.

30. Greenfield S, Kaplan S, Ware JE Jr. Expanding patient involvement in care: effects on patient outcomes. Ann Intern Med. 1985; 102(4):520–8. 31. Kaplan SH, Greenfield S, Ware JE Jr. Assessing the effects of physician-patient interactions on the outcomes of chronic disease. Med Care. 1989;27(3 Suppl):S110–27. 32. Bradley C. Patients’ preferences and randomised trials. Lancet. 1996;347(9008):1118–9. 33. Cooper KG, Bain C, Parkin DE. Comparison of microwave endometrial ablation and transcervical resection of the endometrium for treatment of heavy menstrual loss: a randomised trial. Lancet. 1999;354(9193):1859–63. 34. Cooper KG, Parkin DE, Garratt AM, Grant AM. Two-year follow up of women randomised to medical management or transcervical resection of the endometrium for heavy menstrual loss: clinical and quality of life outcomes. Br J Obstet Gynaecol. 1999; 106(3):258–65. 35. Janevic MR, Janz NK, Dodge JA, et al. The role of choice in health education intervention trials: a review and case study. Soc Sci Med. 2003;56:1581–94. 36. Noel PH, Larme AC, Meyer J, Marsh G, Correa A, Pugh JA. Patient choice in diabetes education curriculum: nutritional versus standard content for type 2 diabetes. Diabetes Care. 1998;21 (6):896–901.

25. Ubel PA. Is information always a good thing? Helping patients make ‘‘good’’ decisions. Med Care. 2002;40(9 Suppl):V39–44.

37. Rokke PD, Tomhave JA, Jocic Z. The role of client choice and target selection in self-management therapy for depression in older adults. Psychol Aging. 1999;14(1):155–69.

26. Croyle RT, Hunt JR. Coping with health threat: social influence processes in reactions to medical test results. J Pers Soc Psychol. 1991;60(3):382–9.

38. Rucker G. A two-stage trial design for testing treatment, selfselection and treatment preference effects. Stat Med. 1989;8(4): 477–85.


625