Patients’ Perceptions of Sharing in Decisions

A Systematic Review of Interventions to Enhance Shared Decision Making in Routine Clinical Practice
  • France Légaré
  • Stéphane Turcotte
  • Dawn Stacey
  • Stéphane Ratté
  • Jennifer Kryworuchko
  • Ian D. Graham
Systematic Review

Abstract

Background: Shared decision making is the process in which a healthcare choice is made jointly by the health professional and the patient. Little is known about what patients view as effective or ineffective strategies to implement shared decision making in routine clinical practice.

Objective: This systematic review evaluates the effectiveness of interventions to improve health professionals’ adoption of shared decision making in routine clinical practice, as seen by patients.

Data Sources: We searched electronic databases (PubMed, the Cochrane Library, EMBASE, CINAHL, and PsycINFO) from their inception to mid-March 2009. We found additional material by reviewing the reference lists of the studies found in the databases; systematic reviews of studies on shared decision making; the proceedings of various editions of the International Shared Decision Making Conference; and the transcripts of the Society for Medical Decision Making’s meetings.

Study Selection: In our study selection, we included randomized controlled trials, controlled clinical trials, controlled before-and-after studies, and interrupted time series analyses in which patients evaluated interventions to improve health professionals’ adoption of shared decision making. The interventions in question consisted of the distribution of printed educational material; educational meetings; audit and feedback; reminders; and patient-mediated initiatives (e.g. patient decision aids).

Study Appraisal: Two reviewers independently screened the studies and extracted data. Statistical analyses considered categorical and continuous process measures. We computed the standardized effect size for each outcome at the 95% confidence interval. The primary outcome of interest was health professionals’ adoption of shared decision making as reported by patients in a self-administered questionnaire.

Results: Of the 6764 search results, 21 studies reported 35 relevant comparisons. Overall, the quality of the studies ranged from 0% to 83%. Only three of the 21 studies reported a clinically significant effect for the primary outcome that favored the intervention. The first study compared an educational meeting and a patient-mediated intervention with another patient-mediated intervention (median improvement of 74%). The second compared an educational meeting, a patient-mediated intervention, and audit and feedback with an educational meeting on an alternative topic (improvement of 227%). The third compared an educational meeting and a patient-mediated intervention with usual care (p = 0.003). All three studies were limited to the patient-physician dyad.

Limitations: To reduce bias, future studies should improve methods and reporting, and should analyze costs and benefits, including those associated with training of health professionals.

Conclusions: Multifaceted interventions that include educating health professionals about sharing decisions with patients and patient-mediated interventions, such as patient decision aids, appear promising for improving health professionals’ adoption of shared decision making in routine clinical practice as seen by patients.

Research shows that a significant number of patients prefer to play an active role in decisions concerning their health, especially after they understand the benefits of doing so.[1] Patients’ participation in decision making is associated with better health outcomes and with greater satisfaction with the consultation process.[2,3] Patient decision aids[4] are one strategy to encourage patients’ participation, and shared decision making is often considered to lie at the crux of patient-centered care.[5]

While there is no single definition of shared decision making, it can, in general, be considered the process in which a healthcare choice is made jointly by the health professional and the patient.[6,7] A systematic review that identified 31 shared decision-making components and summarized the key ones in an integrative model has posited that shared decision making occurs when the parties acknowledge that a decision is needed, are aware of and understand the best available evidence concerning the risks and the benefits of every option, and make a decision that takes not only the professional’s recommendations but also the patient’s values and preferences into account.[8] Shared decision making thus leverages the use of best evidence in clinical practice by capitalizing on the patient-professional consultation to involve both members of the decision-making dyad — the healthcare professional(s) and the patient — in the decision. In other words, shared decision making does not assume that the professional is the only party to need access to evidence in order for patients to experience evidence-based practice. Rather, it assumes that both the professional and the patient need access to the best evidence and need to discuss this evidence in light of the patient’s values and preferences and the practitioner’s recommendations. By practicing shared decision making, professionals can thus be expected to boost their use of evidence in their practice and to improve patients’ experience of the healthcare system.[9]

Shared decision making has earned the approval of policy makers for several reasons. Implementing shared decision making in clinical practice has the potential to reduce the overuse of options whose benefits to the majority are unclear.[10] It can also enhance the use of options shown to be beneficial,[11] reduce unwarranted variations in practices,[12] and make healthcare systems more sustainable.[13] Shared decision making is attractive to healthcare professionals because it helps them discuss evidence with patients, bridging the gap between patient-centered care and evidence-based practice — with some stakeholders perceiving the latter as anything but centered on the patient.[14]

This said, some authors present an alternative view of shared decision making.[15, 16, 17] They suggest that some patients may not wish to participate in decisions, or that patient satisfaction may only be low when patients’ preferences with regard to decision making do not match their actual decisional responsibility.[17] Some of these authors have found that if professionals do not explicitly ask about and respect a patient’s preference of role, they may or may not re-align their approach during the consultation so as to defer to the patient’s preference.[17] Others argue that shared decision making is being defined by a consumer-oriented decision-making model, giving policy makers cause to worry that more shared decisions will increase the demand for unnecessary, costly, or harmful procedures, and undermine the equitable allocation of healthcare resources.[16] More recently, Charles et al.[15] have highlighted other challenges relevant to the implementation of shared decision making in clinical practice: a lack of consistency in the definition and conceptualization of shared decision making; differences within and between patients’ and health professionals’ preferences for active patient involvement; and unrealistic expectations of the impact of shared decision making on population health. These challenges may help explain why health professionals have not adopted shared decision making on a widespread basis,[18] and why, in certain clinical settings, patients exhibit low involvement in consultations.[19, 20, 21]

To inform the debate on shared decision making, in 2010 we published a Cochrane systematic review[22] of interventions to increase health professionals’ adoption of shared decision making as observed by a third party. Although our findings were of interest to decision makers, they only disseminated information pertaining to the effect of interventions evaluated using an objective instrument. There is evidence that third-party observations of the decision-making process that occurs during consultations do not match patients’ experience and may not reflect what had occurred during the clinical encounter.[23] Moreover, Stewart et al.[24] have shown that patients’ perceptions of how much patient-centered care occurs during a consultation predict patient outcomes, but that third-observer perspectives do not. Consequently, this systematic review seeks to evaluate patients’ perceptions of the effectiveness of interventions to improve health professionals’ adoption of shared decision making in routine clinical practice.

1. Methods

1.1 Data Sources

We systematically searched electronic bibliographic databases from inception to mid-March 2009. Table I specifies the search strategy used for MEDLINE’s PubMed database. We modified this strategy to search the following databases: EMBASE, CINAHL, PsycINFO, and the Cochrane Library (namely, the Cochrane Database of Systematic Reviews; the Cochrane Central Register of Controlled Trials, which includes the Effective Practice and Organisation of Care [EPOC] Group specialized register; the Database of Abstracts of Reviews of Effectiveness; the NHS EED [Economic Evaluation Database]; and the Health Technology Assessment Database). We found additional material by consulting the reference lists of studies found in the databases; reviews about shared decision making; proceedings of the 2003, 2005, and 2007 editions of the International Shared Decision Making Conference;[25, 26, 27] and proceedings of the 2004, 2005, 2006, and 2007 annual meetings of the Society for Medical Decision Making.[28, 29, 30, 31]
Table I

Search strategy used for PubMed

1.2 Screening and Data Collection

Pairs of reviewers independently screened all titles and abstracts to determine eligibility for inclusion. Four reviewers (FL, SR, DS, and ST) screened full copies of potentially eligible studies. Two reviewers (SR and ST) used a modified version of the EPOC data collection checklist[32] to independently extract the primary outcome of interest and the characteristics of the settings and the interventions. We also extracted information about the duration of the consultation, a secondary outcome of interest. When necessary, we contacted the primary authors for clarification. Disagreements were resolved in discussions with FL and, when necessary, the entire team.

1.3 Selection Criteria

1.3.1 Study Design

We included randomized controlled trials (RCTs) or well designed quasi-experimental studies, controlled clinical trials, controlled before-and-after studies, and interrupted time series analyses. For the latter, we required a well defined timeframe and at least three data points before and after the intervention.

1.3.2 Participants

We defined health professionals as providers who were (i) either licensed to practice or in training, and (ii) responsible for patient care.

1.3.3 Interventions

We targeted interventions designed to increase health professionals’ adoption of shared decision making. Using the EPOC taxonomy of interventions,[32] we included interventions targeting health professionals (e.g. printed educational material, educational meetings, audit and feedback, reminders) and patient-mediated interventions. We defined the latter as interventions aimed at changing health professionals’ behavior either through provider-patient interactions or through information provided by or to the patient (e.g. a patient decision aid to help prepare patients for participating in decision making).[33]

1.3.4 Outcome Measures

Our primary outcome of interest was health professionals’ adoption of shared decision making, assessed by a self-administered patient questionnaire. For example, we considered agreement with the following response to indicate that a health professional had adopted shared decision making: ‘My doctor and I made the decision together.’[34]

1.4 Quality Assessment

Two reviewers (SR and ST) independently assessed the quality of the studies using a modified version of the quality criteria described in the EPOC data collection checklist.[32] One important difference concerned the baseline measurement; should it not be reported, the risk of bias was reported as ‘no’ and not ‘unclear’. They assessed seven criteria: allocation concealment; follow-up with professionals; follow-up with patients; blinded assessment of the participant-reported outcome; baseline data for the participant-reported outcome; a reliable primary outcome; and protection from contamination. Each criterion was assessed as ‘yes’ or ‘no.’ Quality criteria that did not apply were excluded from the score of bias.

1.5 Data Analysis

For all studies, we reported results for both categorical and continuous primary outcomes. For categorical measures, we calculated the absolute difference in the main outcome of interest between the intervention and the control arms. We calculated standard effect sizes for continuous measures by dividing the difference in the mean scores of the intervention group and comparison group of each study by the pooled estimated standard deviation for the two groups (the intervention group and the comparison group). A standard effect size of <0.2 was considered to be neither clinically or statistically significant; an effect size of 0.2 to <0.5 was potentially statistically significant, but was not considered clinically significant; an effect size of 0.5 to <0.8 was considered to be statistically significant and moderately clinically significant; and lastly, an effect size of ≥0.8 was considered to be both statistically and clinically significant. For categorical and continuous outcomes, we constructed 95% confidence intervals (CIs) to compare groups before and after the intervention. In the case of continuous data, we constructed CIs using the interval inversion approach.[35] The absence of a zero value within a CI indicated that the baselines differed or that the intervention had a statistically significant effect compared with the control intervention or with usual care. When no baseline was reported, we considered groups to be similar prior to the intervention. For studies for which quantitative data were missing or insufficient to calculate effect size, we contacted the authors of the studies to obtain the information. Barring the authors’ response, we reported qualitative data as presented by the authors in the article. We did not conduct a meta-analysis. For ease of interpretation, we produced forest plots that depict the results for categorical and continuous measures. We used Cohen’s d method to calculate CIs in the forest plots. Consequently, small variations can be observed between the lower and upper limits of the CI for the forest plots and the limits of the interval, for which we used the interval inversion approach.

1.5.1 Unit-of-Analysis Issues

Comparisons that randomize or allocate clusters (health professionals or organizations) but do not account for clustering during analysis have potential unit-of-analysis errors that can produce artificially significant p-values and overly narrow CIs.[36] When authors provided information that was previously missing, we attempted to re-analyze studies with potential unit-of-analysis errors. When the missing information was unavailable, we did not establish CIs.

1.6 Assessment of Heterogeneity

To explore heterogeneity, we prepared a table that compared the standard effect sizes of the studies and presented absolute differences in the main outcome of interest. We considered the following variables: type of intervention; characteristics of the intervention; clinical setting; type of health professionals involved; and health professionals’ level of training. Because of the broad nature of our primary outcome (ten measures), we did not perform a meta-regression.

2. Results

2.1 Characteristics of the Studies and the Interventions

Our search strategy yielded 6764 potentially relevant references, 21 of which were eligible for inclusion (figure 1). These 21 studies made 35 comparisons. All 21 were RCTs and all were published in Europe, North America, or Australia (table II). Fifteen studies (71.4%) were published after 2003,[37, 38, 39, 40,42, 43, 44, 45, 46, 47, 48, 49,52,54,55] and the rest were published between 1995 and 2003.[34,41,50,51,53,56] Ten studies were conducted in primary care clinical settings,[40, 41, 42,46, 47, 48, 49, 50, 51,55] and the others were conducted in specialized care.[34,37, 38, 39,43, 44, 45,52, 53, 54,56] Only two studies were conducted in a non-ambulatory care setting.[43,54] None of the studies involved simulated patients, making this review’s reporting on outcomes for routine clinical practice even more relevant.
Fig. 1

Flow diagram of the search results.

Table II

Characteristics of the studies and the interventions

Twenty studies (95.2%) targeted physicians, and one study targeted other types of health professionals.[48] Only 13 RCTs (61.9%) reported the number of participating professionals (this number ranged from 2 to 91).[34,37, 38, 39,41,42,46,47,49,52,53,55,56] In contrast, all studies reported the number of participating patients (between 26 and 10 070). The most commonly observed patient profile was patients with a cancer-related clinical condition (eight studies).[34,38,39,41,46,53,54,56] Only 12 studies reported exclusively on patient-mediated interventions.[34,38,39,41,44, 45, 46,50,52, 53, 54,56] Table II describes the studies in more detail.

2.2 Characteristics of the Outcome Measures

The studies used ten measures to assess the primary outcome of interest. The most commonly used instruments were a variant of the ‘Control Preference Scale’ evaluating the perceived level of control in decision making (used in ten studies)[34,38, 39, 40, 41,44,46,50,51,54] and COMRADE (combined outcome measure for risk communication and treatment decision-making effectiveness) [used in three studies].[42,43,55]

2.3 Quality Assessment

Table II gives our assessment of the quality of the 21 studies. Overall, we assessed quality as ranging from 0% (the study met none of the six quality criteria that applied)[40] to 83.3% (the study met five of the six quality criteria that applied).[42] More specifically, the quality assessment criteria for the evaluation of bias scored ≥50% for six of the 15 patient RCTs (range 50–80%).[41,44, 45, 46,50,56] For cluster RCTs, four of the six studies scored over 50% (range 60–83.3%).[42,47,51,55]

2.4 Effects of Interventions

2.4.1 Single Interventions Compared with Usual Care

Of the 21 studies, three[46,50,54] compared a single intervention with usual care. These studies reported a total of one continuous measure[54] and four categorical measures[46,50,54] of health professionals’ adoption of shared decision making. All three studies used a patient-mediated intervention, such as a patient decision aid. The study[54] that used a continuous measure was not considered in the forest plot of these outcomes (see figure 2), as data were incomplete.
Fig. 2

Forest plot for continuous measures. Some identified studies were not considered in the forest plot because a confidence interval (CI) could be not be constructed due to the provision of either incomplete data (Vodermaier et al.,[54] Loh et al.,[49] Wetzels et al,[55] and Krones et al.[47]) or only qualitative data (Street et al.[53] and Lalonde et al.[48]). COMRADE = combined outcome measure for risk communication and treatment decision-making effectiveness.

In the three studies that used categorical measures (see figure 3),[46,50,54] two studies found no statistically or clinically significant difference between intervention groups.[46,54] One study did find a statistically significant effect favoring the control, but it was not clinically significant.[50] The non-clinically relevant study was the only intervention which used an interactive videodisc as a decision aid.
Fig. 3

Forest plot for categorical measures. One study (Brown et al.[38]) was not considered in the forest plot because it reported qualitative, not quantitative, data.

2.4.2 Single Interventions Compared with Other Single Interventions

Ten studies compared a single intervention with another single intervention[34,38,39,41,44, 45, 46,52,53,56] These studies compared different patient-mediated interventions and reported on one continuous measures[53] and ten categorical measures[34,38,39,41,44, 45, 46,52,53,56] involving the adoption of shared decision making. Two of these studies were not considered in the relevant forest plots (figures 2 and 3), as data were qualitative.[38,53] None of these instruments recorded a significant effect.[34,38,39,41,44, 45, 46,52,53,56]

2.4.3 Multifaceted Interventions Compared with Usual Care

Four RCTs compared a multifaceted intervention with usual care.[43,49,51,55] Together, these RCTs reported three continuous measures[43,49,55] and one categorical measures[51] of health professionals’ adoption of shared decision making. Two studies[49,55] that had used a continuous measure were not considered in the forest plot, as data were either incomplete[49] or qualitative.[55] The results of the other two studies.[43,51] are shown in figures 2 and 3.

Of the four RCTs, only one found a significant effect.[49] This study reported that, compared with usual care, an educational meeting for physicians and a patient-mediated intervention improved patients’ perception of shared decision making (p = 0.003). It was not possible to calculate a standardized effect size with a CI for this study because the raw data were unavailable.

2.4.4 Multifaceted Interventions Compared with Single Interventions

Three RCTs compared a multifaceted intervention with a single intervention,[37,40,47] and reported a total of two continuous[37,47] and one categorical measure[40] of the adoption of shared decision making. One study’s outcome[47] that used a continuous measure was not considered in the forest plot, as data were incomplete. The results of the other two studies[37,40] are shown in figures 2 and 3.

Two of the three multifaceted RCTs produced a significant effect.[37,47] The first study compared a two-faceted intervention (a patient-mediated intervention and 18 hours of physician-oriented educational meetings) with a patient-mediated intervention.[37] This study reported three continuous measures for health professionals’ adoption of shared decision making, and showed improvements in the physician-patient interaction of 74.3% (95% CI 25, 124), 50.6% (95% CI 2, 99), and 89.0% (95% CI 38,139) at first consultation, 3 months, and 12 months, respectively.[37] The second study compared a three-faceted intervention (a patient-mediated intervention, two educational meetings of 2 hours each with physicians, and audit and feedback) with a physician-oriented educational meeting.[47] This study reported one continuous measure of the adoption of shared decision making (the other measure could not be evaluated) and measured an improvement of 611.0% (95% CI 582, 640) on the patient participation and satisfaction scale.[47] The precision of the lower and upper limits of the CI for the standardized difference between means reported here differs slightly from that in figure 2, because we calculated the CI with the interval inversion method for the results section and not with the Cohen’s d method used for the forest plot. The third study, a neutral study (i.e. one that did not record an impact), did not involve training health professionals.[40]

2.4.5 Multifaceted Interventions Compared with Other Multifaceted Interventions

Two RCTs compared two multifaceted interventions,[42,48] and reported nine continuous measures of health professionals’ adoption of shared decision making. The study[48] that used a continuous measure was not considered in the forest plot, as data were incomplete; the results of the eight continuous measures in the other study[42] are shown in figure 2. Neither study found a statistically significant difference between the professionals’ adoption of shared decision making

2.5 The Duration of the Consultation and Potential Effect Modifiers

The duration of the patient-practitioner consultation was reported in seven studies,[38,46,49,50,54, 55, 56] none of which found a statistically significant change in the duration of the consultation after the intervention. Additionally, no studies showed heterogeneity for the following variables: the type of intervention; the type of health professional; the clinical setting; and whether the health professional was licensed or was in training.

3. Discussion

This systematic review identified 21 unique studies that evaluated the effectiveness of interventions to improve the adoption of shared decision making in clinical practice from a patient’s perspective. Of these, 12 compared a single intervention with usual care[46,50,54] and/or another single intervention.[34,38,39,41,44, 45, 46,52,53,56] All 12 assessed the impact of a lone patient-mediated intervention, such as a patient decision aid, and none showed a significant impact. The nine remaining studies[37,40,42,43,47, 48, 49,51,55] compared multifaceted interventions with usual care,[43,49,51,55] a single intervention[37,40,47] or another multifaceted intervention.[42,48] Of these studies, six included training health professionals as an intervention.[37,42,43,47,49,51] These six studies had different conceptions of training health professionals. Two studies trained professionals who were responsible for using the patient-mediated intervention with patients (all were patient decisions aids) but were not responsible for making the decision.[43,51] Neither study showed a significant impact. The four remaining studies tested training health professionals who were responsible for sharing the decision with the patient.[37,42,47,49] Of these, three combined this training with a patient-mediated intervention (i.e. a patient decision aid),[37,47,49] and one combined the training with audit and feedback.[42] Interestingly, only the three studies that combined training health professionals who were sharing the decision with the patient with the use of a patient-mediated intervention reported a positive impact on the outcome of interest.[37,47,49] In other words, of the 21 studies included in our review, only three combined a patient-mediated intervention (in all cases, patient decision aids) with training health professionals who were responsible for sharing the decision with the patient.[37,47,49] This unique finding suggests that combining interventions that target health professionals responsible for sharing the decision with the patient (in this case, educational meetings) with patient-mediated interventions (in this case, patient decision aids) is a promising means of translating shared decision making into routine clinical practice.

Of course, since shared decision making is a mutual process, it makes sense to intervene at the level of both the health professional and the patient; in other words, it is logical that both members of the dyad must be activated for shared decision making to occur. From a theoretical perspective, however, it is interesting that this deduction is congruent with a recently published study that showed that physicians’ attitudes influenced both their patients’ and their own willingness to engage in shared decision making.[57]

For the three positive studies that included training health professionals as an intervention, the duration of the consultation was not observed to vary according to professionals’ aptitudes in shared decision making, as perceived by patients.[37,47,49] This finding is important given health professionals’ concern that shared decision making increases the length of the consultation.[58] Of course, the length of the consultation does not include the time needed to prepare health professionals and patients to share decisions in routine clinical practice. For example, one of the three positive studies reported spending 18 hours training health professionals.[37] We agree with Edwards and Elwyn[17] that further exploration with health professionals is needed to design efficient, as well as effective, training programs for translating shared decision making into routine clinical practice.

These results complement the results of other reviews. Between our previous Cochrane review[22] and the present work, we found five studies in which interventions to increase shared decision making in clinical practice had a significant effect.[37,47,49,59,60] In one[60] of the two positive studies[59,60] identified in our previous review,[22] the authors compared usual care with a multifaceted intervention consisting of a coaching protocol, health professional training in shared decision making, and performance feedback. This study used simulated patients to assess health professionals’ behaviors. The second study[59] in our previous review compared a pamphlet with a patient decision aid to be used during the clinical consultation, and found that the patient decision aid increased shared decision making during the consultation. In both studies, only third-party observers assessed health professionals’ shared decision-making behavior. The use of a third-observer measurement could explain the second study’s finding that the patient decision aid was effective in the absence of health professional training,[59] albeit our present conclusions suggest that training health professionals should be half of a two-pronged approach. In other words, the difference in measures (a third-party instrument vs a patient-administered instrument) leads to a difference in the construct assessed and may explain the difference in results. We remain cautious when comparing the results of studies conducted in routine clinical practice settings with the results of studies using simulated patients (such as the first study[60] of our previous review[22]), since it is unclear how interpersonal perceptions and the bidirectional influence between the two members of the dyad — health professional and patient — differ in simulated versus routine clinical settings.[23]

Meanwhile, the two studies that overlap our previous Cochrane review and the present review are consistent.[39,42] In neither did third-party observers or patients find that interventions to promote health professionals’ practice of shared decision making had an effect. Supporting our conclusions here is that neither study combined training health professionals in shared decision making with the use of a patient-mediated intervention.

Our review shares eight studies[34,40,41,48,50,51,53,56] with a Cochrane review of the impact of patient decision aids.[3] In contrast with the Cochrane review,[3] our current review showed none of the common studies as having a clinically significant effect on our primary outcome of interest: health professionals’ adoption of shared decision making, assessed with a self-administered patient questionnaire. This difference can be explained by the fact that our primary outcome of interest (whether the patient considered that a joint decision had been made) differed from the primary outcome of interest of the Cochrane review (whether the decision was not made by the practitioner).[34] Two more systematic reviews on shared decision-making interventions consist of narrative summaries of the studies and do not re-analyze or evaluate the material as we have done.[61,62]

It should be noted that one statistically significant study, identified in the patient decision aid review,[3] favored the control.[50] This study closely resembles the characteristics of the other studies in the review, although, the decision aid was slightly different: it used an interactive videodisk, consulted on site. Furthermore, according to the paper itself, the intervention groups were more likely to independently decide their treatment option than the control group.[50] Consequently, it is possible that the decision aid was enabling independent patients rather than shared decision making. More importantly, based on our reference framework, this study was identified as non-clinically significant. Consequently, it is not among the studies which allowed us to assess potentially effective intervention strategies.

Our systematic review reveals that the study of implementing shared decision making in routine clinical practice is a recently established field of inquiry, with more than half of the 21 studies included in our review published after 2003.[37, 38, 39, 40,42, 43, 44, 45, 46, 47, 48, 49,52,54,55] It also reveals that the evidence base comes from diverse clinical settings: this underscores the relevance of shared decision making to a spectrum of clinical specialties and health conditions. This is important because debate about the relevance of shared decision making to certain clinical settings is ongoing.[63] For example, it is not clear whether shared decision making should be implemented in primary care, where several decisions may be made during a single consultation[64] and where patients often present with multiple diseases.[65]

Most studies covered by our review focused on the implementation of shared decision making in physicians’ clinical practice, not in the clinical practice of other kinds of health professionals. This seriously limits our understanding of the implementation of shared decision making in nursing, rehabilitation, dietary medicine, and other clinical settings. It also limits our understanding of the implementation of shared decision making in clinical practice in general, not to mention its implementation within the interprofessional, team-based approach to patient care towards which many healthcare systems are moving.[66]

Finally, none of the 21 studies evaluated the costs and benefits of their interventions, despite the evident need of cost-benefit information to evaluate the efficacy of the interventions. This is especially important given the length and the intensity of some shared decision-making training programs. It may not be realistic to expect large groups of health professionals to attend many hours of training, even if the training is found to be effective. For this reason, we consider that future studies of the implementation of shared decision making should take into account the broader nature of clinical work in healthcare systems, and assess the costs and benefits of implementation interventions.

Notwithstanding its interesting results, this systematic review has several limitations. First, despite our thorough search strategy, it is possible that we missed eligible studies. Second, the studies retained for this review used ten unique instruments (scales or subscales) to assess health professionals’ adoption of shared decision making as perceived by patients. This made it difficult to combine data and draw conclusions. Third, most studies provided little information on the context for changes in behavior. Also, we found no relationship between the quality assessment of the studies and the significance of their results. For these reasons, we advise that study designs be improved methodologically, and that reports on behavioral changes in clinical practice be more extensive.[67] Finally, when we computed effect size, we could not account and control for co-variates. Therefore, our interpretation of the effectiveness of interventions may differ from that of the original authors.

4. Conclusions

Our study provides evidence that, judging from patients’ perspectives of the physician-patient consultation, interventions that target both the health professional responsible for sharing a decision with the patient (one or more health professional educational meetings) and the patient himself/herself (a patient-mediated intervention, such as a patient decision aid) appear promising in translating shared decision making into routine clinical practice. That said, because of the small number of studies in our review, our results do not allow us to draw absolute conclusions on the most effective types of intervention for promoting health professionals’ adoption of shared decision making. Our results do, however, point to the importance of targeting the health professionals most likely to benefit from a shared decision-making training program: namely, those professionals who will share the decision with the patient. This distinction has not yet been made in the literature or in training programs. In addition, our results indicate the desirability of interventions that engage both members of the dyad. They do not, however, acknowledge more interprofessional involvement in sharing decisions, and our results lack information to help decision makers weigh the costs and benefits of implementation of interventions, including interventions that require significant hours of health professional training. For this reason, future studies of the implementation of shared decision making should use consistent measures and stronger methods, better report on the elements of interventions, and analyze the costs and benefits of the overall process of care, including the costs associated with training health professionals.

Notes

Acknowledgments

France Légaré, Stéphane Turcotte and Dawn Stacey conceived this study. All the authors helped refine the study design and analyze and interpret the data. France Légaré drafted the initial manuscript, and the other authors suggested critical revisions. All authors approved the final manuscript for publication. Jennifer Petrela edited the manuscript. France Légaré received salary support from the Government of Canada’s Research Chairs Program. The views expressed are those of the authors and not of the funding agency. The authors declare that they have no conflicts of interest.

References

  1. 1.
    Kiesler DJ, Auerbach SM. Optimal matches of patient preferences for information, decision-making and interpersonal behavior: evidence, models and interventions. Patient Educ Couns 2006; 61(3): 319–41PubMedCrossRefGoogle Scholar
  2. 2.
    Hack TF, Degner LF, Watson P, et al. Do patients benefit from participating in medical decision making? Longitudinal follow-up of women with breast cancer. Psychooncology 2006; 15(1): 9–19PubMedCrossRefGoogle Scholar
  3. 3.
    O’Connor AM, Bennett CL, Stacey D, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev 2009; 3: CD001431PubMedGoogle Scholar
  4. 4.
    Coulter A, Ellins J. Effectiveness of strategies for informing, educating, and involving patients. BMJ 2007; 335(7609): 24–7PubMedCrossRefGoogle Scholar
  5. 5.
    Weston WW. Informed and shared decision-making: the crux of patient-centered care. CMAJ 2001; 165(4): 438–9PubMedGoogle Scholar
  6. 6.
    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): 681–92PubMedCrossRefGoogle Scholar
  7. 7.
    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): 651–61PubMedCrossRefGoogle Scholar
  8. 8.
    Makoul G, Clayman ML. An integrative model of shared decision making in medical encounters. Patient Educ Couns 2006; 60(3): 301–12PubMedCrossRefGoogle Scholar
  9. 9.
    Brouwers M, Stacey D, O’Connor A. Knowledge creation: synthesis, tools and products. CMAJ 2010; 182(2): E68–72PubMedGoogle Scholar
  10. 10.
    Evans R, Edwards A, Brett J, et al. Reduction in uptake of PSA tests following decision aids: systematic review of current aids and their evaluations. Patient Educ Couns 2005; 58(1): 13–26PubMedCrossRefGoogle Scholar
  11. 11.
    O’Connor AM, Bennett C, Stacey D, et al. Do patient decision aids meet effectiveness criteria of the international patient decision aid standards collaboration? A systematic review and meta-analysis. Med Decis Making 2007; 27(5): 554–74PubMedCrossRefGoogle Scholar
  12. 12.
    Wennberg JE. Practice variation: implications for our health care system. Manag Care 2004; 13(9 Suppl.): 3–7PubMedGoogle Scholar
  13. 13.
    Coulter A. Engaging patients in their healthcare: how is the UK doing relative to other countries. Oxford: Picker Institute Europe, 2006Google Scholar
  14. 14.
    Bensing J. Bridging the gap: the separate worlds of evidence-based medicine and patient-centered medicine. Patient Educ Couns 2000; 39: 17–25PubMedCrossRefGoogle Scholar
  15. 15.
    Charles C, Gafni A, Freeman E. Implementing shared decision making and treatment decision aids: a cautionary tale. Psicooncologica 2010; 7(2–3): 243–55Google Scholar
  16. 16.
    Coulter A. Partnerships with patients: the pros and cons of shared clinical decision-making. J Health Serv Res Policy 1997; 2(2): 112–21PubMedGoogle Scholar
  17. 17.
    Edwards A, Elwyn G. Inside the black box of shared decision making: distinguishing between the process of involvement and who makes the decision. Health Expect 2006; 9(4): 307–20PubMedCrossRefGoogle Scholar
  18. 18.
    Guimond P, Bunn H, Edwards A, et al. Validation of a tool to assess health practitioners’ decision support and communication skills. Patient Educ Couns 2003; 50(3): 235–45PubMedCrossRefGoogle Scholar
  19. 19.
    Elwyn G, Hutchings H, Edwards A, et al. The OPTION scale: measuring the extent that clinicians involve patients in decision-making tasks. Health Expect 2005; 8(1): 34–42PubMedCrossRefGoogle Scholar
  20. 20.
    Gagnon S, Labreque M, Njoya M, et al. How much do family physicians involve pregnant women in decisions about prenatal screening for Down syndrome? Prenat Diagn 2010; 30(2): 115–21PubMedCrossRefGoogle Scholar
  21. 21.
    Siriwardena AN, Edwards AG, Campion P, et al. Involve the patient and pass the MRCGP: investigating shared decision making in a consulting skills examination using a validated instrument. Br J Gen Pract 2006; 56(532): 857–62PubMedGoogle Scholar
  22. 22.
    Légaré F, Ratté S, Stacey D, et al. Interventions for improving the adoption of shared decision making by healthcare professionals. Cochrane Database Syst Rev 2010; 5: CD006732PubMedGoogle Scholar
  23. 23.
    Saba GW, Wong ST, Schillinger D, et al. Shared decision making and the experience of partnership in primary care. Ann Fam Med 2006; 4(1): 54–62PubMedCrossRefGoogle Scholar
  24. 24.
    Stewart M, Brown JB, Donner A, et al. The impact of patient-centered care on outcomes. J Fam Pract 2000; 49(9): 796–804PubMedGoogle Scholar
  25. 25.
    Elwyn G, Edwards A. Proceedings of the 2nd International Shared Decision Making Conference; 2003 Sep 2–4; Swansea [online]. Available from URL: http://204.187.39.28/isdm2005/SDM_Conference_2003_Book.pdf [Accessed 2011 Oct 31]Google Scholar
  26. 26.
    O’Connor A, Graham I. Implementing shared decision making in diverse health care systems and cultures: programs and abstracts. Proceedings of the 3rd International Shared Decision Making Conference; 2005 Jun 14–16; Ottawa (ON) [online]. Available from URL: http://decisionaid.ohri.ca:8080/ISDM2005/program.html [Accessed 2011 Oct 31]Google Scholar
  27. 27.
    Härter M, Simon D, Loh A. Shared decision-making in diverse health care systems: translating research into practice. Proceedings of the 4th International Shared Decision Making Conference; 2007 May 30–Jun 1; Freiburg [online]. Available from URL: http://2010.patient-als-partner.de/files/isdm_2007_-_conference_program.pdf [Accessed 2011 Oct 31]Google Scholar
  28. 28.
    Society of Medical Decision Making. Proceedings of the 26th Annual Meeting for the Society for Medical Decision Making; 2004 Oct 17–20; Atlanta (GA) [online]. Available from URL: http://smdm.confex.com/smdm/2004ga/techprogram/MEETING.HTM [Accessed 2011 Oct 31]Google Scholar
  29. 29.
    Society of Medical Decision Making. Proceedings of the 27th Annual Meeting for the Society for Medical Decision Making; 2005 Oct 21–24; San Francisco (CA) [online]. Available from URL: http://smdm.confex.com/smdm/2005ca/techprogram/MEETING.HTM [Accessed 2011 Oct 31]Google Scholar
  30. 30.
    Society of Medical Decision Making. Proceedings of the 28th Annual Meeting for the Society for Medical Decision Making; 2006 Oct 15–18; Boston (MA) [online]. Available from URL: http://smdm.confex.com/smdm/2006ma/techprogram/MEETING.HTM [Accessed 2011 Oct 31]Google Scholar
  31. 31.
    Society of Medical Decision Making. Proceedings of the 29th Annual Meeting for the Society for Medical Decision Making; 2007 Oct 20–24; Pittsburgh (PA) [online]. Available from URL: http://smdm.confex.com/smdm/2007pa/techprogram/MEETING.HTM [Accessed 2011 Oct 21]Google Scholar
  32. 32.
    Grimshaw J. Effective practice and organisation of care group. Ottawa (ON): University of Ottawa, 2002 [online]. Available from URL: http://epoc.cochrane.org/ [Accessed 2011 Oct 23]Google Scholar
  33. 33.
    Entwistle VA, Watt I, Rosser J, et al. Patient mediated interventions: development of materials and early findings. 11th Annual Meeting of the International Society of Technology Assessment in Health Care; 1995 Jun 4–7; StockholmGoogle Scholar
  34. 34.
    Davison BJ, Degner LF. Empowerment of men newly diagnosed with prostate cancer. Cancer Nurs 1997; 20(3): 187–96PubMedCrossRefGoogle Scholar
  35. 35.
    Hess MR. Effect sizes, significance tests, and confidence intervals: assessing the influence and impact of research reporting protocol and practice [dissertation]. Tampa (FL): University of South Florida, 2003Google Scholar
  36. 36.
    Ukoumunne OC, Gulliford MC, Chinn S, et al. Methods for evaluating area-wide and organisation-based interventions in health and health care: a systematic review. Health Technol Assess 1999; 3(5): iii–92PubMedGoogle Scholar
  37. 37.
    Bieber C, Muller KG, Blumenstiel K. Long-term effects of a shared decision-making intervention on physician-patient interaction and outcome in fibromyalgia: a qualitative and quantitative 1 year follow-up of a randomized controlled trial. Patient Educ Couns 2006; 63(3): 357–66PubMedCrossRefGoogle Scholar
  38. 38.
    Brown RF, Butow PN, Sharrock MA, et al. Education and role modelling for clinical decisions with female cancer patients. Health Expect 2004; 7(4): 303–16PubMedCrossRefGoogle Scholar
  39. 39.
    Butow P, Devine R, Boyer M, et al. Cancer consultation preparation package: changing patients but not physicians is not enough. J Clin Oncol 2004; 22(21): 4401–9PubMedCrossRefGoogle Scholar
  40. 40.
    Deschamps MA, Taylor MG, Neubauer SL, et al. Impact of pharmacist consultation versus a decision aid on decision making regarding hormone replacement therapy. Int J Phar Pract 2004; 12(1): 21–8CrossRefGoogle Scholar
  41. 41.
    Dolan JG, Frisina S. Randomized controlled trial of a patient decision aid for colorectal cancer screening. Med Decis Making 2002; 22(2): 125–39PubMedGoogle Scholar
  42. 42.
    Elwyn G, Edwards A, Hood K, et al. Achieving involvement: process outcomes from a cluster randomized trial of shared decision making skill development and use of risk communication aids in general practice. Fam Pract 2004; 21(4): 337–46PubMedCrossRefGoogle Scholar
  43. 43.
    Hamann J, Langer B, Winkler V, et al. Shared decision making for in-patients with schizophrenia. Acta Psychiatr Scand 2006; 114(4): 265–73PubMedCrossRefGoogle Scholar
  44. 44.
    Kasper J, Köpke S, Muhlhauser I, et al. Informed shared decision making about immunotherapy for patients with multiple sclerosis (ISDIMS): a randomized controlled trial. Eur J Neurol 2008; 15(12): 1345–52PubMedCrossRefGoogle Scholar
  45. 45.
    Köpke S, Kasper J, Muhlhauser I, et al. Patient education program to enhance decision autonomy in multiple sclerosis relapse management: a randomized-controlled trial. Mult Scler 2009; 15(1): 96–104PubMedCrossRefGoogle Scholar
  46. 46.
    Krist AH, Woolf SH, Johnson RE, et al. Patient education on prostate cancer screening and involvement in decision making. Ann Fam Med 2007; 5(2): 112–9PubMedCrossRefGoogle Scholar
  47. 47.
    Krones T, Keller H, Sonnichsen A, et al. Absolute cardiovascular disease risk and shared decision making in primary care: a randomized controlled trial. Ann Fam Med 2008; 6(3): 218–27PubMedCrossRefGoogle Scholar
  48. 48.
    Lalonde L, O’Connor AM, Duguay P, et al. Evaluation of a decision aid and a personal risk profile in community pharmacy for patients considering options to improve cardiovascular health: the OPTIONS pilot study. Int J Pharm Pract 2006; 14(1): 51–62CrossRefGoogle Scholar
  49. 49.
    Loh A, Simon D, Willis CE, et al. The effects of a shared decision-making intervention in primary care of depression: a cluster-randomized controlled trial. Patient Educ Couns 2007; 67(3): 324–32PubMedCrossRefGoogle Scholar
  50. 50.
    Murray E, Davis H, Tai SS, et al. Randomised controlled trial of an interactive multimedia decision aid on benign prostatic hypertrophy in primary care. BMJ 2001; 323(7311): 493–6PubMedCrossRefGoogle Scholar
  51. 51.
    O’Cathain A, Walters SJ, Nicholl JP, et al. Use of evidence based leaflets to promote informed choice in maternity care: randomised controlled trial in everyday practice. BMJ 2002; 324(7338): 643PubMedCrossRefGoogle Scholar
  52. 52.
    Stiggelbout AM, Molewijk AC, Otten W, et al. The impact of individualized evidence-based decision support on aneurysm patients’ decision making, ideals of autonomy, and quality of life. Med Decis Making 2008; 28(5): 751–62PubMedCrossRefGoogle Scholar
  53. 53.
    Street Jr RL, Voigt B, Geyer Jr C, et al. Increasing patient involvement in choosing treatment for early breast cancer. Cancer 1995; 76(11): 2275–85PubMedCrossRefGoogle Scholar
  54. 54.
    Vodermaier A, Caspari C, Koehm J, et al. Contextual factors in shared decision making: a randomised controlled trial in women with a strong suspicion of breast cancer. Br J Cancer 2009; 100(4): 590–7PubMedCrossRefGoogle Scholar
  55. 55.
    Wetzels R, Wensing M, van Weel C, et al. A consultation leaflet to improve an older patient’s involvement in general practice care: a randomized trial. Health Expect 2005; 8(4): 286–94PubMedCrossRefGoogle Scholar
  56. 56.
    Whelan T, Sawka C, Levine M, et al. Helping patients make informed choices: a randomized trial of a decision aid for adjuvant chemotherapy in lymph node-negative breast cancer. J Natl Cancer Inst 2003; 95(8): 581–7PubMedCrossRefGoogle Scholar
  57. 57.
    Légaré F, St-Jacques S, Gagnon S, et al. Prenatal screening for Down syndrome: a survey of willingness in women and family physicians to engage in shared decision-making. Prenat Diagn 2011; 31(4): 319–26PubMedCrossRefGoogle Scholar
  58. 58.
    Légaré F, Ratte S, Gravel K, et al. Barriers and facilitators to implementing shared decision-making in clinical practice: update of a systematic review of health professionals’ perceptions. Patient Educ Couns 2008; 73(3): 526–35PubMedCrossRefGoogle Scholar
  59. 59.
    Nannenga MR, Montory VM, Weymiller AJ, et al. A treatment decision aid may increase patient trust in the diabetes specialist: the Statin Choice randomized trial. Health Expect 2009; 12(1): 38–44PubMedCrossRefGoogle Scholar
  60. 60.
    Stacey D, O’Connor AM, Graham ID, et al. Randomized controlled trial of the effectiveness of an intervention to implement evidence-based patient decision support in a nursing call centre. J Telemed Telecare 2006; 12(8): 410–5PubMedCrossRefGoogle Scholar
  61. 61.
    Duncan E, Best C, Hagen S. Shared decision making interventions for people with mental health conditions. Cochrane Database Syst Rev 2010 Jan; 1: CD007297PubMedGoogle Scholar
  62. 62.
    Joosten EA, DeFuentes-Merillas L, de Weert GH, et al. Systematic review of the effects of shared decision-making on patient satisfaction, treatment adherence and health status. Psychother Psychosom 2008; 77(4): 219–26PubMedCrossRefGoogle Scholar
  63. 63.
    Braddock 3rd CH. The emerging importance and relevance of shared decision making to clinical practice. Med Decis Making 2010; 30(5 Suppl.): 5–7SCrossRefGoogle Scholar
  64. 64.
    Braddock 3rd CH, Edwards KA, Hasenberg NM, et al. Informed decision making in outpatient practice: time to get back to basics. JAMA 1999; 282(24): 2313–20PubMedCrossRefGoogle Scholar
  65. 65.
    Fortin M, Bravo G, Hudon C, et al. Prevalence of multimorbidity among adults seen in family practice. Ann Fam Med 2005; 3(3): 223–8PubMedCrossRefGoogle Scholar
  66. 66.
    Légaré F, Stacey D, Gagnon S, et al. Validating a conceptual model for an inter-professional approach to shared decision making: a mixed methods study. J Eval Clin Pract 2011; 17(4): 554–64PubMedCrossRefGoogle Scholar
  67. 67.
    Michie S, Fixsen D, Grimshaw JM, et al. Specifying and reporting complex behaviour change interventions: the need for a scientific method. Implement Sci 2009; 4: 40PubMedCrossRefGoogle Scholar

Copyright information

© Adis Data Information BV 2012

Authors and Affiliations

  • France Légaré
    • 1
  • Stéphane Turcotte
    • 1
  • Dawn Stacey
    • 2
  • Stéphane Ratté
    • 1
  • Jennifer Kryworuchko
    • 3
  • Ian D. Graham
    • 4
  1. 1.St-François d’Assise HospitalQuebec University Hospital Research CentreQuébec CityCanada
  2. 2.School of Nursing, Faculty of Health SciencesUniversity of OttawaOttawaCanada
  3. 3.College of NursingUniversity of SaskatchewanSaskatoonCanada
  4. 4.Knowledge TranslationCanadian Institute of Health ResearchOttawaCanada

Personalised recommendations