Main findings
In this study, we aimed to report participant’s perspectives on using an SDM intervention to reduce decisional conflict of a preference-sensitive decision; the use of thromboprophylaxis during pregnancy. Our main finding is that all three components of the SDM intervention (evidence-based information, value elicitation exercises and decision analysis) reduce the overall decisional conflict. We found no decisional conflict (DCS<25) for the overall DCS score and the majority of the subscales, for all three groups; however, this trend was not statistically significant. We found a high level of decisional conflict (>37.5) for the uncertainty subscale due to the low certainty of the evidence, however value elicitation exercises showed to have the potential to reduce the uncertainty of the process.
Effect of the SDM intervention components on DCS
Evidence-based information reduced conflict in the decision-making process. However, when the information is based on low quality evidence, uncertainty increases. A similar study conducted in Spain(30), assessing a decision aid for breast cancer screening also noted the importance of providing evidence-based information to improve decision-making: women positively value receiving information regarding the benefits and harms of breast cancer screening. As in previous studies(31,32), we found that graphical representation of risks and benefits using pictograms showing the number of people experiencing an event with and without medication, reduces decisional conflict by clarifying the numerical information provided. Participants in our study also noted that patients’ health literacy should be assessed to ensure adequate understanding of the information. As shown by several authors who explored the relationship between health literacy and DCS, a better understanding of health information can significantly decrease decisional conflict(33).
Value elicitation exercises were useful to understand what is most important (risks or benefits) in a decision, thus supporting and facilitating the weighing activity (pros and cons) in the decision-making process. These exercises also reduced uncertainty in the decision-making process by helping participants better clarify (‘what choice is best for me’) their decision. This finding is consistent with IPDAS recommendations(11,15). Furthermore, exploring patients’ values and preferences contributes to patient engagement in the decision-making process, improving self-efficacy. Supportive of these findings, a recent cross-sectional study(33) assessing factors contributing to a lower decisional conflict found that respondents, who reported higher ability to actively engage and participate in the decision-making process, had lower decisional conflict.
The provision of a decision analysis recommendation decreased uncertainty (lowest level of decisional conflict for the uncertainty subscale) and improved self-efficacy with the decision process. It helped tip the balance of pros and cons, helping participants to be more confident with their decision. In addition, participants noted the need for health professionals when implementing the decision analysis technique in the clinical encounter, to support the cognitively-demanding activity of integrating the evidence with their preferences(34–36). On this regard, Dumont and colleagues(37) have referred to the use of decision analysis, as a decision support technique that promotes a meaningful dialogue between providers and patients on preferences, options, concerns, risks and benefits, leading to an informed and more satisfactory decision for both parties.
Decisional conflict scale as an instrument to assess the SDM process
In the context of SDM, decisional conflict is one of the most frequently reported outcomes in studies assessing decision support interventions(10,17,20,28), and the DCS appears to be an optimal instrument to measure the quality of the process(29). All the subscale items are in line with other instruments used to measure the quality of SDM interventions, such as the widely used SDM-Q-9, MAPPIN'SDM, and OPTION (16). However, a review assessing the quality of the SDM process highlighted that their common usage does not imply that these measures have adequate congruence with the conceptualization of SDM used to develop the intervention, as they do not necessarily capture the effect of the interactions among the decision-makers (i.e. patients, clinicians, family)(22). As seen in our study, the support subscale (how supported do patients feel in the decision-making process) needs further attention, especially the role of health professionals to support the process. For example, the CollaboRATE scale (38) further explores the support from clinicians in decision-making with items like ‘how much the provider listened to them about their health issue’. The need for health professionals as decisional partners was also highlighted by Legaré and colleagues(28) when developing a modified decisional conflict scale (D-DCS) with the aim of evaluating the decision-making process in SDM encounters, concluding that the patient-clinician relationship affects the quality of the decision. Furthermore, there is a need to understand the impact of peer pressure on decision-making. For example, in our decision context, some authors(8,39) have reported that the opinions and support from the husband of a pregnant woman going through this decision, as well as experiences from other women who went through this same condition may be important to support them.
The different DCS subscales have normally been compared in relation to usual care(9,10,20,40), less frequently when comparing SDM interventions(21,22), or for decision analysis as an intervention for SDM(13). In a study(40) evaluating the DCS for measuring the quality of end-of-life decisions, authors found significant differences in DCS scores between usual care (higher DCS scores) and the intervention (containing an evidence-based component and value elicitation exercises), and these were due to factors contributing to uncertainty and the efficacy of their decisions. They highlight some of the factors contributing to high uncertainty; feeling uninformed, feeling unclear about personal values, and feeling unsupported. Our study also showed that the subscale showing high conflict between groups was the uncertainty subscale (how clear and sure do patients feel about what to choose) and was attributed to the low certainty of the evidence and the support from others (specially clinicians) in the decision-making process. Despite this, value elicitation exercises did help clarify personal values. Other authors(13,34–36) have also reported on the contribution of decision analysis to support SDM and improve the uncertainty and effectiveness of the process; as Robinson and colleagues(35) explain: decision analysis was of value as it seeks to create a rational framework for evaluating complex medical decisions and to provide a systematic way of integrating potential outcomes with probabilistic information. However, our findings, as well as a scoping review on SDM containing decision analysis(13) highlighted the difficulties on how to implement decision analysis recommendations in clinical decision-making. Our results reveal that some of these challenges are related to how to present recommendations in the clinical encounter, and to deliver the information in a timely manner.
Limitations and strengths
Our sample consisted of students enrolled in a master program and, therefore, we cannot extrapolate our results to a more general population of women with a previous VTE event. This limitation was partly due to the COVID-pandemic, which hindered the recruitment of participants(41). Therefore, we conducted this study in parallel to a study our team was developing with the target population(12). However, our focus was to understand the quality of the decision-making process (i.e., how decisional conflict increased or decreased) with respect to each SDM intervention component. To this end, because our participants were simultaneously working as health professionals, they had helpful insights to understand the potential sources of conflict that may arise when implementing SDM interventions in a clinical context. In addition, we acknowledge the small sample size of our study as well as the different specialties of the health professionals included in our study and not having target clinicians such as gynecologists, obstetricians or hematologists for the decision assessed. Despite this limitation, we observed trends that were consistent with the qualitative findings.
Using a mixed method approach, and presenting the data in a joint manner, are some of the main strengths of our study. As other authors(24,25,42) have also reported, mixed-methods designs facilitate the understanding of complex phenomena and overcome the limitation that quantitative data have in understanding complex decision-making processes.
Implications for practice and research
We highlight four main implications of our study that should be addressed in future research and clinical practice:
First, high certainty of the evidence is needed to construct decision aids that aim to improve informed decision-making. This is especially important and challenging when there is equipoise regarding the efficacy of alternative treatments. Hence, more studies with larger sample sizes are needed to assess women’s values and preferences for the use of LMWH in pregnancy, thus providing high quality evidence to develop SDM interventions.
Second, we highlight the importance of including components that specifically explore patients’ values and preferences, such as value elicitation exercises, to reduce decisional conflict. Simple exercises exploring factors such as their previous experience with the condition or treatment, should be included in the development of SDM interventions.
Third, decision analysis has the potential to add value by reducing uncertainty and improving the efficacy and satisfaction with the SDM process. The cognitive reasoning activity of balancing pros and cons could be eased by an algorithm (decision analysis) that combines preferences with evidence. More implementation research is needed on how to deliver the decision analysis recommendation in clinical practice.
Fourth, it is essential to assess the interaction between patient and health professional, as well as include health professionals in the development of SDM tools(43) to better understand the feasibility when implementing them in the clinical encounter(44).