Approaches to SDM

SDM likely best occurs in the setting where a physician and a patient have an established relationship such that the physician knows the patient’s values and goals, informing how options are described and weighed during SDM.7 Even in the acute setting, such as the emergency department or during hospitalisation, understanding a patient’s background (eg, employment) and values prior to formal decision-making can provide important context for decisions.

There are multiple models for SDM with different numbers of outlined steps.8–10 When a clinical decision is needed, SDM starts by engaging patients and key supports in the process (table 1). This step requires the clinician to understand who the patient desires to participate in decision-making, such as a family member. In situations where a patient is incapacitated and unable to participate, SDM occurs with the surrogate decision maker. Even if a patient ultimately desires to defer the decision to a physician or family member, it is important to actively engage him/her in the SDM process.

Once patients and families are engaged, the second step (table 1) is specifically describing the decision and outlining the different options. In describing the different options, clinicians should use the available medical evidence to inform patients about the potential benefits and risks. It is also important to highlight when there is something unknown about the options (uncertainty) and to describe how the options are distinct. Differences between options include potential benefits and harms and considerations such as cost and invasiveness. This discussion should be individualised—for example, the balance of benefits and harms of clopidogrel for secondary stroke prevention will be different between a person with a prior history of bleeding gastric ulcers and a person with no such medical history.

Once the evidence is presented, the intersection between the options and the patient’s values and goals is explored (step 3, table 1). If a patient’s circumstances, values and goals were known prior to the initiation of SDM, the presentation of the options in step 2 should occur in that context. For example, if a patient lives alone and prioritises continued independence, the potential benefits and risks of each intervention are specifically presented with a reference to the likelihood of maintaining independence (eg, from successful treatment) or putting it at risk (eg, due to side effects). Regardless of prior knowledge of patient values, in step 3 a patient’s values as they specifically relate to the decision are explored. What is most important to the patient in this situation—expected functional recovery? Amount of risk? Cost?

Finally, a decision is made. Ideally, the patient makes the decision with the help of those friends or family members whom he or she has chosen for involvement. Sometimes patients prefer clinicians to make final decisions. In these circumstances, patients often still express a desire for participating in SDM,11 but request that the clinician select the best strategy after discussion. When the patient defers to clinicians, the burden is on the physician to target the decision to the patient’s stated values and goals, thus still using SDM to make the best individualised decision for that patient in that circumstance.

In many circumstances, particularly those encountered in the outpatient setting, re-evaluation is an important component of SDM. Anticipated ongoing benefits and risks may change based on the development of comorbidities; patients’ values and priorities may change based on their experiences with a medication or shifting life circumstances. In the outpatient clinic, for example, decisions regarding anticoagulation for secondary stroke prevention in the setting of atrial fibrillation should be reassessed over time. This is in contrast to certain acute stroke decisions, such as those regarding tissue plasminogen activator (t-PA), where the window for SDM is small, with little opportunity for re-evaluation.

SDM and stroke

Most recent publications on SDM in stroke care focus on oral anticoagulation for stroke prevention in atrial fibrillation.16–20 This is a decision where SDM clearly plays an important role given differences in individual risks based on comorbidities, multiple options with different potential benefits, risks, costs and time requirements (eg, for international normalized ratio [INR] monitoring), and obvious value assessments relating to potential outcomes such as stroke and bleeding.

Less research exists for other decisions relating to stroke, and currently available DAs may not meet decision aid standards.21 A 2013 review of patient tools designed for decision-making regarding thrombolytic treatment identified that available tools lacked key development stages, presented outcome probabilities poorly and failed to completely describe potential benefits and risks.22 Subsequently, the COMPuterized decsion Aid for Stroke thrombolysiS (COMPASS) tool, a computerised DA for thrombolysis in acute stroke, was developed with clinicians, patients, families and modelling techniques. Using the tool took a median time of only 2.8 min in early pilot testing, but additional study is required.23 24 DAs have particular potential for improving care in this emergent setting, where SDM is challenged by the time limitations for effective thrombolysis, the need to engage patients and families and convey knowledge in the context of the shock and effects of an acute stroke, and the need to incorporate personal values into a decision that relies heavily on physician expertise.25